Open Access
Add to BibSonomyAbstract
In the search for alternative materials to replace indium-tin-oxide in transparent electrodes we have structured copper and aluminum thin films (between 5 an 40 nm) for tailoring their optical properties. Micrometer scaled holes were produced using the direct laser interference patterning (DLIP) technique. We compared the optical and electrical parameters of nanosecond and picosecond processed thin films. It was found that the optical transmittance of the structured layers was relatively increased between 25 to 125% while the electrical resistance was marginally influenced. In addition, the laser treatment enhanced the diffuse to total transmission ratio (HAZE) by values ranging from 30 to 82% (relative) as a potential advantage of μm structuring. The results also show that both of the studied metals succeed to match the target which is set by typical applications of indium thin oxide (ITO) films. Furthermore, numerical simulations are performed in order to understand the ablation process of thin film material for ps and ns pulses.
© 2016 Optical Society of America
1. Introduction
The growing field of opto-electronics with a broad variety of applications requires electrodes with a sufficiently high electrical conductivity and an optical transmittance similar to plain glass. Presently, vacuum-sputtered indium tin oxide (ITO) is used for the state of the art production of thin film electrodes, for example in organic photovoltaic modules or organic LED-devices [1–4]. Depending on the layer thickness and the deposition method ITO coatings present optical transmittances in the visible spectrum from 80 to 91% and electrical sheet resistances between 5 and 100 Ω/sq [1, 2, 5]. However, the price of indium as a rare earth material has constantly increased since the beginning of the so called polytronics era. In consequence, there is a strong demand to replace ITO-based electrodes with alternative coatings [2, 5, 6].
As an alternative material for ITO several studies have been focused on the development of aluminum doped zinc oxides (AZO) films. These studies could demonstrate in some cases optical transmittances even higher than ITO with similar electrical conductivities [7–9]. However, AZO electrodes have shown lower mechanical stability as they tend to crack in continuous stress tests [8,10]. Optical properties of materials can be also modified by periodic surface patterns in the micrometer or submicrometer range [11, 12]. This was demonstrated for instance in the fabrication of structured Si-thin film solar cells using interference lithography [13, 14]. However, due to its multiple (especially chemical) step process, this method cannot be applied for a high throughput fabrication.
Alternatively, a direct production of surface patterns can be achieved by the application of high-energy short pulse laser systems with a coherence length in the centimeter range. A periodic interference pattern is produced by an angular overlap of two or more coherent laser beams. In contrast to the interference lithography a selective and periodic ablation process is obtained for the irradiated material [15]. Thus, the so called Direct Laser Interference Patterning (DLIP) is a one step fabrication process [16]. Furthermore, the DLIP method is a low cost (about 1 USD/m2) technology with great flexibility that works with fast speed (about 0.5 – 1 m2/min) on a broad variety of materials. Especially for spatial periods around 5 μm the DLIP process is about 10 times faster than established laser scanning methods [17, 18]. Furthermore, the resolution of scanning based technologies is generally limited to 5 – 15 μm, that can be overcome using DLIP. The structuring happens inside an interference volume where a vast multitude of holes or lines is created with each single laser pulse.
Beneath studies following the usage of alternative transparent conductive oxides (TCOs) as replacement materials for ITO, a variety of different approaches have been researched throughout the last decade. Two alternative concepts for electrode design were followed. On the one hand many works reported on the production of a web of nanowires or fibers and subsequent lamination onto a substrate. On the other hand the structuring of a coated substrate by etching, embossing or laser processing were introduced. Since the number of materials for transparent electrode applications is low, it turns out, that many of the solution approaches work with surface modifications. One possible candidate is currently seen in the nanowire technology. Wires with diameters of several Ångstromes provide nearly ideal electrical conductivity due to quantum mechanical effects [19, 20]. Considering the fact that such wires would cover the surface of a substrate like a disordered network, these ultrathin nanothreads would allow a large part of light to pass through. There are two different major ways of creating nanowire-based transparent electrodes, a wet-chemical synthesis and a lithography method [21–28]. For both of these techniques, nanowires require pre-production phase, which includes either a wet-chemical process or a multi-step procedure including exposure-, developing- and cleansing procedures. This leads to an increase in process complexity compared to simple metal or metal oxide coating procedures. As another fact, the fabrication of nanowire coatings is more time consuming than the most of the established vacuum-based deposition solutions [20,29]. Since silver is one state of the art material for nanowires, the cost factor will eventually play a role for a production in bulk. More cost-efficient materials like e.g. Aluminum could help to keep the production costs low. Percolated films using self-assembly of nano particles are another approach to produce transparent thin film electrodes. In this four-step process, an amount of polymer-based nanospheres are spread evenly on a metal film or a TCO-substrate. During the distribution phase, the nanoballs start to assemble in a hexagonally ordered hole-like pattern. In order to fix this texture on the substrates surface, an etching agent is applied. The electrode area for this method is limited to several square centimeters [30–32], while DLIP in combination with continuous vacuum deposition can achieve more can 1 m2/min using two single steps.
A very effective surface modification are nano-imprinted networks where a nano-structured embossing roller acts as imprint tool which can be used in a fast and reliable roll-to-roll process. However, the runtime of an embossing roller is limited while a replaced tool requires an invest of approximately 100,000 USD per roller. Presumably, the repetition rates of existing pulsed DLIP laser systems operating at the mJ-level can be scaled to several 10 kHz allowing then a similar production speed. Furthermore, the pattern, the structuring depth and the spatial period can easily be changed [18, 33, 34].
In this study, advanced thin film metallic electrodes consisting on periodic arrays of holes are fabricated using DLIP. Coatings of copper and aluminum with thicknesses between 5 and 40 nm on float glass are exposed to three-beam-interference patterns using ps and ns-pulsed laser sources. These materials have been selected due to their relative low cost compared to other metals such gold or silver. In order to analyze the temporal evolution of the temperature distribution within substrates and thin films initiated by a laser pulse we performed a numerical simulation. Instead of a complex in situ observation the model allows an approximate assessment of the obtained experimental results. The topographical, optical and electrical properties of the films are systematically characterized as function of the laser processing parameters as well as the film thickness. Finally, the properties of the produced electrodes are compared to standard properties of ITO electrodes.
2. Interference structuring of Cu and Al thin films
Thin metallic films of Al and Cu were deposited on squared float glass substrate (1 in2) with film thicknesses ranging from 5 to 40 nm, using physical vapor deposition method (PVD). For the experiments, two different solid state laser systems were utilized, with pulse durations of 6 ns and 35 ps. The longer pulses (6 ns) were provided by a Nd:YAG laser system with a wavelength λ = 1064 nm. The picosecond laser system is based on an Yb:YAG crystal disk with a fundamental wavelength of λ = 1030 nm. To perform the interference experiments, a diffractive optical configuration was used [16, 35].
Figure 1 shows the experimental setup. The principal laser beam is split into four equivalent symmetrical partial beams by a 2×2 diffractive optical element (DOE). After splitting, the partial beams are sent through a tailored pyramid prism to parallelize the beams before they pass through a convex focusing lens. In order to obtain a three-beam setup, one individual beam was blocked. The number of laser beams was selected in order to produce a periodic arrange of holes which is necessary to maintain the electrical conductivity of the films in two dimensions. The lens guides the three remaining beams towards the interfering zone were they overlap. By controlling the angle between the beams, the spatial period of the interference pattern was changed between 1.7 and 2.7 μm. The smallest hole size was 300 nm at a spatial period of 1.7 μm with pulse energies ≥100% AT (ablation threshold). The largest holes had a diameter of 2.3 μm at a period of 2.7 μm and a laser fluence of 200% AT. The hole diameter of all produced structures averaged between 40 – 60% regarding to the spatial period.
Fig. 1 Experimental setup: The spatial period Λ is changed by moving the prism along the A-axis. The beam profile is optimized by moving the substrate along the Z-axis.
To treat the whole surface of the substrate a precise translation stage system was used to move the sample in the XY-plane. A separation of 800 μm was chosen to position Gaussian-shaped laser spots with a beam diameter (1/e2) of approximately 450 μm. The distance between the spots was optimized in order to keep the global intensity profile homogeneous and to avoid the destruction of the film at multi-shot illumination. For a global analysis of the electrical and optical properties a samples size of 6 mm× 6 mm was selected. The optical properties of the films (total and diffuse transmittance) were measured with a photo spectrometer (Shimadzu UV-3100//MPC-3100). The sheet resistance was determined using the 4-point probes method. Surface characterization of the films was performed using an Scanning Electron Microscope operating at 5 kV (A NEON 40ESB) as well as an Atomic Force Microscope (Jeol JSPM 5200).
2.1. Structuring with ns-laser pulses
Figure 2 shows characteristic atomic force microscope (AFM) and scanning electron microscope (SEM) images of processed Al thin metallic films (20 nm) using 6 ns laser pulses. The non-symmetrical hexagonal geometry of the pattern results from the utilized configuration of the laser beams with non-symmetric azimuthal angles. The oval shape of the holes is also the result of this configuration [35]. Due to the high energy density at the interference maxima positions, the material is locally heated and ablated. Ablation effects that involved the glass substrate (also in case of ps pulses) are not critical to the opto-electrical active layer of organic electronics, since the thin film stack is deposited onto the structured electrode. Thus, the stack surface forms like the base substrate which gives no drawbacks for the device (also shown in [36]).
Fig. 2 (a – c) AFM images of structured aluminum thin films with 20 nm thickness and a spatial period of Λ = 1.7 μm. The surfaces were irradiated at (a) 1.0 J cm−2, (b) 1.3 J cm−2 and (c) 1.6 J cm−2. In (d), a SEM image of an Al sample exposed to 1.3 J cm−2 is shown. The pulse duration was 6 ns.
The structuring of aluminum thin films resulted in regular patterns without a significant amount of debris or residual particles. In Fig. 2(a) – 2(c), AFM images of treated 20 nm thin films are displayed. Each surface shown in the figure was exposed to a different average laser fluence (energy per unity of area): (a) 1.0 J cm−2, (b) 1.3 J cm−2 and (c) 1.6 J cm−2. It can be seen that with increasing laser fluence, the diameter of the produced micro-holes also increases due to the stronger local ablation of the film at the maxima. In consequence, the size of the interstitial metal bridges, which are necessary to build the conductive and transparent network of the surface electrode, shrink with increasing laser fluence.
In addition, the surface roughness of the Al films after the laser treatment is very low (Ra < 3 nm) what can be explained by the strong melting of the film. Furthermore, transport of the melt from the interference maxima to minima positions can be observed, especially for high laser fluences (> 1.2 J cm−2). This effect is demonstrated by the square-like characteristic of the produced patterns, even showing pillars of re-solidified material at the minima positions, that can be only explained by melt transport (see Fig. 2(c)). The driven force for this process can be attributed to Marangoni convection induced by surface tension gradients which at the same time result from the temperature difference between maxima and minima positions [37].
The SEM image in Fig. 2(d) shows a larger section of the thin film displayed in Fig. 2(b). Here, the smooth surface of the treated Al film can be confirmed. In addition, some spherical particles of solidified melt can be observed. The image also shows that the bottom holes correspond to the glass substrate, revealing that the metal layer was completely removed during the ablation process.
For the 20 nm layer, the maximal structure depths was 60 nm, corresponding to a laser fluence of 1.6 J cm−2 (spatial period Λ = 1.7 μm). The crater ribs featured a height of 15 nm (material over the surface level of the unstructured film). The calculation of the ablated volume and the volume of the crater corona reveal that approximately only 1.0 to 2.5% of the molten material remains at the interference maxima especially in the peripheral zones surrounding the holes. Thus, the major part of the material is completely ablated from the film. Similar results were observed for other layer thicknesses (5 – 40 nm).
In the case of the Cu thin films, the DLIP treatment produced different results, mainly depending on the spatial period. For a pitch of 1.7 μm, the copper layer was totally detached from the glass substrate. A typical AFM image of such a surface is shown in Fig. 3(a). In this case, the formerly 20 nm thick copper layer was totally molten at a laser fluence of 2.1 J cm−2. The residual melt spread irregularly on the substrate. This effect can be explained from the relatively high surface tension of molten copper that is 1.36 J m−2 at a temperature of 1360 K [38]. Since aluminum has a much lower surface tension of 0.87 J m−2 at its melting point [39], the adhesion to the float glass is stronger and the metallic thin film sticks to the substrate material despite the fact that it is completely liquefied for a short period of time.
Fig. 3 AFM images of ns-structured copper thin films with a thickness of 20 nm. (a) Cu film treated with a single laser pulse at 2.1 J cm−2 and a spatial period of 1.7 μm. In this case, the whole metal layer has been molten by the laser treatment. (b) Cu film treated with 1 laser pulse at 2.3 J cm−2 and a spatial period of 2.7 μm. Although the textures are characterized by many irregularities, the holes are pronounced clearly. The pulse duration was 6 ns.
For larger spatial periods, e.g. Λ = 2.7 μm, the hole-like pattern is clearly pronounced, although it shows several flaws and irregularities (Fig. 3(b)). The dependence of the pattern quality on the pitch and the pulse duration (6 ns) can be related with the high thermal conductivity of copper. The longer the laser pulse, the more heat can flow laterally through the metal layer. Therefore, the possibility of totally melt the Cu film increases with reducing the spatial period.
2.2. Structuring with ps-laser pulses
Figure 4 shows characteristic scanning electron microscope (SEM) images of processed Al and Cu thin metallic films using ps-laser pulses. In the case of the Cu films, also hole-like periodic structures with smaller spatial periods (Λ = 2.0 μm) could be fabricated (see Fig. 4(a)) in comparison to the ns-pulse experiments. Figure 4(a) also shows light delamination effects on the edges of the holes (sample irradiated with 0.76 J cm−2). In addition, a small and narrow amount of solidified material can be observed at the ablated regions forming a corona and particles of metal can be found inside the holes. The melt particles vary in size between 50 and 250 nm in diameter. The surrounding area of the holes remains unaffected, what is demonstrated by the unchanged surface roughness of the film (nanocrystalline structure of the copper grains).
Fig. 4 Surface topography of ps-structured Cu (a, c) and Al (b, d) thin films with a spatial period of Λ = 2.0 μm. Samples irradiated with a single laser pulse and laser fluences of (a) 0.76 and (c) 0.90 J cm−2 for Cu, (b, d) 0.54 J cm−2 for Al. The pulse duration was 35 ps.
Figure 4(c), depicts an AFM image of a Cu sample irradiated at 0.90 J cm−2. The residual particles of solidified material can also be observed in this image and their diameters range in the same size class. The structure depths depended strongly on the laser fluence at which the metallic films were exposed. For the 20 nm film thickness, the structure depth varied between 20 and 28 nm, corresponding to laser fluences between 0.61 and 0.93 J cm−2, respectively.
The results for ps-ablated aluminum films show a different behavior. The SEM-scan in Fig. 4(b) also displays oval shaped holes, even more regular as they were typical for the copper ablation. However, in this case the edge zones of the holes are softened and do not show a sharp morphology as in the case of copper. There are no observable delamination effects and the residual melt is negligible. The pattern depths for ps-ablated aluminum (20 nm) ranged between 750 nm, corresponding to laser fluences between 0.43 and 0.65 J cm−2.
3. Simulation
Numerical simulations of the laser treatment were realized utilizing the model reported in [40]. The calculations are based on a numerical solution of the two-dimensional heat diffusion equation, taking into consideration the energy required to melt and/or vaporize the material:
with T being the temperature, ρ the density, cp the specific heat, χ the thermal diffusivity, and qL, qF and qV are the laser heat flow of the interference pattern, the melting and vaporization enthalpies, respectively. The interference pattern is described shown in where I0 is the peak intensity of both laser beams together, k is the wave vector, and α is the angle between the two interfering beams. The magnitude qL is finally calculated combining (2), together with the linear absorption of the material and the temporal shape of the laser beam where R stands for the optical reflection of the material, tp for the pulse duration and .Numerical calculations of the temperature distribution profiles of Al and Cu films were performed to understand the mechanism of the structuring process. The thermal properties were taken from [41–44] and the specific optical properties from [45,46]. The spatial period was set to 1.7 μm. The simulated pulse duration amounted 6 ns at 1.5 J cm−2 for aluminum and 35 ps with a fluence of 0.9 J cm−2 for copper thin films.
Figure 5(a) – 5(c) show the results of the calculation for a 20 nm thick aluminum-on-float-glass substrate for a 6 ns laser pulse. Figure 5(a) visualizes the temporal evolution of the temperatures at interference minima and maxima. The simulation explains that the film is strongly heated at the maxima positions but also the temperature at the minima increases due to lateral thermal conduction. Furthermore, 40 ns after the pulse arrival the temperature at maxima and minima reach values close to 1250 K. Figure 5(b) and 5(c) show the temperature distribution inside the film and the molten regions, respectively. These graphs including a lateral (with 5 or 8 spatial periods) and an altitude dimension over the film correspond to 10 ns after the pulse arrival. A maximum temperature of up to 1700 K can be predicted while the interference minima are heated up to 800 K. Due to the high temperatures, the material at the interference maxima is strongly molten (melting point of Al: 933 K). Furthermore, also the glass substrate under the metallic film is molten (up to 40 nm in depth). At the interference minima the situation is different. During the first 10 ns after pulse arrival the material remains solid (as seen in Fig. 5(c)) while it becomes melted lateron. Similar results were observed for Cu-films irradiated with 6 ns pulses (not presented here).
Fig. 5 Numeric simulation of DLIP-based ablation on metallic thin films with a thickness of 20 nm. Four monitoring points i – iv are set while points i and iii are situated at an interference maximum and points ii and iv at a minimum. Points i and ii are placed at the film surface and iii and iv 20 nm below the surface. The pulse arrival corresponds to t = 0 and a spatial period of 1.7 μm was chosen. (a – c) Results for aluminum films at nanosecond pulses (6 ns) with a fluence of 1.5 J cm−2; (d – f) Results for copper films at picosecond pulses (35 ps) with a fluence of 0.9 J cm−2; Temperature distribution of five spatial periods (b, e) and molten regions of eight interference periods (c, f) at 10 ns for ns-pulses and 50 ps for ps-pulses.
The results of the thermal simulation of copper thin films for 35 ps laser pulses are presented in Fig. 5(d) – 5(f). Differently from the ns-pulses simulation, the temperature at the interference minima does not reach the melting point after the laser heat treatment (Fig. 5(d)). The temperature at the minima shows a slow rise and distinctly stays below the copper melting point at a maximum of 650 K. The temperature distribution within the sample (see Fig. 5(e)) behaves differently compared to ns as well. The temperature gradients are stronger and the depth of penetration by the heat flux is lower than for aluminum simulation (see Fig. 5(a) – 5(c)). The temperature in the Cu-film ranges between 300–2470 K. An interesting aspect to mention is the vaporization (ablation) of the glass substrate occurring at the same time where the copper film is molten (see Fig. 5(f)). In this case, the pressure from the evaporated glass substrate pushes the molten copper zones out with ejection of molten particles. Some of these particles can be observed in Fig. 4(a).
According to these simulation results it can be stated that for ps-DLIP the temperature gradient between the maxima- and minima zones is much higher than in the case of ns-DLIP. Due a longer light-material interaction at ns the glass substrate is molten up to higher depths. Furthermore, thermal diffusion within the aluminum film leads to temperatures above the melting point also at interference minima. For ps-DLIP simultions, the interference minima zones remain solid while the glass substrate is molten only up to 10 nm.
4. Experimental results
4.1. Optical properties
For the evaluation of optical properties, the total and the diffuse transmittance spectra of the laser treated films were determined. A spectral range between 350 – 800 nm was selected where organic photovoltaic (OPV) elements generally achieve their highest efficiencies and the emission spectrum of organic light emitting diodes (OLEDs) is also included. The information depicted in the spectra (e.g. see Fig. 6(a), 6(b), 6(e) and 6(f)) can be quantified in a single data point by integrating the optical transmission (total and diffuse) of the films in the whole measured wavelength range and comparing it to the Air Mass (AM) sun spectrum using
where τgS represents the transmittance weighed by the AM1.5 sun spectrum S(λ) [47]. The diffuse transmission τd of the films can be also characterized by calculating the ratio between the diffuse and total transmission (haze). In the case of periodic surface structures the haze is correlated to the diffraction efficiency. Direct τ0 and total τt transmittance were analyzed using a Shimadzu UV-3100//MPC-3100 photo spectrometer.Fig. 6 Optical characterization of DLIP structured metallic thin films, for aluminum (a – d) and copper (e – h) films; Representative total (a, e) and diffuse (b, f) transmission spectra; AM1.5 weighted total transmission (c, g) and haze as function (d, h) vs. layer thickness (5 – 40 nm) at various laser process parameters (spatial period, laser fluence, pulse duration). The used laser fluences are given in terms of the ablation threshold (AT) for a better representation of the results.
From the point of view of applications, high haze values are desirable to obtain increased optical path lengths by diffraction leading to larger solid angles within the radiation active layers. This permits improved efficiencies of optical devices such as solar cells or light emitting diodes (organic or inorganic) [11, 36]. Figures 6(a) and 6(b) show representative optical transmission spectra of treated Al-thin films (15 nm layer thickness, 1.7 μm spatial period, 6 ns pulse duration) treated at different laser fluence values. The used laser fluences are given in terms of the ablation threshold (AT) for a better representation of the results. The incidence of light for the measurement of the transmittance was perpendicular to the substrate. The diffuse transmittance was measured with an integrating sphere including diffraction angles between 5 – 175°, while the samples were always illuminated on the patterned side.
In general, it can be observed that the laser treatment is capable to increase both, the total and diffuse transmission of the films. The result of the weighted optical spectra (AM 1.5) as function of the layer thickness (5 – 40 nm) and the laser treatment process parameters (spatial period, laser fluence, pulse duration) are shown in Fig. 6(c) and 6(d). The maximal transmittance for ns-structured samples at a pitch of 1.7 μm is gained at a film thickness of 5 nm with a value of 71%. At 20 nm layer thickness, there is a local maximum of 79%. For 2.7 μm spatial period, a maximum of 77% can be found for 15 nm thin films. This can be explained by the observation of a minimum ablation threshold for film thicknesses around 20 nm. The corresponding haze values (see Fig. 6(d)) show the best diffraction characteristics for layer thicknesses of 15 and 20 nm. For a spatial period of 1.7 μm, the maximum haze was 61% while for 2.7 μm an overall diffraction efficiency of approximately 73% is reached. The improvements in the diffuse transmission can be explained by the diffraction effects of the transmitted light induced by the well ordered periodic structures. The aspect ratio (quotient between the structure height and spatial period) of the produce hole-like patterns was in the range of 0.03 to 0.06, which are relatively high considering the small film thickness [6, 14, 16]. The diagrams also show that both, total transmittance and haze characteristic, rise with increasing laser the fluence. This effect can be related with a higher rate of material removal.
In the case of the Al-films structured with ps laser pulses, relatively high transmittances (between 64 – 80%) have been observed. On the other hand, the measured haze values were significantly lower (15 to 23%) compared to the ns-treated samples. This can be explained by the characteristic ablation process observed for the ps-pulses. In the last case, a negligible amount of melt is produced and thus, forming periodic structures with lower height and a lower diffraction efficiency.
Similar results were also observed for the laser treated Cu-films. Examples of characteristic total and diffuse transmittance spectra of 15 nm copper samples are shown in Fig. 6(e) and 6(f). The total transmittance and the haze values for film thicknesses from 5 to 40 nm and treated with ps and ns-pulses are shown in Fig. 6(g) and 6(h).
The measured total transmittance has a maximum for a 5 nm film thickness and drops continuously with the layer thickness (Fig. 6(e)). Compared to the Al-film, any significant improvement could be obtained using ns-laser pulses. Furthermore, the differences in the total transmittance of ns-treated and non-treated films is not as high as for the case of Al.
On the other hand, the ps-experiments produced patterns with total transmittances ranging from 41 to 82%, which correlate with the increasing intensity. The maximum haze measured for the ps-treated films was 25% for a film thickness of 20 nm. This is related to a high quality of the structures that are not completely molten due to the shorter laser interaction time of the ps compared to the ns pulses.
4.2. Electrical properties
The sheet resistance of the exemplary processed Al and Cu thin films is shown in Fig. 7 as function of the film thickness and the laser processing parameters. The sheet resistances of the untreated films start at 1 – 2 Ω/sq for 40 nm (see reference in Fig. 7) and increase with a reduced film thickness up to 50 – 70 Ω/sq for 5 nm layers. This behavior is related to the increasing number of grain boundaries with decreasing the film thicknesses as reported in [48].
Fig. 7 Electrical characterization of laser treated (a) Cu and (b) Al thin films with spatial periods of 2.7 μm for ns- and 2.0 μm for ps-pulses.
For the laser treated samples, different effects were found. In general, an increase of the electrical resistance is observed with increasing the laser fluence. This can be explained by the reduction of available material at the electrode due to the selective ablation process. On the other hand, the remained conductivity of the films denotes that the hexagonal shaped metallic network is still interconnected. If still higher laser fluences are used, this network is destroyed and thus significantly increasing the sheet resistance (e.g. larger than 30,000 Ω/sq for 15 nm Cu films irradiated at > 1.2 J cm−2).
In the case of the Al electrodes treated with ns-pulses, low variations in the sheet resistance were observed for low and moderate laser fluences. This behavior is different from the Cu films, where the electrical resistance was increased at all investigated laser parameters (see Fig. 7). This behavior was observed especially for film thicknesses between 20 and 40 nm.
In the case of ps-pulse laser experiments, sheet resistances values between 5.5 – 1,050 Ω/sq were observed for Al, corresponding to 20 nm films irradiated with laser fluences between 0.43 and 0.65 J cm−2, respectively. For the Cu films (also 20 nm thick), the sheet resistance varied between 4.6 to 155 Ω/sq for laser fluences between 0.62 and 0.93 J cm−2, respectively.
The differences between the observed behaviors for the samples irradiated with ns and ps laser pulses can be also be explained from the thermal simulations. For ps-pulses, the heat remains selectively at the interference maxima positions due to the very short available times for heat diffusion. In consequence, the electrical resistance increases since less material is available to transport electrons. In case of ns pulses, thermal diffusion leads to melting of the surface or recrystallization of the nano-crystalline films also at the interference minima as demonstrated by the thermal simulations. Both effects produce a positive contribution in the electrical conductivity since the amount of grains (and therefore of grain boundaries) is reduced. Therefore, this effect compensates the reduction in the electrical conductivity of the films due to the ablation process. Experimental evidence of recrystallization effects induced by interference patterns using ns-laser pulses were also reported before [49].
4.3. General performance
Ideal transparent electrodes for example usable for organic electronic devices simultaneously require high values of electrical conductivity, optical transmission, and haze. To evaluate the general performance of the produced metallic thin films electrodes, the optical and electrical properties of selected treated films were summarized in Fig. 8. According to that, an ideal electrode is placed in the frontal upper right corner of the diagram (low sheet resistance, high transmittance and haze).
Fig. 8 Optical and electrical properties of selected Al and Cu films, structured using ps and ns-pulses. Only structured substrates exceeding the limits of 40% in total transmittance and sheet resistances bellow 250 Ω/sq, with a haze higher than 5% are shown. The target range denotes electrical and optical parameters of ITO thin film electrodes given in [50–54].
Characteristic ITO-based transparent electrodes are also included in this diagram as comparison [3,8,9,55]. Treated metallic electrodes which might substitute ITO materials, should have electrical resistance below 150 Ω/sq, total optical transmittance higher than 60% (preferentially over 80%). It can be observed that all ns-structured substrates stand outside the region of interest. Two electrodes patterned with medium laser fluence at Λ = 2.7 μm are situated very close to the target region. The haze achieved with ns-ablated aluminum samples could be measured between 35 – 78%. In the case of the Cu electrodes, also treated with ns-pulses, excellent electrical resistivity properties could be obtained (below 25 Ω/sq). However, these films presented very low optical transmittance (∼ 40%).
Among the ps-structured substrates, two samples could match the properties of existing ITO electrodes: one Cu thin film (20 nm) exposed to 0.9 J cm−2, and one Al electrode (also 20 nm) irradiated at 0.75 J cm−2 of laser fluence. In addition, due to the periodic geometry of the fabricated pattern, the structured samples provide a total transmission of 80% and haze values of 19 and 34% for Al and Cu, respectively.
5. Conclusion
Aluminum and copper thin films deposited on glass substrates were processed using three-beam direct laser interference patterning. Using this method, hole-like structures with spatial periods between 1.7 to 2.7 μm were produced. The measurements of topographic characteristics in combination with the results of finite element method based heat-flow simulation revealed two different ablation mechanisms for ns- and ps-laser pulses. For metallic films irradiated with ps pulses, the structuring process is mainly driven by local vaporization effects of film and the glass substrate in combination with convective melt flows and evaporation pressure from inside the melt bath. For films irradiated with ns pulses, both the film and the glass substrate are molten. Due to the longer time for thermal diffusion, the metallic film is also molten at the interference minima positions, specially for shorter spatial periods.
Concerning the optical and electrical performance of the produced electrodes, it could be shown that the laser treatment method permitted to increase both, the total and diffuse transmissions significantly compared to unstructured electrodes. Best performances were obtained for electrodes irradiated with ps-laser pulses. Furthermore, among the ps-structured substrates, two conditions could match the properties of existing ITO electrodes, with electrical sheet resistance between 25 and 50 Ω/sq, transmittances up to 81% and a haze between 19 and 34%.
We observed an optimum performance of the copper and aluminum films with thicknesses between 15 – 25 nm. Based on the results of preceded ablation test series, a minimum in the ablation threshold was confirmed for this thicknesses as well. This can be explained by an optimum between the film thickness and the coupling of the laser pulse energy into the layer. In this way, a relative maximum of material can be ablated at a minimum fluence. Thus, structures with the highest quality and an optimal ratio of spatial period and hole diameter can be produced, which also explains the higher transmission and haze rates achieved here.
Although the DLIP-structured electrodes cannot show the highest field performance regarding to transmittance and sheet resistance, they provide 20 – 28% haze, which is a major requirement for light harvesting in OPV [36] and for light de-coupling in OLED devices [56]. Future work is now in progress to investigate picosecond structuring of metallic films with different thicknesses.
Acknowledgments
A. Lasagni and M. Siebold acknowledge the German Ministry of Economics and Technology (BMWi) through the German Federation for Industrial Research (Arbeitsgemeinschaft industrieller Forschungsvereinigungen, AiF) under IGF Contract 18359 N for partial financial support. S. Eckhardt acknowledges the European Social Fonds (ESF) and the Sächsische Aufbaubank (SAB).
References and links
1. J. Meiß, New Material Concepts for Organic Solar Cells (Optimus Mostafa Verlag, Göttingen, 2011).
2. N.S. Sariciftci and S.-S. Sun, Organic Photovoltaics: Mechanism, Materials, and Devices (Taylor & Francis, Boca Raton, 2005). [CrossRef]
3. B. Kang, J. Shi, Q. Gao, in 2011 International Conference on Electrical and Control Engineering (ICECE): Sept. 16–18, 2011, Yichang, China Proceedings ( Institute of Electrical and Electronics Engineers (IEEE)), IEEE. [Piscataway, N.J.] 2011, p. 3017.
4. D.S. Ghosh, T.L. Chen, N. Formica, J. Hwang, I. Bruder, and V. Pruneri, “High figure-of-merit Ag/Al:ZnO nano-thick transparent electrodes for indium-free flexible photovoltaics, Solar Energy Materials and Solar Cells,” 107, 338–343 (2012).
5. Polinares Consortium, Fact Sheet: Indium (European Research Area, Dundee, 2012).
6. H. Paul, Lexikon der Optik (Spektrum, Akad. Verl., Heidelberg, Berlin, 2003).
7. S. Ray, R. Banerjee, N. Basu, A. K. Batabyal, and A. Barua, “Properties of tin doped indium oxide thin films prepared by magnetron sputtering,” J. Appl. Phys. 54, 3497 (1983). [CrossRef]
8. Y. Waseda and J. M. Toguri, The Structure and Properties of Oxide Melts: Application of Basic Science to Metallurgical Processing (World Scientific, Singapore, River Edge, 1998). [CrossRef]
9. M. Yamaguchi, A. Ide-Ektessabi, H. Nomura, and N. Yasui, “Characteristics of indium tin oxide thin films prepared using electron beam evaporation,” Thin Solid Films 447, 115–118 (2004). [CrossRef]
10. X. Pang, H. Ma, K. Gao, H. Yang, X. Wu, and A. A. Volinsky, “Fracture Toughness and Adhesion of Transparent Al:ZnO Films Deposited on Glass Substrates,” J. Mater. Eng. Perform. 22, 3161–3167 (2013). [CrossRef]
11. E.G. Loewen, M. Nevière, and D. Maystre, “Grating efficiency theory as it applies to blazed and holographic gratings,” Appl. Opt. 16, 2711–2721 (1977). [CrossRef] [PubMed]
12. K.Q. Le and A. Alù, “Plasmonic gratings for enhanced light-trapping in thin-film organic solar cells,” in Conference on Lasers and Electro-Optics 2012, OSA Technical Digest (online) (Optical Society of America, 2012), paper CF3J.5.
13. S. Sivasubramaniam and M. M. Alkaisi, “Inverted nanopyramid texturing for silicon solar cells using interference lithography,” Microelectron. Eng. 119, 146–150 (2014). [CrossRef]
14. Y.M. Song, J.S. Yu, and Y.T. Lee, “Antireflective submicrometer gratings on thin-film silicon solar cells for light-absorption enhancement,” Opt. Lett. 35, 276–278 (2010). [CrossRef] [PubMed]
15. B. Voisiat, M. Gedvilas, S. Indrisinas, and G. Raciukaitis, “Picosecond-laser 4-beam-interference ablation as a flexible tool for thin film microstructuring,” Physics Procedia 12, 116–124 (2011). [CrossRef]
16. A. Lasagni, T. Roch, M. Bieda, D. Benke, and E. Beyer, “High speed surface functionalization using direct laser interference patterning, towards 1 m2/min fabrication speed with sub-μm resolution,” Proc. SPIE 8968, 89680A (2014). [CrossRef]
17. M. Theuring, V. Steenhoff, S. Geißendörfer, M. Vehse, K. von Maydell, and C. Agert, “Laser perforated ultrathin metal films for transparent electrode applications,” Opt. Express 23, A254–A262 (2015). [CrossRef] [PubMed]
18. A. Lasagni, T. Roch, J. Berger, T. Kunze, V. Lang, and E. Beyer, “To use or not to use (direct laser interference patterning), that is the question,” Proc. SPIE 9351, 935115 (2015). [CrossRef]
19. Z.L. Wang, Nanowires and Nanobelts. Materials, Properties and Devices. Volume 1: Metal and Semiconductor Nanowires (Sringer, Boston, 2003).
20. A. Sedlmayr, Experimental Investigations of Deformation Pathways in Nanowires (KIT Scientific Publishing, Karlsruhe, 2012)
21. S. De, T.M. Higgins, P.E. Lyons, E.M. Doherty, P.N. Nirmalraj, W.J. Blau, J.J. Boland, and J.N. Coleman, “Silver Nanowire Networks as Flexible, Transparent, Conducting Films: Extremely High DC to Optical Conductivity Ratios,” ACS Nano 28, 1767–1774 (2009). [CrossRef]
22. H. Wu, L. Hu, M.W. Rowell, D. Kong, J.J. Cha, J.R. McDonough, J. Zhu, Y. Yang, M.D. McGehee, and Y. Cui, “Electrospun Metal Nanofiber Webs as High-Performance Transparent Electrode,” Nano Lett. 10, 4242–4248 (2010). [CrossRef] [PubMed]
23. L. Hu, H.S. Kim, J.-Y. Lee, P. Peumans, and Y. Cui, “Scalable Coating and Properties of Transparent, Flexible, Silver Nanowire Electrodes,” ACS Nano 4, 2955–2963 (2010). [CrossRef] [PubMed]
24. J. Groep, P. Spinelli, and A. Polman, “Transparent Conducting Silver Nanowire Networks,” Nano Lett. 12, 3138–3144 (2012). [CrossRef] [PubMed]
25. E.C. Garnett, W. Cai, J.J. Cha, F. Mahmood, S.T. Connor, M.G. Christoforo, Y. Cui, M.D. McGehee, and M.L. Brongersma, “Self-limited plasmonic welding of silver nanowire junctions,” Nature Materials 11, 241–249 (2012). [CrossRef]
26. G.Y. Margulis, M.G. Christoforo, D. Lam, Z.M. Beiley, A.R. Bowring, C.D. Bailie, A. Salleo, and M.D. McGehee, “Spray Deposition of Silver Nanowire Electrodes for Semitransparent Solid-State Dye-Sensitized Solar Cells,” Adv. Energy Mater. 3, 1657–1663 (2013). [CrossRef]
27. H. Wu, D. Kong, Z. Ruan, P.-C. Hsu, S. Wang, Z. Yu, T.J. Carney, L. Hu, S. Fan, and Y. Cui, “A transparent electrode based on a metal nanotrough network,” Nat. Nanotechnol. 8, 421–425 (2013). [CrossRef] [PubMed]
28. J.A. Fairfield, C. Ritter, A.T. Bellew, E.K. McCarthy, M.S. Ferreira, and J. J. Boland, “Effective Electrode Length Enhances Electrical Activation of Nanowire Networks: Experiment and Simulation,” ACS Nano 89542–9549 (2014). [CrossRef] [PubMed]
29. O. Eibl, Thermoelectric Bi2Te3 Nanomaterials (Wiley, Weinheim, 2015)
30. T. Nakanishi, E. Tsutsumi, K. Masunaga, A. Fujimoto, and K. Asakawa, “Transparent Aluminum Nanomesh Electrode Fabricated by Nanopatterning Using Self-Assembled Nanoparticles,” Appl. Phys. Express 4, 025201 (2011). [CrossRef]
31. F. Afshinmanesh, A.G. Curto, K.M. Milaninia, N.F. van Hulst, and M.L. Brongersma, “Transparent Metallic Fractal Electrodes for Semiconductor Devices,” Nano Lett. 14, 5068–5074 (2014). [CrossRef] [PubMed]
32. T. Gao, B. Wang, B. Ding, J. Lee, and P.W. Leu, “Uniform and Ordered Copper Nanomeshes by Microsphere Lithography for Transparent Electrodes,” Nano Lett. 14, 2105–2110 (2014). [CrossRef] [PubMed]
33. M.-G. Kang and L.J. Guo, “Nanoimprinted Semitransparent Metal Electrodes and Their Application in Organic Light-Emitting Diodes,” Adv. Mater. 19, 1391–1396 (2007). [CrossRef]
34. J. van de Groep, F. Gupta, M.A. Verschuuren, M.M. Wienk, R.A. Janssen, and A. Polman, “Large-area soft-imprinted nanowire networks as light trapping transparent conductors,” Sci. Rep. 5, 11414 (2015). [CrossRef] [PubMed]
35. S. Eckhardt, L. Müller-Meskamp, M. Loeser, M. Siebold, and A. F. Lasagni, “Fabrication of highly efficient transparent metal thin film electrodes using Direct Laser Interference Patterning,” Proc. SPIE 9351, 935116 (2014).
36. L. Müller-Meskamp, Y.H. Kim, T. Roch, S. Hofmann, R. Scholz, S. Eckardt, K. Leo, and A.F. Lasagni, “Efficiency Enhancement of Organic Solar Cells by Fabricating Periodic Surface Textures using Direct Laser Interference Patterning,” Adv. Mater. 24, 906–910 (2012). [CrossRef] [PubMed]
37. J. Lawrence, D.G. Waugh, and N. Aoyagi, Laser Surface Engineering: Processes and Applications (Woodhead Publishing, Cambridge, 2015).
38. G F. Deyev and D.G. Deyev, Surface Phenomena in Fusion Welding Processes (CRC Taylor & Francis, Boca Raton, 2006).
39. J.E. Hatch, Aluminum: Properties and Physical Metallurgy, American Society for Metals (Metals Park, Ohio, 1984).
40. A. Lasagni and F. Mücklich, “FEM simulation of periodical local heating caused by Laser Interference Metallurgy,” J. Mater. Process. Tech. 209, 202–209 (2009). [CrossRef]
41. S.L. Chaplot, R. Mittal, and N. Choudhury, Thermodynamic properties of solids: Experiment and modeling (Wiley-VCH, Weinheim, 2010). [CrossRef]
42. R. Hultgren, R.L. Orr, P.D. Anderson, and K.K. Kelley, Selected Values of the Thermodynamic Properties of Metals and Alloys (John Wiley & Sons Inc., New York, Heidelberg, 1963).
43. IAEA, Thermophysical Properties of Materials for Nuclear Engineering: A Tutorial and Collection of Data (International Atomic Energy Agency, Vienna, 2008).
44. L.D. Pye, A. Montenero, and I. Joseph, Properties of Glass-Forming Melts (Taylor & Francis, Boca Raton, 2005). [CrossRef]
45. M. Wakaki, K. Kudo, and T. Shibuya, Physical Properties and Data of Optical Materials (CRC Press, Boca Raton, 2007). [CrossRef]
46. L. Ward, The Optical Constants of Bulk Materials and Films (Institute of Physics Pub, Bristol, Philadelphia, 1994).
47. K. Mertens, Photovoltaics: Fundamentals, Technology and Practice (John Wiley & Sons, 2014).
48. A.F. Mayadas and M. Shatzkes, “Electrical-Resistivity Model for Polycrystalline Films: the Case of Arbitrary Reflection at External Surfaces,” Phys. Rev. B 1, 1382 (1970). [CrossRef]
49. G. Aichmayr, D. Toet, M. Mulato, P. V. Santos, A. Spangenberg, and R.B. Bergmann, “Growth mechanisms in laser crystallization and laser interference crystallization,” J. Non-Cryst. Solids 227, 921–924 (1998). [CrossRef]
50. T. Suzuki, T. Yamazaki, and H. Oda, “Properties of indium oxide/tin oxide multilayered films prepared by ion-beam sputtering,” J. Mater. Sci. 23, 3026–3030 (1988). [CrossRef]
51. Y. Akagi, K. Hanamoto, H. Suzuki, T. Katoh, M. Sasaki, S. Imai, M. Tsudagawa, Y. Nakayama, and H. Miki, “Low-Resistivity Highly Transparent Indium-Tin-Oxide Thin Films Prepared at Room Temperature by Synchrotron Radiation Ablation,” Jpn. J. Appl. Phys. 38, 6846–6850 (1999). [CrossRef]
52. H. Agura, A. Suzuki, T. Matsushita, T. Aoki, and M. Okuda, “Low resistivity transparent conducting Al-doped ZnO films prepared by pulsed laser deposition,” Thin Solid Films 445, 263–267 (2003). [CrossRef]
53. R. Ghodssi and P. Lin, MEMS Materials and Processes Handbook (Springer, Boston, 2011). [CrossRef]
54. H. Askari, H. Fallah, M. Askari, and M.C. Mohmmadieyh, “Electrical and optical properties of ITO thin films prepared by DC magnetron sputtering for low-emitting coatings,” eprint arXiv:1409.5293 (2014).
55. D.S. Ghosh, Ultrathin Metal Transparent Electrodes for the Optoelectronics Industry (Springer, New York, 2013). [CrossRef]
56. C. Fuchs, T. Schwab, T. Roch, S. Eckardt, A. Lasagni, S. Hofmann, B. Lssem, L. Müller-Meskamp, K. Leo, C. Gather, and R. Scholz, “Quantitative allocation of Bragg scattering effects in highly efficient OLEDs fabricated on periodically corrugated substrates,” Opt. Express 21, 16319–16330 (2013). [CrossRef] [PubMed]
