Micro-patterning of magnetron sputtered titanium 2 dioxide coatings and their efficiency for 3 photocatalytic applications 4

: Titanium dioxide thin films were deposited onto sola-lime glass substrates by reactive 10 magnetron sputtering. Fine stainless steel mesh sheets with different aperture sizes were applied as 11 masks over glass substrates to allow the deposition of the coatings with micro-patterned structures 12 and, therefore, enhanced surface area. Non-patterned titania films were deposited for comparison 13 purposes. The titanium dioxide films were post-deposition annealed at 873K for crystallinity 14 development and then extensively analysed by a number of analytical techniques, including 15 SEM/EDX, optical and stylus profilometry, XRD, XPS and UV-vis spectroscopy. Photocatalytic 16 activity of non-patterned and micro-patterned titania films was assessed under UV light irradiation 17 by three different methods; namely methylene blue, stearic and oleic acid degradation. The results 18 revealed that the micro-patterned coatings significantly outperformed non-patterned titania in all 19 types of photocatalytic test, due to their higher values of the surface area. Increasing the aperture of 20 the stainless steel mesh resulted in lower photocatalytic activity and lower surface area values, 21 compared to the coatings deposited through smaller aperture mesh.


26
Over the past few decades photocatalytic processes have gained recognition as simple, yet 27 sustainable methods of air / water / surfaces depollution and disinfection [1][2][3][4]. Despite the fact that 28 the overall focus of photocatalytic research seems to be shifting towards the discovery of novel 29 photocatalytic materials [5], conventional titanium dioxide (or titania)-based photocatalytic materials 30 still remain by far the most studied and practically used photocatalysts, owing to the low cost of the 31 material, high chemical and biological stability and low toxicity [6]. Titanium dioxide -based 32 photocatalytic surfaces find practical applications in such fields as self-cleaning surfaces, building 33 materials, antimicrobial materials and non-fogging surfaces [7]. It is clear that, for an efficient 34 photocatalytic process, the area of contact between the catalyst and the pollutant should be rather 56 positively charged ions (usually argon) generated in a glow discharge plasma, followed by the 57 condensation of the target atoms on the substrate to form a thin film. Reactive gases, such as oxygen, 58 can be introduced to the process to react with the sputtered metal atoms, resulting in oxide film 59 formation.

60
Sputtered coatings are usually conformal to their substrate, which means that their surface area 61 is very similar to that of the uncoated substrate. As a result, thin solid film photocatalysts deposited 62 onto plane surfaces cannot provide surface areas comparable to nanoparticulated materials.

63
Therefore, a plethora of techniques aimed at increasing the available surface area are being developed 64 and tested. Frequently reported methods include surface structuring techniques, such as the 65 formation of nanocolumns and nanorods [25,26], surface etching [27], oblique angle deposition 66 [28,29], use of high surface area substrate materials [30], etc. While each of these techniques has, to a 67 certain extent, proven to successfully increase surface area of the photocatalysts, their practical 68 application is very limited. Most of these methods are not necessarily suitable for up-scaling, but 69 rather limited to laboratory scale deposition. Here we present and assess the efficiency of increasing 70 the surface area of photocatalytic titania coatings using patterning via masked deposition. The idea 71 of masked deposition is not new on its own right; it has been mentioned in several patents -e.g., in patterned titanium dioxide surfaces, they typically use polymer / colloidal masks that have to be 76 dissolved / removed post-deposition [33][34][35]. Instead, we have attempted to use fine stainless steel 77 mesh as a mask to obtain micro-patterned titanium dioxide surfaces with higher surface areas in a 78 one-step process by reactive magnetron sputter deposition onto glass substrates. Non-patterned 79 titania coatings (produced without a mesh) were deposited for comparison purposes. The coatings

85
Coating deposition was performed in a Teer UDP350 sputtering rig, the schematic of the rig can 86 be found elsewhere [36]. In brief, the deposition was performed from one 300 mm × 100 mm type II 87 unbalanced planar magnetron, installed through the chamber wall. A directly cooled titanium (99.5% 88 purity) target was installed on the magnetron. A base pressure of 2 × 10 −3 Pa or below was achieved 89 through a combination of rotary (Edwards 40) and turbomolecular (Leybold i450) pumps. The 90 magnetron was driven in pulsed DC mode, using an Advanced Energy Pinnacle plus power supply 91 at 1 kW time-averaged power, 100 kHz pulse frequency and a duty cycle of 50% for all deposition structure development and then allowed to cool gradually in air for 10h to avoid the formation of 110 thermal stresses in the coatings (experimentally pre-defined cooling regime).      Energy-dispersive X-ray spectroscopy (EDX) was used for quantitative characterisation of film 193 composition; with the composition of each coating analysed at four points to assess uniformity -the 194 variation of the results was no greater than 2% for the same sample. No significant difference in 195 composition of the films was observed with the EDX (data are given in Table 2 Figure 8. Additionally, reaction rate constants were calculated for quantitative 297 representation of the degradation efficiency; the values are given in Table 3. Data presented in Figure   298 8 and Table 3 clearly reveal that micro-patterned titania films, and sample TiO2-M26 in particular,

299
were considerably more efficient, compared to non-patterned TiO2. Thus, for non-patterned titania, 300 full disappearance of the stearic acid peaks was observed only after 48h of UV irradiation, while for 301 micro-patterned films this time varied from 16h (for sample TiO2-M26) to 40h (for sample TiO2-302 M149). It should be noted here that reaction rate constants were calculated in each case based on data 303 points before full disappearance of the stearic acid IR peaks (e.g. for sample TiO2-M26 on 0h, 8h and 304 16h data points).

327
there are also some factors limiting its application in photocatalysis. In particular, it is frequently 328 reported that titanium dioxide coatings deposited under conventional sputtering conditions are 329 smooth and densely packed [44,45], hence surface area values are rather low.

330
It is a well-known fact that in photocatalysis higher surface area contributes to higher overall and oleic acid degradation), the micro-patterned surfaces were clearly more efficient, compared to 363 the non-patterned titania coating. While for the methylene blue degradation test, photocatalytic 364 activity followed the same trend, the improvement in activity for the patterned surfaces was not quite 365 as striking. We suggest that the observed phenomenon can be explained by the fact that for both 366 stearic and oleic degradation tests the pollutant is in direct contact with the photocatalyst surface, 367 while in the MB degradation test a transfer step is required. Therefore, the increase of surface area 368 resulted in higher efficiency most notably for the testing methods where no transfer step is required.

369
Summarising the above observations, since the best improvement of the photocatalytic activity 370 in this case was achieved for the tests where the model pollutant was in direct contact with the 371 photocatalyst surface, rather than for the liquid phase one, we suggest that the proposed method may 372 find better practical application in self-cleaning surfaces, rather than e.g. water treatment materials.

373
It should be noted that the present work only presents early results, and precise optimisation of the 374 deposition parameters, including the optimum mesh aperture, optimum thickness of the coating, etc.,

375
is the subject of a follow-up stage of work currently in progress.

377
In summary, we proposed a simple, yet efficient, method of photocatalytic thin film surface area the manuscript were all carried out by Marina Ratova. David Sawtell contributed the surface area analysis. Peter should be labeled starting with 'A', e.g., Figure A1, Figure A2, etc.