Inkjet printed interdigitated cells for photoelectrochemical oxidation of aqueous pollutants
Abstract — Planar, interdigitated photoelectrochemical cells were made by ink jet printing. The electrode fingers had widths from 200 to 1500 µm and were revealed by printing a positive protective polymer mask on FTO glass slides and subsequent etching. One finger family was overprinted by an inkjettable sol-gel composition based on titanium propoxide which was then converted into TiO2 by annealing in air. The device was finalized by printing a masking frame defining its active area. The influence of electrode geometry and titania thickness on the electrochemical properties of resulting cells is discussed in detail. Due to the interdigitated layout, photoelectrochemical response was not suffering from iR drop down to low electrolyte ionic strengths. The photoelectro-catalytic degradation of an aqueous solution of terephthalic acid by UVA illumination and electric bias of 1 V was demonstrated by monitoring the fluorescence of the OH-substituted oxidation intermediate.
Keywords: advanced oxidation process, material printing, inkjet, electrochemistry, water treatment, photocatalysis
Faculty of Chemistry,
Brno University of Technology,
Purkynova 118, 612 00 Brno,
European Institute of Technology,
Brno University of Technology,
Technicka 3058/10, 616 00 Brno,
Groupe d’Étude de la Matière Condensée (GEMaC),
CNRS/Université de Versailles, 45,
avenue des États-Unis, 78035
Versailles CEDEX, France
Photocatalytic systems employing slurried powders of TiO2 are able to deliver excellent performance due to their very high catalyst surface area . However, catalyst immobilization inevitably results in the decrease of active catalyst surface area which is accopanied by a loss of catalytic performance due to limited mass transfer . Yet, immobilized TiO2 is the preferred form of photocatalyst for industrial applications, because with an immobilized catalyst there is no need for separation of the catalyst from the liquid phase and continuos reactors are thus easier to design.
The enhancement of photocatalytic activity of an immobilized semiconductor photocatalyst by the application of external electrical bias has been convincingly demonstrated. The strategy is based on promoting photogenerated electron– hole pair separation by applying an electrical potential which results in an increase of the quantum yield of the pollutant degradation. This is only possible when the photocatalyst is deposited on an electrically conducting substrate[4-6]. However, when using electrolytes of low conductivity, iR drop is one of the factors limiting high current throughput at moderate bias. Supporting electrolyte may be added to increase the electrical conductivity in traditional electrochemical cells, but if the treatment of inevitably low ionic strength media (drinking water) is envisaged, other means for minimizing the iR drop must be secured.
The iR drop may be reduced by the use of a parallel plate reactor with two opposite electrodes and a small space between them where the electrolyte is passed through. However, this confugiration is prone to the pressure build-up in a module consisting of many such cells. This drawback can be avoided by placing both electrodes onto a single substrate and shaping them into a close mutual proximity with a very long boundary, i.e. creating interdigitated electrodes (IDE). The working electrode consists of an electrical conductor covered by semiconducting titanium dioxide. The counter electrode material is not critical as long as sufficient electrical conductivity and corrosion resistance is provided and interdigital geometry is respected. Such a design ensures two key functions: (1) it suppresses the main obstacle to efficient use of absorbed photons, i.e. the recombination of photogenerated charge carriers, by applying external electrical bias to the semiconducting photocatalyst and (2) it avoids the reduction of the generated photocurrent due to iR drop, even in electrolytes of low ionic strength. These features qualify the device as an interesting candidate for electrophotocatalytic purification of drinking water. Decomposition of model pollutants has been observed on centimeter-scale prototype devices fabricated by conventional lithographic techniques using optical copying through contact masks for resist patterning. Although the lithographic approach allows for very fine patterns to be fabricated, it is totally unsuitable for the fabrication of large foot-print cell modules.
Recently, major developments in the technology of electronic component fabrication have taken place. The dominant position of subtractive fabrication processes based on sequential etching through temporary resists masks has been challenged by a new additive approach. The so called material printing techniques  seem to be a promising microfabrication method well applicable for the production of planar structures with a low level of integration. The technique is based on sequential laying of patterned functional layers by means of modified conventional printing techniques. Strictly speaking, material printing has been around for several decades, but limited to screen printing which has been widely applied in the electronic industry. During the past decade, the concept of material printing has been significantly broadened and the portfolio of applicable techniques has included off-set, gravure, flexo, and inkjet printing too.
While generally all traditional printing techniques can be adopted for printing functional layers and patterns, inkjet printing occupies a prominent position. Despite its quite narrow viscosity and particle size limits, it seems to be the most suitable technique for lab scale prototype development as no hardware printing form is necessary, i.e., patterns designed on a printer driving computer can be printed directly without the need for physical printing form manufacturing. Moreover, up-scaling is very smooth and easy, because industrial inkjet printers with several meters working width are readily available[11-13], so transfer from prototype level to a small series level is reduced to switching to a bigger printer.
In this paper, we report on further improvement in the design and fabrication technology of planar interdigitated photoelectrochemical cells fabricated by inkjet printing. The concept of fully printed electrochemical cell introduced recently  is further expanded and more design variables are discussed in this paper. The adoption of inkjet printing provided a great freedom in the design of the cells and thus samples of various geometries and active layer thickness could be easily fabricated and their properties investigated.
A. Cell Fabrication Procedure
FTO glass sheets (Sigma-Aldrich, 11 Ω/□) were cut down to 26 x 76 mm slides, cleaned by sonication in Neodisher LM cleaning agent, rinsed in ethanol and fired at 450 °C, rinsed in
1 vol% ethanolic solution of an aliphatic hydrocarbon based hydrophobization agent (Toko Waterstop, Tokowax, Switzerland), rinsed in 1 vol% aqueous solution of sodium dodecylbenzene sulfonate (Enaspol Inc., Czech republic) and dried with a stream of nitrogen.
Printing of all functional and auxiliary layers was performed with an experimental inkjet printer Fujifilm Dimatix 2831. The printer has been successfully employed for the deposition of a wide variety of functional and auxiliary layers[15-22] and during the past years has de facto become the industrial standard tool for ink development and testing.
Generally, the following procedure was repeated for each printing step: the prepared printing formulations were sonicated for 5 min and then loaded into syringes. 0.45 µm membrane filters (Pall Corporation, USA) and blunt needles were attached to the syringe luer ports. The printing formulations were filtered and filled into the Dimatix ink tanks. Dimatix 10 pL printing heads were attached to the tanks and sequentially mounted into the Dimatix printer. The interdigitated electrode base pattern as well as the patterns for other layers were drawn as vector graphics and exported to 1- bit BMP files to be used for driving the printer. The electrode fingers had widths ranging from 200 to 2000 µm.
First, the FTO base pattern serving as the current collecting interdigitated electrode was revealed by printing a positive protective polymer mask on the FTO glass slides and subsequent etching. The polymer mask was printed by a commercial UV-curable ink (Svang Cyan, Grapo Ltd., Czech Republic), cured off-line under a mercury vapor medium pressure lamp (25 J/cm2), and baked on a hot plate at 250 °C for 10 minutes. Next, the FTO slide was etched in a mixture of zinc powder and 15 % aqueous HCl in order to remove the naked FTO while preserving those areas covered by the protective printed mask. After the etching operation was finished, the mask was fired at 450 °C for 30 min.
The second layer, i.e. the titania working electrode, was printed using our previously developed inkjetable sol-gel composition. The thickness of the titania layer after calcination was 50 nm per single layer. It was observed that it is possible to print 2 layers of sol in wet-to-wet manner followed by single calcination (heating rate 3 °C per min, 30 min at 450 °C) yielding 100 nm titania layers without any cracking in one deposition cycle. Attempts to print more sol layers followed by single calcination resulted in cracked and/or peeling layers. Therefore, thicker layers need to be prepared by repeating the complete cycle of printing double sol layers and calcination.
The device was finalized by printing a masking frame around its edges defining the active area. The same UV curable ink and curing procedure was employed for this task. Proper aligning of the titania active layer and the insulator mask was achieved by means of the Dimatix fiducial camera, which enables sample observation and print origin alignment with 5 µm accuracy. In this manner, devices with electrode finger widths ranging from 200 to 1000 µm were conveniently fabricated. Apart from regularly spaced devices, several other lay-outs were also fabricated and their properties investigated, including reduced gap devices or devices with different finger widths of working and counter electrodes. Moreover, the thickness of the titania working electrode layer was varied in the 100-500 nm range.
B. Investigation of IDE Device Properties
Printed layer quality was monitored by a Nikon Eclipse E200 optical microscope equipped with a polarized light unit and a Nikon D5000 digital camera and a Nikon Camera Control Pro 2 software. SEM imaging and elemental analysis was performed on a ZEISS EVO LS 10 scanning electronic microscope. The same machine was used for layer thickness estimation by observing sample cross sections. Phase composition of calcined titania layers was confirmed by a Panalytical Empyrean XRD system and its diffuse reflectance UV-Vis spectra were recorded by an Ocean Optics Redtide spectrophotometer with a reflectance fiber probe. Layer thickness was investigated by a Dektak XT stylus profilometer.
Photoelectrochemical characterization was performed using a two-electrode setup with the titania overprinted FTO finger family as the working electrode and the opposite naked FTO finger family as the counter electrode. This setup was fitted into a custom-build 15 x 40 x 70 mm quartz cuvette. The cuvette was filled with 0.1 M sodium sulfate solution (15 mS cm-1) and fitted onto an optical bench equipped with a Sylvania Lynx-L 18 W fluorescent UV-A lamp. The lamp emission was monitored by a Gigahertz Optic X97 Irradiance Meter with a UV-3701 probe and the irradiance was set to 5 mW cm-2 (integral irragiance in the 320-400 nm range) by adjusting the lamp to cuvette distance. A magnetic stirrer was placed beneath the cuvette and a magnetic flea inside the cuvette provided efficient electrolyte mixing. Electrochemical measurements were performed with a computer controlled electrometer in combination with a National Instruments Labview platform supplying a linear voltage sweep of 5 mV s-1 from -0.5 to 2 V. For chopped response curves, the lamp was manually obscured and revealed at 5 second intervals.
Electrophotocatalytic experiments were conducted with the same cell and light source as the electrochemical response curves were measured with. Recently, terephtalic acid was suggested as a model compound for monitoring the oxidative activity of valence band holes generated in the immobilized photocatalyst. Upon oxidation, presumably resulting from the attack of a hydroxyl radical, terephtalic acid is oxidized into hydroxyterephtalic acid, which gives a strong fluorescence signal at 425 nm. This approach proved to be very convenient for our experimental setup, because a single radiation source could be used for both the activation of the photocatalyst as well as the excitation of the fluorescent probe generated during the course of the reaction. The emitted fluorescence was collected by a quartz collimating lens mounted in the lateral wall of the cell holder and projected into an optical fiber attached to an Ocean Optics Redtide spectrometer. The spectrometer driving software allowed for a convenient automated recording of the fluorescence intensity. Calibration was performed using hydroxyterepthalic acid standard (Sigma Aldrich).
III. RESULTS AND DISCUSSION
A. Physical Properties of Devices
For routine on-the-fly checking during sample fabrication, optical microscopy is the preferred tool. However, as all the layers are transparent, images tend to be faint and have low contrast. For post-fabrication checking, profilometry and profilometric scanning provides valuable information about the thickness, roughness and homogeneity of deposited layers. Figure 3. depicts the record of a stylus profilometric mapping across 2 fingers of a 500-500-500-500 m device fabricated with 3 cycles of titania deposition. The thickness of the FTO finger (appr. 550 nm) and of the FTO + titania layer (appr. 300 nm) can be exactly determined and the overlap of the wider titania strip is also clearly visible. Figures 5 and 6 illustrate the optical properties and phase composition of the reverse micelles originated titania layers. Strong interference coloring observed visually expresses itself in the typical periodic peaks in the diffuse reflectance spectra. The phase composition was investigated by XRD and the presence of pure anatase phase and cassiterite (SnO2) phase was confirmed. Further details about the properties of this particular type of titania coating can be found in our previous communication .
B. Electrical propeties
The photoelectrochemical performance of the cell, i.e. the photocurrent generated as a function of applied voltage, is the key property for the evaluation of the IDE device quality and suitability for water treatment. The best devices were selected for further water treatment experiments. However, the actual photoelectrochemical performance of the cell is determined by several variables which are discussed below.
The irradiation intensity, wavelength and incidence (from the electrolyte side or from the substrate side) determine the amount of photons reaching the semiconductor and being absorbed by it and the corresponding rate of electron-hole pair generation. This parameter was not altered in this study and the irradiation was set to 5 mW/cm2 UVA broadband irradiation incident on the photocatalyst (front) side.
The generated photocurrent depends on the electrical resistance of the circuit, which is determined by the electrode design and electrolyte choice. Since interdigitated design has been employed in this study, the concept of the (conductivity) cell constants, κ, of the interdigitated finger device  may be applied to evaluate the influence of electrode design on the observed photocurrents. It was observed that κ values were close to the ones calculated following Olthuis at al. . A structure of 200 µm wide fingers and spaces was already fine enough to obtain a κ value below 0.01 cm-1, other cells reported in this study, i.e. cells with 500 and 1000 µm wide fingers, featured cell constants of 0.0136 and 0.0271 cm-1, respectively. Thus, in the case of a very fine fingered cell, the influence of decreasing specific conductivity of a diluted electrolyte can be compensated by finer fingered cell with a lower cell constant.
A typical photoelectrochemical response of one of the finger devices under continuous UV irradiation and in the dark is depicted at Figure 6. Three complete cyclic scans were recorded in this measurement in order to confirm the stability and repeatability of the setup. The “UV curve” describes the profile of the total photocurrent delivered by the device, i.e. the number of free charge carriers generated in the semiconductor and drawn into the external circuit by the applied voltage. The photocurrent steeply increases in the 0–0.5 V range, and then reaches a plateau where the current density is essentially independent on the voltage.
The influence of titania semiconductor thickness on the photocurrent density is depicted in Figure 7. under rectangular chopping, allowing the simultaneous recording of the i-V response in the dark and under illumination. The plateau region of all the curves is very well developed and clearly indicates the linear additive trend of photocurrent with the number of deposition. The same general trend was observed for all other sample series, i.e. the photocurrent was always linearly additive with respect to the number of deposition cycles.
The current magnitude was obviously depending on the area of the working electrode related to the total device area (the coverage). Figure 8. illustrates this by a comparative plot of response curves from samples C8, C9 and C10 having working electrode coverage of 25, 33 and 50 %, respectively. Although the general trend of increasing current with increasing coverage is respected, the absolute values do not correlate well, i.e. the photocurrent delivered by C9 (50 % working electrode coverage) is not twice as high as the photocurrent of C8 (25 % coverage). This effect can be attributed to the isotropic nature of the etching bath which creates significant undercuts removing FTO even under the printed mask. Therefore the real working electrode coverage of finer-fingered devices tends to be more reduced than the area of wider-fingered devices. The effect is even more obvious upon the comparison of two sample series with the same surface coverage by titania but different finger density, e.g. the C7 and C8 samples (Figure 9. ). The photocurrents generated by the finer-fingered series were consistently smaller than the wider-fingered one, despite the fact that the active area should be equal. In order to overcome this problem, mask files used for the fabrication of further sample series were modified to compensate for this phenomenon.
The positive effect of interdigitated electrode design on the cell performance can be illustrated by comparing the response curves in various electrolytes (Figure 10. ). Generally, in a conventional electrochemical cell, with decreasing electrolyte conductivity the photocurrent should decrease as well. However, with the interdigitated design, the ohmic drop is (partially) compensated and therefore the observed current is well conserved – while we got the photocurrent of 80 µA at 1V bias in the sodium sulphate solution of specific conductivity of 4450 µS/cm, still it is observed 70 µA in a more diluted solution of conductivity 266 µS/cm and 65 µA in even more diluted solution of conductivity 30 µS/cm.
Plots in Figure 11 further illustrate the positive effect of finer fingers by comparing the behaviour of two devices of the same working electrode coverage but different finger density. While there is essentially no difference in the shape of the C7 and C8 device photocurrent response curves in the sodium sulphate solution (15 mS/cm), the shape of the curves in demineralized water is different. The C8 device is able to deliver more current at a particular potential because the finer- fingered device leads to less iR drop. In other words, the finer fingers in the case of the 500 µm device result in lower cell constants, which are beneficial for suppression of iR drop in electrolytes of low conductivities.
Finally, an example of photodegradation of a solution of terephthalic acid under UVA irradiation and assisted by electrical bias of 1 V, using the 1000-1000-1000-1000 device with 200 nm thick TiO2 fingers is shown at Figure 12. During this experiment the photogenerated holes act as strong oxidizing species presumably producing OH radicals. Although the exact mechanism is still being debated, the net oxidative effect is evident and can be easily monitored by various probes yielding fluorescent oxidized products[24, 28, 29]. During the course of the experiment, the fluorescence due to an intermediary product (OH-substituted) increased and would eventually decrease down to zero as the fluorescent intermediate is further oxidised. The fluorescence at 425 nm plotted as a function of time reflects the initial reaction rate. While the blank checks show essentially no reaction, we can observe a doubling of the reaction rate (with respect to the unbiased case) when an external bias of 1 V was applied. From the initial slopes of the traces, the production rates, v, can be calculated taking into account the total volume of the solution. These production rates, considering Farday’s law of electrolysis, are related to the electrical charge passed (photocurrent, iphoto). The Faradaic efficiency, f, of the process is calculated using f = vF / iphoto. The value of 0.009 obtained for the experiment shown in Fig. 12 is satisfactory given the fact that a very low concentration of electroactive species (1.10-5 M) was used. Polychromatic light centered at 365 nm was used for this experiment. The IPCE (incident photon-to- current efficiency) was 0.13 at 365 nm for 200 nm thick electrodes. Values of f were found close to values obtained for the degradation of phthalic acid in a parallel plate reactor . While the datasets of the irradiated samples were measured continuously in 10 s intervals, the dark ones needed to be measured discontinuously only a few times during the reaction duration because a single radiation source was used for both catalyst activation and fluorescent probe excitation.
Inkjet material printing was successfully employed for the fabrication of planar interdigitated photoelectrochemical cells. A mixed subtractive and additive approach was adopted and all the functional and auxiliary layers were printed by an experimental inkjet printer Fujifilm Dimatix 2831. Inkjet printing proved to be an elegant method for sol delivery to the substrate. It provides a complete control over the deposition process parameters together with an excellent efficiency of precursor use. While interdigitated electrode systems have been routinely fabricated by screen printing for the past 50 years, our fully digital printing workflow eliminated the need for physical printing forms, saved valuable ink and allowed unlimited freedom in the design of the electrode patterns.
The process of FTO patterning by etching with the help of a protecting printed mask is very convenient and can be easily adopted for pattering of other materials as well. Although the resolution of printed masks can not compete with the well established resist technologies, it may be well suited for the fabrication of large footprint devices requiring a low degree of integration (solar cells, electroluminiscent modules etc.). The absence of a resist development step makes the fabrication process much more environment friendly. Moreover, it does not require any special material or equipment as it can be performed with ordinary flatbed or roll printers and UV- curable inks.
Devices with equal finger width (w) and space width (s) ranging from 200 to 1000 µm were conveniently fabricated and so were several other lay-outs including ones with w > s or devices with different finger widths (w1) of working and counter electrodes (w2). The benefits of finer fingers are clearly demonstrated, resulting in lower cell constants which are beneficial for the suppression of iR drop in electrolytes of low conductivities. With the present technology, 200 μm fingers seem to be fine enough for work with electrolytes with ionic strength typical for drinking water.
However, with the presented technology, 200 μm fingers seem to be the finest routinely achievable. Fabrication of finer fingered devices would require a better control over the etching process and printing in a cleanroom environment in order to avoid dust accumulation on wet printed patterns.
Comparative photocatalytic and electrophotocatalytic experiments with terephtalic acid as model contaminant proved the beneficial role of external electrical bias in suppressing photogenerated electron-hole recombination in the semiconducting photocatalyst. In this way, more efficient charge separation in the electric field of the IDE device has been demonstrated by the acceleration of terephtalic acid oxidation, which was conveniently monitored by the fluorescent signal of its dominant oxidation product, hydroxyterephtalic acid.
The authors thank the Ministry of Education, Youth and Sports of the Czech Republic for support through projects CZ.1.07/2.3.00/30.0005 and LD14131.
 R. van Grieken, J. Marugan, C. Sordo, C. Pablos, Comparison of the photocatalytic disinfection of E. coli suspensions in slurry, wall and fixed-bed reactors, Catalysis Today, 144 (2009) 48-54.  M. Tasbihi, C.R. Ngah, N. Aziz, A. Mansor, A.Z. Abdullah, L.K. Teong, A.R. Mohamed, Lifetime and regeneration studies of various supported TiO2 photocatalysts for the degradation of phenol under UV-C light in a batch reactor, Industrial & Engineering Chemistry Research, 46 (2007) 9006-9014.  M. Neumann-Spallart, Aspects of photocatalysis on semiconductors: Photoelectrocatalysis, Chimia, 61 (2007) 806-809.  G. Waldner, A. Brüger, N.S. Gaikwad, M. Neumann-Spallart, WO3 thin films for photoelectrochemical purification of water, Chemosphere, 67 (2007) 779-784.  P. Fernandez-Ibanez, S. Malato, O. Enea, Photoelectrochemical reactors for the solar decontamination of water, Catalysis Today, 54 (1999) 329-339.  I.M. Butterfield, P.A. Christensen, A. Hamnett, K.E. Shaw, G.M.Walker, S.A. Walker, C.R. Howarth, Applied studies on immobilized titanium dioxide films as catalysts for the photoelectrochemical detoxification of water, Journal of Applied Electrochemistry, 27 (1997) 385-395.  P.S. Shinde, P.S. Patil, P.N. Bhosale, A. Brüger, G. Nauer, M.Neumann-Spallart, C.H. Bhosale, UVA and solar light assisted photoelectrocatalytic degradation of AO7 dye in water using spray deposited TiO2 thin films, Applied Catalysis B-Environmental, 89 (2009) 288-294.  M. Neumann-Spallart, Photoelectrochemistry on a planar, interdigitated electrochemical cell, Electrochimica Acta, 56 (2011) 8752-8757.  F.C. Krebs, Fabrication and processing of polymer solar cells: A review of printing and coating techniques, Solar Energy Materials and Solar Cells, 93 (2009) 394-412.  G.D. Martin, S.D. Hoath, I.M. Hutchings, Iop, Inkjet printing - the physics of manipulating liquid jets and drops, Conference on Engineering in Physics - Synergy for SuccessLondon, ENGLAND, 2006.  P. Calvert, Inkjet Printing for Materials and Devices, Chemistry of Materials, 13 (2001) 3299-3305.  H. Sirringhaus, T. Shimoda, Inkjet Printing of Functional Materials, MRS Bulletin, 28 (2003) 802-806.  E. Tekin, P.J. Smith, U.S. Schubert, Inkjet printing as a deposition and patterning tool for polymers and inorganic particles, Soft Matter, 4 (2012) 703-713.  P. Dzik, M. Veselý, M. Králová, M. Neumann-Spallart, Ink-jet printed planar electrochemical cells, Applied Catalysis B: Environmental.  W. Tang, L.R. Feng, J.Q. Zhao, Q.Y. Cui, S.J. Chen, X.J. Guo, Inkjet printed fine silver electrodes for all-solution-processed low-voltage organic thin film transistors, Journal of Materials Chemistry C, 2 (2014) 1995-2000.  M.W. Wang, D.C. Pang, Y.E. Tseng, C.C. Tseng, The study of light guide plate fabricated by inkjet printing technique, Journal of the Taiwan Institute of Chemical Engineers, 45 (2014) 1049-1055.  G. Tarapata, M. Marzecki, Methodology and technological aspects of the flexible substrate preparation for Ink-jet Printing technology, in: R.S. Romaniuk (Ed.) Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 20132013.  J.T. Kwon, S.H. Eom, B.S. Moon, J.K. Shin, K.S. Kim, S.H. Lee, Y.S.Lee, Studies on Printing Inks Containing Poly 2-methoxy-5-(2- ethylhexyl-oxyl)-1,4-phenylenevinylene as an Emissive Material for the Fabrication of Polymer Light-Emitting Diodes by Inkjet Printing, Bulletin of the Korean Chemical Society, 33 (2012) 464-468.  P. Dzik, M. Morozova, P. Kluson, M. Vesely, Photocatalytic and Self- cleaning Properties of Titania Coatings Prepared by Inkjet Direct Patterning of a Reverse Micelles Sol-gel Composition, Journal of Advanced Oxidation Technologies, 15 (2012) 89-97.  J.D. Swartz, L.F. Deravi, D.W. Wright, Bottom-Up Synthesis of Biologically Active Multilayer Films Using Inkjet-Printed Templates, Advanced Functional Materials, 20 (2010) 1488-1492.  E. Sowade, J. Hammerschmidt, T. Blaudeck, R.R. Baumann, D. Polster, T. Baumgartel, H. Graaf, C. von Borczyskowski, R. Wagner, F. Cichos, Sist, Inkjet Printing of Polymer Microspheres, 2009.  L.F. Deravi, J.D. Swartz, D.W. Wright, Piezoelectric Inkjet Printing of Biomimetic Surfaces for Enzyme Encapsulation, 2008.  M. Cerna, M. Vesely, P. Dzik, Physical and chemical properties of titanium dioxide printed layers, Catalysis Today, 161 (2011) 97-104.  U. Cernigoj, M. Kete, U.L. Stangar, Development of a fluorescence- based method for evaluation of self-cleaning properties of photocatalytic layers, Catalysis Today, 151 (2010) 46-52.  W. Olthuis, W. Streekstra, P. Bergveld, Theoretical and experimental- determination of cell constants of planar-interdigitated electrolyte conductivity sensors, Sensors and Actuators B-Chemical, 24 (1995) 252-256.  J.M. Herrmann, Heterogeneous photocatalysis: fundamentals and applications to the removal of various types of aqueous pollutants, Catalysis Today, 53 (1999) 115-129.  D. Lawless, N. Serpone, D. Meisel, Role of oh. radicals and trapped holes in photocatalysis - a pulse-radiolysis study, Journal of Physical Chemistry, 95 (1991) 5166-5170.  K. Ishibashi, A. Fujishima, T. Watanabe, K. Hashimoto, Detection of active oxidative species in TiO2 photocatalysis using the fluorescence technique, Electrochemistry Communications, 2 (2000) 207-210.  M. Wojtoniszak, B. Zielinska, R. Kalenczuk, E. Mijowska, Photocatalytic performance of titania nanospheres deposited on graphene in coumarin oxidation reaction, Materials Science-Poland, 30 (2012) 32-38.  M. Neumann-Spallart, S.S. Shinde, M. Mahadik, C.H. Bhosale, Photoelectrochemical degradation of selected aromatic molecules, Electrochimica Acta, 111 (2013) 830-836.