# Fileset

[MDR用ppAAonMCS&LASAW_MainText&figures_23Apr2018.pdf](https://mdr.nims.go.jp/filesets/9f778381-7a3e-4ef8-9a4d-953f31f07715/download)

## Creator

Masaya Toda, Koji Miyake, Li-Qiang Chu, Marjan Zakerin, Renate Förch, Rüdiger Berger, [Akiko N. Itakura](https://orcid.org/0000-0001-5783-141X)

## Rights

[In Copyright](http://rightsstatements.org/vocab/InC/1.0/)

## Other metadata

[Youngs modulus of plasma-polymerized allylamine films using micromechanical cantilever sensor and laser-based surface acoustic wave techniques](https://mdr.nims.go.jp/datasets/e58720f7-4cff-4386-b87f-5dae14395bf9)

## Fulltext

DOI: 10    - 1 - Plasma Processes and Polymers 1 Article type: Full Paper 2  3 Title: Young’s modulus of plasma-polymerized allylamine films using micromechanical 4 cantilever sensor and laser-based surface acoustic wave techniques 5  6 Author(s), Corresponding Author(s)*  7 Masaya Toda*, 8 Tohoku University, 6-6-01 Aramaki-Aza-Aoba, Aoba-ku, Sendai 980-8579, Japan 9 E-mail: mtoda@nme.mech.tohoku.ac.jp 10  11 Koji Miyake, 12 National Institute of Advanced Industrial Science and Technology, 1-2-1 Namiki, Tsukuba, 305-8564, Japan 13  14 Li-Qiang Chua, Marjan Zakerin, Renate Förchb, Rüdiger Berger* 15 Max Planck Institute for Polymer Research, Ackermannweg 10, Mainz D-55128, Germany 16  17 Akiko N. Itakura* 18 National Institute for Materials Science, 1-2-1 Sengen, Tsukuba 305-0047, Japan 19 E-mail: itakura.akiko@nims.go.jp 20  21  22 ––––––––– 23 Abstract 24  25 Mechanical properties of ultra-thin organic films are of fundamental importance considering 26 coating applications. Micromechanical cantilever sensor (MCS) and laser-based surface 27 acoustic wave (LA-SAW) techniques were both used to measure the Young's moduli of plasma 28 polymerized films at different humidities. For plasma polymerized allylamine (ppAA) films 29 deposited at 5 W and 90 W, Young’s moduli of 1400 ± 350 MPa and 110 ± 20 MPa at humidities 30 between 10 – 40 %, and 1070 ± 250 MPa and 32 ± 10 MPa at humidities between 70 – 80 % 31 were measured. LA-SAW technique revealed Young’s moduli lower than 60 % of those by 32 MCS technique. The difference suggests an enhanced swelling at the air interface or a gradient 33 of cross-linking density. (120 words) 34   35  a Currently at Tianjin University of Science & Technology,  No 29, 13th Avenue, TEDA, 300457 Tianjin, China. E-mail: chuliqiang@tust.edu.cn b Currently at Eura Consult AG, Max-Eyth-Strasse 2, 73479 Ellwangen, Germany. E-mail: renate.foerch@euraconsult.de     - 2 - 1. Introduction  36 Organic thin films are an attractive approach to immobilize molecules [1-3] and to realize 37 superhydrophobic [4, 5] and dielectric coatings. [6, 7] Organic thin films can be directly deposited 38 on almost any substrate by using a plasma-assisted polymerization process. [8] In order to 39 achieve efficient functionality of plasma polymerized organic thin films, specific organic 40 molecules can be used during the deposition process. [9-11] In addition to the chemical 41 functionality, the mechanical stability of plasma polymerized films plays a major role. For 42 example, the coatings of mechanical components need to be highly resistant to wear, [12, 13] 43 corrosion, [14, 15] and delamination. Plasma-assisted polymerization processes offer a unique 44 option in thin film preparation, since the mechanical stability can be adapted by varying the 45 deposition conditions. This in turn affects the cross-linking density and mobility of polymeric 46 chains. [16] In general, it has been shown that films deposited at a higher input power level 47 exhibit a higher cross-linking density. [17] Thereby Young’s modulus, [18] hardness of film [19] 48 and its wear resistance can be increased. [20] 49 Dynamic mechanical analysis (DMA) is a standard characterization method used to determine 50 mechanical properties of polymers. However, DMA requires thick films (> 1µm) and is thus 51 not appropriate for the study of thin (< 1µm) films. Nano-indentation technique is considered 52 to be one of the most useful methods for determining Young’s modulus (E) of thin films. [21-25] 53 However, for ultra-thin films up to a thickness of 20 nm the indentation depth is limited to about 54 10 – 20 % of the film thickness. [26, 27] In other words, indentions of < 2 nm have to be made in 55 order to avoid the influence of substrate materials. Such small indentations are possible by using 56 scanning force microscopy methods. However, scanning force microscopy methods are able to 57 sense local mechanical properties, which can be different to mechanical properties of entire 58 film. Consequently, several measurements have to be performed at different positions. 59 Mechanical properties of plasma-deposited films may vary according to their thickness. 60     - 3 - Therefore, alternative methods have to be developed and evaluated to measure ultra-thin 61 polymer films.  62 Recently, we described the possibility of calculating E of polymer films made by layer-by-layer 63 deposition of polyelectrolytes (≤ 20 nm thickness) on the micromechanical cantilever sensor 64 (MCS). [28-35] The swelling of polymer films in a solvent vapor environment transduces into a 65 measurable bending of MCS. [36, 37] In this way, 𝐸 of the thin polymer film can be calculated by 66 determining the magnitude of bending and the swelling ratio. This method should in theory, be 67 applicable for ultra-thin films made by plasma deposition.  68 Alternatively, E of ultra-thin films can be determined by laser-based surface acoustic wave 69 (LA-SAW) technique. [38-41] Here, a surface acoustic wave (SAW) is generated by laser 70 irradiation on top of the ultra-thin film. The phase velocity of SAW is dependent on E. Typically, 71 LA-SAW technique performs well for ultra-thin films with a 𝐸 > 300 GPa. [24] However, it is 72 an open question whether thin polymer films can be measured using LA-SAW technique, and 73 whether it is reliable in the case of ultra-thin plasma polymerized films. 74 In this study we used plasma polymerized allylamine (ppAA) as a model material. [42] The 75 ppAA films deposited on MCS swell upon exposure to humidity, [37] and we have previously 76 shown that the resulting changes in their mechanical properties can be measured by the bending 77 of MCS. [43] Our current work focuses on calculating E of ultra-thin plasma polymerized films 78 using (i) the bending characteristics of ppAA coated MCS, and (ii) LA-SAW propagation at 79 surface in response to humidity changes. Ellipsometry was used to determine the material 80 parameters additionally required for ppAA films to calculate E. 81  82 2. Experimental Section 83 I. Preparation of ppAA films 84 Allylamine monomer (99 %, Sigma-Aldrich) was plasma-polymerized with input power levels 85 of 5 W and 90 W under continuous wave conditions, and a process pressure of 10 Pa. This 86     - 4 - corresponds to our previous published experimental work using a 13.56 MHz Pyrex plasma 87 reactor. [37, 42] The ppAA films deposited at a higher plasma power level (P = 90 W) resulted in 88 a higher cross-linking density than those deposited at lower power level (P = 5 W). [23] Generally, 89 films deposited at higher input power showed a higher roughness and cross-linking density [17]. 90 The energetic ion bombardment and UV irradiation during the plasma deposition process 91 determines the cross-linking density of the film.  [44, 45] 92  93 II. Thickness measurement of ppAA films 94 Imaging ellipsometry (EP3, Nanofilm Technologie GmbH, Germany) was used to measure the 95 thickness changes (d/d) of ppAA films under defined humidity conditions. The samples were 96 mounted in the fluid cell (SL-cell, Nanofilm, volume 7 cm3) with windows at an angle of 60° 97 relative to the sample plane. A laser with a wavelength of 532 nm at an incident angle of 60° 98 was used. For ellipsometric measurements, ppAA film was deposited directly onto a piece of 99 silicon wafer (nSi = 3.879 – 0.0257i). The refractive index of ppAA film, nppAA, was calculated 100 from a fit to a model consisting of a single uniform layer on a substrate for each measurement. 101 For a 5 W film of nppAA = 1.564 – 0.0i, and for a 90 W film of nppAA = 1.624 – 0.0i resulted in 102 the best fit. The errors of fits in both cases were ± 0.004. A scanning force microscope was 103 used to measure the root mean square (RMS) surface roughness of ppAA films, which was 0.6 104 nm for a corresponding area of 1 µm2. 105  106 III. Humidity (H) control 107 A humidity controller was used to control the partial vapor pressure in a fluid cell which enabled 108 relative H conditions from 0 to 100 %. [46] In the dry state, the fluid cell was flushed with N2 109 gas at a defined H by using a specially-constructed gas mass flow setup (flow controller: Model 110 80s, McMillan Company, Texas, USA). Different relative concentrations of N2 and H2O were 111 obtained by mixing dry N2 gas and N2 gas saturated with water vapor at T = 20 °C. The setup 112     - 5 - was operated at a constant flow speed of 0.5 L/min. After buffering the vapor, a small amount 113 of the vapor was fed into the fluid cell by using an electric pump (NMP-30, Neuberger Inc., 114 New Jersey, USA). For the measurements, a constant flow of 100 µL/min through the fluid cell 115 was established (volume of the fluidic cell ~30 µL and volume of the connecting tubes ~20 µL). 116 Given this set-up, the fluid cell environment changes between dry and different H states in 117 desired steps of 10 %. After reaching the desired H, the system was flushed with dry N2 and 118 then the H was increased further. 119  120 IV. MCS bending measurement 121 The deflection and resonance frequency of MCS were measured by an optical beam deflection 122 technique in the above-mentioned fluidic cell (SCENTRIS, VEECO Instruments, USA, Figure 123 1). We measured the resonant frequencies of eight MCSs arranged in a linear array. We used 124 MCSs fabricated from Si with a thickness of 1.2 – 1.9 m, a length of 500 m, a width of 90 125 µm and a pitch of 250 μm (Octosensis, Micromotive GmbH, Germany). The resonant 126 frequencies of uncoated MCS were 10730 ± 30 Hz and 6900 ± 20 Hz and were used for 5 W 127 and 90 W ppAA coatings, respectively. The films of ppAA were prepared on the MCS chip and 128 on Si wafer substrates in the same process described previously [42]. Each of the eight-MCS 129 arrays had at least one MCS without a ppAA film (Figure 1). These were used as references 130 and were required to compensate thermal drift in the setup that arises during a 24 h experiment.  131  132 3. Principle of Measurements 133 I. 𝑬 calculation by MCS technique 134 Swelling of films deposited on MCS generates surface stress, which is proportional to the 135 difference in expansion coefficients between the film and the substrate (Figure 1). [47] The 136 deflection behavior of MCS can be described as being similar to the bending of a bi-metallic 137     - 6 - strip upon temperature changes. [48] The change in deflection ∆δ of a bi-material strip upon 138 humidity changes ∆𝐻 is given by [49] 139 ∆δ =𝑙22𝑅=3𝑙2𝐸Si𝐸film𝑡Si𝑡film(𝑡Si+𝑡film)(𝐸Si𝑡Si2)2+(𝐸film𝑡film2)22𝐸Si𝐸film𝑡Si𝑡film(2𝑡Si2+3𝑡Si𝑡film+2𝑡film2)(𝛼film − 𝛼Si)∆𝐻   (1)  , 140 where 𝛼film and 𝛼Si are linear expansion coefficients of film and Si at a given H, 𝐸Si and 𝐸film  141 are 𝐸 of Si and film, 𝑡Si and 𝑡film are thicknesses of Si cantilever beam and film, 𝑙 is a length of 142 cantilever beam. The geometric values of 𝑙, 𝑡Si and 𝑡film are given by the fabrication design of 143 MCS. Thus,  𝛼film and 𝐸film are the unknown parameters in equation (1). In order to check 144 which magnitude of 𝐸film values are accessible by this method, we calculated the change in 145 deflection, ∆δ, at a typical 𝛼film of 0.0001 for polymers for a ∆𝐻 of 10 % (Figure 2). For 𝑡film 146 < 𝑡Si (𝑡film /𝑡Si = 0.05), the deflection of MCS due to swelling of the film becomes almost 147 directly proportional to 𝐸film in the range of 107 – 1011 Pa.  Typically, the deflection of MCS 148 can be measured with an accuracy of < 1 nm. Thus 𝐸film  values of 10 MPa or more are 149 accessible.  150  151 II. 𝑬 calculation by LA-SAW technique: 152 Laser acoustic thin film analyzer (LA wave, IWS, Germany) was used to estimate 𝐸 of thin 153 ppAA films near the air surface. SAWs were generated by illumination of the surface with short 154 laser pulses (wavelength: 337 nm, pulse duration of 0.5 ns, and power of 0.4 mW). This method 155 is based on measuring the phase velocity of generated SAW. The SAW propagates along the 156 surface of materials and its amplitude decays exponentially. In the case of thick bulk samples 157 (homogeneous and isotropic materials without any film on top), the phase velocity 𝑐bulk of 158 SAW is given by equation (2) [50] 159 𝑐bulk =0.87+1.12𝜈21+𝜈√𝐸2𝜌(1+𝜈)                                                 () 160     - 7 - where ν is Poisson ratio of a bulk sample material. Thus the phase velocity of SAW for bulk 161 samples in the propagation direction depends on both 𝐸 and 𝜌, and is, in particular, not related 162 to the frequency of SAW. [51]  163 However, in the case of thin films on top of a bulk sample the depth of propagating 164 SAW needs to be considered. The depth of propagating wave becomes proportional to 165 wavelength [52] and decreases with diminishing frequency. Correspondingly, the phase velocity 166 of SAW depends on the film’s thickness and the frequency of SAW. [53] The phase velocity is, 167 in particular, more relevant for thin films when the SAW is generated at a higher frequency. 168 For a homogeneous and isotropic thin film and a homogeneous and isotropic substrate, the 169 phase velocity 𝑐film+substrate  depends on 𝐸 of the film and the substrate (𝐸′film and 𝐸substrate), 170 on their Poisson’s ratios (𝜈film  and 𝜈substrate), their densities (𝜌film  and 𝜌substrate), the film 171 thickness 𝑡, and the frequency of SAW , as represented by the dispersion relation, shown in the 172 following equation (3). 173 𝑐film+substrate = g(𝐸substrate, 𝐸film , 𝜈substrate, 𝜈film , 𝜌substrate , 𝜌film , 𝑡, 𝑓)   () 174 In our measurements, a pulsed nitrogen laser was used to generate a wide band surface 175 wave. The laser beam was shaped into a line on sample surface by a cylindrical lens. The 176 generated SAW impulses were detected by a transducer for at least two different distances, 𝑥1 177 and 𝑥2, from the transducer to the laser focus line. The surface wave velocity spectra 𝑐(𝑓) is 178 obtained from: 179 𝑐(𝑓) =2𝜋𝑓(𝑥2−𝑥1)[𝜙2(𝑓,𝑥2)−𝜙1(𝑓,𝑥1)]   ,        () 180 where 𝜙1(𝑓, 𝑥1) and 𝜙2(𝑓, 𝑥2) are SAW phase values for frequency of 𝑓 at positions 𝑥1 and 181 𝑥2, respectively. The 𝑐(𝑓) at a specific position can be obtained from a Fourier transformation 182 of SAW signals in the time domain registered by the transducer at both distances. Then 𝐸′film 183 is obtained by fitting equation (4) to the dispersion data. 184  185     - 8 - 4. RESULTS 186 I. Vertical expansion: Thickness and swelling measurements 187 In order to calculate Efilm by MCS method, the film at different H values of thin ppAA films 188 were required. Here, we applied ellipsometry to determine the relative changes in the film 189 thicknesses (d/d) of ppAA film at different H values (Figure 3 (a)). The ppAA films deposited 190 in dry nitrogen at a plasma power level of 5 W and 90 W, had thicknesses of 86.4 ± 0.1 nm 191 and 98.1 ± 0.4 nm, respectively. We found that the thickness of ppAA films increased linearly 192 up to H = 60 %. In addition, the increase in thickness appeared to be independent of the plasma 193 power level used until H = 60 % was reached. Thus, the H induced expansion coefficients were 194 similar to the 5 W and 90 W ppAA films in the range of H  = 0 % to H = 60 %. Above H = 195 60 %, the relative film thickness of 5 W deposited ppAA thin film increased significantly more 196 than 90 W ppAA films. This difference in swelling was consistent with the assumption that 5 197 W ppAA films exhibit a lower cross-linking density and can thus incorporate more water 198 molecules.  199 Since the ppAA film expansion was dependent on humidity, we calculated the 200 corresponding humidity induced expansion coefficients, , by H increments of 10 % (Figure 3 201 (b)). We obtained film ⊥ = 0.0005 ± 0.0002 for the H range from 5 % to 60 % for both 5 W 202 and 90 W films. At H > 60 %, film ⊥ becomes 0.0032 ± 0.0002 and 0.0012 ± 0.0004 for the 203 5 W and 90 W ppAA films, respectively. The error bars correspond to the standard deviation 204 of measurements at seven different areas of the same sample. The symbol ⊥ denotes that we 205 measured thickness changes by ellipsometry.  206 However, for MCS technique the humidity induced expansion coefficients of ppAA films 207 along MCS lengths (i.e. film //) are required. For thin films that swell homogeneously and 208 isotropically film // = film ⊥can be assumed. Nevertheless, the ppAA film is anchored to the 209 MCS surface. Thus, the presence of MCS surface hinders the lateral expansion of the ppAA 210     - 9 - film and film // ≤ film ⊥. We will discuss errors in calculating E given this connection in section 211 III-A.   212  213 II.  Lateral Expansion: MCS bending measurement 214 Deflection of MCS by swelling ppAA films 215 Identical ppAA films were prepared on MCS and exposed to different H ranging from 0 % to 216 85 % using a computer-controlled setup (Figure 4 (a) and (b)). Two subsequent H cycles are 217 plotted, showing the observed deflection for a 5 W ppAA film during both increasing and 218 decreasing H cycles (Figure 4 (a)). One typical cycle of increasing and decreasing H for a 90 219 W ppAA film is shown in Figure 4 (b). At time t = 10 minutes (Figure 4(b)) the stability of 220 measured deflection signal was confirmed by keeping H = 20 % for a few hours. At constant 221 H, fluctuations of deflections < 0.2 % were measured and were considered negligible. 222 Furthermore, the drift of the entire setup was measured by determining the response of adjacent 223 MCSs, acting as references that were not coated with a polymer film.  The deflection values of 224 such reference measurements were used to subtract background signals (Supporting 225 information). Our measurements confirm reproducible swelling of ppAA films and indicate 226 no significant or measurable polymer degradation during repeated exposure to humidity over 227 the time frame studied in this experiment (30 hours). 228 Figure 4 (c) and (d) show the bending response of 5 W ppAA film and 90 W ppAA 229 film over time for a stepwise H variation from 0 % to 10 %. After the exposure to each humid 230 environment for 10 minutes (in steps of 1 %, 2.5 %, 5 %, and 10 %) the samples were exposed 231 to dry N2 for 20 minutes before the next exposure to humidity. By switching the single four-232 way valve we were able to create a dry condition in the sample chamber (Figure 1). For each 233 dry exposure, the bending deflection returned to the initial zero point of deflection which 234 confirmed that (i) the gas inside the chamber was replaced completely, and (ii) the reversible 235 swelling of PPAA film (Figure 4 (c) and (d)).  236     - 10 - It is worth to notice that the saturated deflection magnitude at the same H-level between 237 the increasing and decreasing cycle is different. For example, at H = 10 % in the decreasing 238 cycle the saturated deflection of 594 nm was measured. This value was slightly smaller than 239 that deflection magnitude of 630 nm in the increasing cycle for a 5 W ppAA film. The hysteresis 240 could be up to 6 % in deflection magnitude and indicates that the de-swelling of films was 241 affected by remaining water molecules left over from previous humidity exposure to films.  242  243 Stress changes of swelling ppAA films 244 The change in stress within the swollen ppAA film can be calculated from the deflection 245 magnitude of MCS using Stoney’s formula. [54-56] Within the entire H range from 10 % to 80 %, 246 we measured a larger stress change for the ppAA film deposited at 90 W compared with the 247 one deposited at 5 W (Figure 5). Moreover, the experiments that were performed in the 248 increasing H process showed a larger stress change compared to the subsequent experiments 249 performed while reducing H. This behavior was found for both 5 W and 90 W ppAA films and 250 was fully reproducible in the second cycle recorded the next day. At first glance, this additional 251 water residue should lead to a higher stress magnitude while decreasing H, because more water 252 was present in ppAA films. However, the presence of residual water also influences the 𝐸film. 253 Specifically, the presence of residual water decreases 𝐸  of ppAA films. Consequently, the 254 deflection magnitude and the calculated stress change both decrease compared to the 255 experiment where H was increased stepwise. In order to address the overall changes in 𝐸film, 256 we calculated averaged stress changes (lines plotted in Figure 5). The results indicate an 257 increase of the averaged stress change of the swollen ppAA film when exposed to humidity. 258 However, the increase of averaged stress was not directly proportional to the level of H. At the 259 highest measured level of H, the stress value for 90 W ppAA film was 1.6 times larger than that 260 of 5 W ppAA film.  261  262     - 11 - III.  E calculations 263 A. MSC technique. We calculated 𝐸film using equation (1) (circle symbols in Figure 6). For 264 the calculation, the averaged deflections and calculated film ⊥ values based on ellipsometry 265 data (Figure 3 (b)) were used. Furthermore, in humid environments, the values of Si = 0 and 266 ESi =165 GPa remained constant. However, it was important to consider the relationship of 267 film //  film ⊥  which was deduced from ellipsometry measurements. Therefore, our 268 calculations of 𝐸film using film ⊥ can be seen as the lower limits of 𝐸film. In order to estimate 269 an upper limit for 𝐸film, the minimum swelling state of a ppAA film at a low H level was used. 270 In the case of the small volume expansion, the effects from substrate can be disregarded. We 271 determined that the swelling of ppAA film was isotropic and that the film expansion in both 272 parallel (//) and vertical (⊥) directions was identical, i.e. 0.0005 for both of the ppAA films 273 deposited at 5 W and 90 W (Figure 3 (b)).  274 Irrespective of the power level during plasma polymerization, 𝐸film  of both films 275 decreased with increasing H (Figure 6). The 𝐸film calculated for 5 W and 90 W ppAA films 276 were 1070 ± 250 MPa and 1400 ± 350 MPa for H = 0 % and H = 10 % , respectively. These 277 results were slightly lower than E-values obtained by nano-indentation on plasma-polymerized 278 thin films of vinyltriethoxysilane (3 – 10 GPa [18]). We attribute the lower E-values for ppAA 279 films to the different composition of films and their  partially swollen state during  280 measurements.  281 For H < 60 % almost no difference in 𝐸film values between 5 W and 90 W ppAA films 282 was found. For H > 60% the higher cross-linking density in 90 W film leads to a reduced 283 swelling and thus to an increased 𝐸film of 110 ± 20 MPa between H = 70 % and H = 80 % 284 compared to 5 W film (32 ± 1 MPa). The results indicate that the mechanical properties vary 285 less upon H-changes for a film with a higher crosslinking density, i.e. deposited at a higher 286 power. 287     - 12 -  288 B. LA-SAW. LA-SAW measurement was performed over a H range of 20 – 50 % and at a 289 room temperature of 25 ± 3 °C for ppAA films of thickness 100 ± 10 nm deposited on a Si-290 substrate. Then, SAW propagation along <011> direction of Si (100) was measured. SAWs 291 with a frequency of 30 MHz were detected at different distances, l + ∆l, where l was 25 mm 292 and ∆l was varied from 0 to 5 mm. The calculated phase velocity spectra of SAW wave–packets 293 were used to fit 𝑬′𝐟𝐢𝐥𝐦 of ppAA films (square symbols in Figure 6). 𝑬𝐬𝐮𝐛𝐬𝐭𝐫𝐚𝐭𝐞 corresponds to 294 Si substrate (density: 2.33 g/cm3) and we took C11: 165 GPa, C12: 63.5 GPa, C44: 79.6 GPa for 295 each crystal orientation, respectively. X-ray reflectivity measurements (Rigaku RINT ATX-G) 296 revealed a ppAA film density of 1.12 ± 0.06 g/cm3 (5 W) and 1.55 ± 0.04 g/cm3 (90 W). [57, 58] 297 To determine the Poisson’s ratio of films we are taking 0.4, which is a widely used value for 298 polymers [59]. The 𝑬′𝐟𝐢𝐥𝐦of the ppAA film of 5 W decreased from 0.249 GPa to 0.138 GPa for 299 a H change from 15 % to 46 %. These 𝑬′𝐟𝐢𝐥𝐦 values were 41 ± 38 % and 60 ± 30 % lower than 300 ones calculated by MCS technique for 5 W ppAA and 90 W ppAA, respectively (Figure 6). 301 However, the relative differences of 𝑬 between both films at different H levels were confirmed: 302 The 𝑬′𝐟𝐢𝐥𝐦 of the 90 W film was larger compared to the 5 W film, and the 𝑬′𝐟𝐢𝐥𝐦 in high H 303 conditions was smaller than in a low H environment.  304  305 6. Discussion 306 The 𝐸′film in LA-SAW measurement was smaller compared to the 𝐸film  measurement using 307 MCS technique. Based on surface acoustic waves, LA-SAW technique was more sensitive to 308 polymer properties at its air interface and less sensitive to polymer properties at its substrate 309 interface. The difference of calculated 𝐸′film  and 𝐸film  can be explained by two possible 310 theories: (a) ppAA films swell more at polymer/air interface compared to the interface of solid 311 substrate. The presence of substrate may even lead to some inhibition of swelling. (b) The 312 crosslinking density of ppAA films was lower at the air-interface compared to its interface with 313     - 13 - a solid substrate. In contrast, the driving force for cantilever bending occurs at the silicon 314 substrate interface. We are of the opinion that both theories are feasible for ultra-thin ppAA 315 films.  316  317 7. Conclusion and summary 318 We measured Young’s modulus of approximately 100 nm thick polymerized films that 319 were made using a plasma deposition process. We selected plasma-polymerized allylamine 320 (ppAA) as model films that swell upon exposure to humidity. A decrease of Young’s modulus 321 with increasing H was consistently observed. The MCS technique is based on the measurement 322 of the deflection of cantilever beams. The error of the measurement was dominated by the error 323 of the expansion coefficient that was inferred from ellipsometry. In contrast, the LA-SAW 324 technique revealed lower 𝐸 values. We attribute this difference to a lower cross-linking density 325 of ppAA films at the air interface, or to enhanced swelling of polymer close to the air interface. 326 One major outcome of this work is that both techniques MCS and LA-SAW are applicable 327 for measuring ultra-thin plasma polymerized films and represent an alternative to nano-328 indentation measurements. Nano-indentation measurements might not be easily feasible in 329 harsh environments such as polymer solvents. However, MCS technique is not only restricted 330 to experiments involving humidity. It can be readily used in experiments using various polymer 331 solvents and vapors. [46, 60, 61] A simple estimation revealed that MCS is able to measure E down 332 to 10 MPa. In principle, the MCS method can be used to determine 𝐸 for films with a thickness 333 of less than 10 nm, as long as the expansion coefficient can be determined.  334  335  336 Acknowledgements: 337 We would like to thank Mr. Andrey Grinevich, who was an internship student from Charles 338 University, and Dr. Megumi Fukuta from the National Institute of Advanced Industrial Science 339     - 14 - and Technology for the measurements of LA-SAWs. Dr. Kenji Sakurai from NIMS is 340 acknowledged for providing the X-ray refractivity tool and help in determining the film density. 341 We acknowledge the input of our former group member Dr. Shin-ichi Igarashi (now at Shin-342 Etsu Chemical Co., Ltd.), as we started this research with him and the basic idea came from our 343 discussions. This research was supported by funding from the Japanese Science and 344 Technology Agency, National Institute for Material Science, and the Max Planck Institute for 345 Polymer Research, Germany. We acknowledge partial financial support from the Deutsche 346 Forschungsgemeinschaft (DFG) within the Strategic Japanese German Cooperative Program 347 on Nanoelectronics (BE3286/2-1) and (BE 3286/4-1). In addition, we acknowledge the funding 348 received from  the Promotion of Joint International Research by the JSPS KAKENHI Grant 349 Numbers 15KK0225. 350  351  352 Keywords: cantilever sensing, humidity sensor, plasma polymerization, Young’s modulus of 353 ultra-thin films  354   355     - 15 - References 356 [1] Z. H. Zhang, Q. Chen, W. Knoll, R. Foerch, R. Holcomb, D. Roitman, Macromolecules 357 2003, 36, 7689. 358 [2] J. Homola, Chem. Rev. 2008, 108, 462. 359 [3] F. Rusmini, Z. Zhong, J. Feijen, Biomacromolecules 2007, 8, 1775. 360 [4] A. Lazauskas, J. Baltrusaitis, V. Grigaliūnas, D. Jucius, A. Guobienė, I. Prosyčevas, P. 361 Narmontas, Plasma Chem. Plasma Process. 2013, 34, 271. 362 [5] A. Tuteja, W. Choi, M. Ma, J. M. Mabry, S. A. Mazzella, G. C. Rutledge, G. H. 363 McKinley, R. E. Cohen, Science 2007, 318, 1618. 364 [6] Y. Xu, P. R. Berger, J. Cho, R. B. Timmons, J. Electron. Mater. 2004, 33, 1240. 365 [7] C. Reese, M. Roberts, M.-m. Ling, Z. Bao, Mater. Today 2004, 7, 20. 366 [8] E. Radeva, Sensors and Actuators B: Chemical 1997, 44, 275. 367 [9] L. Denis, D. Cossement, T. Godfroid, F. Renaux, C. Bittencourt, R. Snyders, M. Hecq, 368 Plasma Processes and Polymers 2009, 6, 199. 369 [10] Y. Wang, J. Zhang, X. Shen, Mater. Chem. Phys. 2006, 96, 498. 370 [11] R. Morent, N. De Geyter, S. Van Vlierberghe, A. Beaurain, P. Dubruel, E. Payen, Prog. 371 Org. Coat. 2011, 70, 336. 372 [12] H. Holleck,  V. Schier, Surf. Coat. Technol. 1995, 76-77, 328. 373 [13] T. Spalvins, Thin Solid Films 1978, 53, 285. 374 [14] M. Stratmann, R. Feser, A. Leng, Electrochim. Acta 1994, 39, 1207. 375 [15] W. J. van Ooij, D. Zhu, M. Stacy, A. Seth, T. Mugada, J. Gandhi, P. Puomi, Tsinghua 376 Science and Technology 2005, 10, 639. 377 [16] T. Kaule, Y. Zhang, S. Emmerling, S. Pihan, R. Foerch, J. Gutmann, H. J. Butt, R. 378 Berger, U. Duerig, A. W. Knoll, ACS Nano 2013, 7, 748. 379 [17] Y. Zhang, J. Arfsten, S. A. Pihan, T. Kaule, R. Förch, R. Berger, J. Colloid Interface Sci. 380 2010, 351, 532. 381 [18] V. Cech, N. Inagaki, J. Vanek, R. Prikryl, A. Grycova, J. Zemek, Thin Solid Films 2006, 382 502, 181. 383 [19] J. Ahmad, K. Bazaka, M. Jacob, Electronics 2014, 3, 266. 384 [20] R. Berger, Y. Cheng, R. Förch, B. Gotsmann, J. S. Gutmann, T. Pakula, U. Rietzler, W. 385 Schartl, M. Schmidt, A. Strack, J. Windeln, H. J. Butt, Langmuir 2007, 23, 3150. 386 [21] J. Domke,  M. Radmacher, Langmuir 1998, 14, 3320. 387 [22] K. Miyake, N. Satomi, S. Sasaki, Appl. Phys. Lett. 2006, 89, 031925. 388 [23] S. Igarashi, A. N. Itakura, M. Kitajima, A. N. Chifen, R. Foerch, R. Berger, Appl. Phys. 389 Lett. 2006, 88, 143119. 390 [24] T. Chudoba, M. Griepentrog, A. Dück, D. Schneider, F. Richter, J. Mater. Res. 2011, 19, 391 301. 392 [25] V. Cech, J. Lukes, E. Palesch, T. Lasota, Plasma Processes and Polymers 2015, 12, 864. 393 [26] H. J. Butt, B. Cappella, M. Kappl, Surf. Sci. Rep. 2005, 59, 1. 394 [27] D. Passeri, M. Rossi, E. Tamburri, M. L. Terranova, Analytical and Bioanalytical 395 Chemistry 2012, 405, 1463. 396 [28] M. Toda, Y. Chen, S. K. Nett, A. N. Itakura, J. Gutmann, R. Berger, Journal of Physical 397 Chemistry C 2014, 118, 8071. 398 [29] J. H. He, J. K. Luo, H. R. Le, D. F. Moore, Materials Science and Engineering: A 2006, 399 423, 134. 400 [30] G. J. McShane, M. Boutchich, A. S. Phani, D. F. Moore, T. J. Lu, J Micromech 401 Microeng 2006, 16, 1926. 402 [31] S. He, J. S. Chang, L. Li, H. Ho, Sensors and Actuators A: Physical 2009, 154, 149. 403 [32] J. H. Lim, M. M. Ratnam, I. A. Azid, D. Mutharasu, Optics and Lasers in Engineering 404 2011, 49, 1301. 405     - 16 - [33] I. Stachiv, D. Vokoun, Y.-R. Jeng, Appl. Phys. Lett. 2014, 104, 083102. 406 [34] S. Singamaneni, M. E. McConney, M. C. LeMieux, H. Jiang, J. O. Enlow, T. J. Bunning, 407 R. R. Naik, V. V. Tsukruk, Adv. Mater. 2007, 19, 4248. 408 [35] J. H. He, J. K. Luo, H. R. Le, D. F. Moore, Mat Sci Eng a-Struct 2006, 423, 134. 409 [36] C. Ishiyama,  Y. Higo, J Polym Sci Pol Phys 2002, 40, 460. 410 [37] S. Igarashi, A. N. Itakura, M. Toda, M. Kitajima, L. Chu, A. N. Chifene, R. Förch, R. 411 Berger, Sensors and Actuators B-Chemical 2006, 117, 43. 412 [38] X. Xiao, X. M. Shan, Y. Kayaba, K. Kohmura, H. Tanaka, T. Kikkawa, Microelectron. 413 Eng. 2011, 88, 666. 414 [39] H. J. Scheibe,  B. Schultrich, Thin Solid Films 1994, 246, 92. 415 [40] F. Vaz, S. Carvalho, L. Rebouta, M. Z. Silva, A. Paúl, D. Schneider, Thin Solid Films 416 2002, 408, 160. 417 [41] G. Chow, E. Uchaker, G. Cao, J. Wang, Mech. Mater. 2015, 91, 333. 418 [42] L. Q. Chu, W. Knoll, R. Förch, Langmuir 2006, 22, 5548. 419 [43] M. K. Baller, H. P. Lang, J. Fritz, C. Gerber, J. K. Gimzewski, U. Drechsler, H. 420 Rothuizen, M. Despont, P. Vettiger, F. M. Battiston, J. P. Ramseyer, P. Fornaro, E. Meyer, 421 H.-J. Güntherodt, Ultramicroscopy 2000, 82, 1. 422 [44] N. Vandencasteele,  F. Reniers, J. Electron. Spectrosc. Relat. Phenom. 2010, 178, 394. 423 [45] M. S. Silverstein, L. Sandrin, E. Sacher, Polymer 2001, 42, 4299. 424 [46] M. Toda, Y. Joseph, R. Berger, Journal of Physical Chemistry C 2010, 114, 2012. 425 [47] J. R. Barnes, R. J. Stephenson, C. N. Woodburn, S. J. O’Shea, M. E. Welland, T. 426 Rayment, J. K. Gimzewski, C. Gerber, Rev. Sci. Instrum. 1994, 65, 3793. 427 [48] R. Raiteri, M. Grattarola, H.-J. Butt, P. Skládal, Sensors and Actuators B: Chemical 428 2001, 79, 115. 429 [49] A. N. Itakura, M. Toda, K. Miyake, R. Förch, R. Berger, Japanese Journal of Applied 430 Physics 2013, 52, 110111. 431 [50] C. Li, Z. Huang, R. K. Wang, Opt. Express 2011, 19, 10153. 432 [51] A. Neubrand,  P. Hess, J. Appl. Phys. 1992, 71, 227. 433 [52] C. Sun, B. Standish, V. X. Yang, J Biomed Opt 2011, 16, 043001. 434 [53] D. Schneider,  T. Schwarz, Surf. Coat. Technol. 1997, 91, 136. 435 [54] G. Stoney, Proceedings of the Royal Society of London. Series A, 1909, 172. 436 [55] M. Toda, A. N. Itakura, S. Igarashi, K. Buscher, J. S. Gutmann, K. Graf, R. Berger, 437 Langmuir 2008, 24, 3191. 438 [56] R. C. Cammarata, Prog. Surf. Sci. 1994, 46, 1. 439 [57] M. Mizusawa, K. N. Stoev, K. Sakurai, Japanese Journal of Applied Physics 2003, 42, 440 3709. 441 [58] S. Sakata, Y. Inoue, K. Ishihara, Transactions of the Materials Research Society of Japan 442 2016, 41, 51. 443 [59] J. E. Mark, "Polymer Data Handbook", Oxford University Press, 1999. 444 [60] S. S. Bedair,  G. K. Fedder, Transducers '05, Digest of Technical Papers, Vols 1 and 2 445 2005, 2035. 446 [61] R. Zhang, T. Cherdhirankorn, K. Graf, K. Koynov, R. Berger, Microelectron. Eng. 2008, 447 85, 1261. 448  449  450  451  452  453  454  455     - 17 - Figures 456   457 Figure 1. This schematic diagram shows the principle of the MCS bending technique. The 458 deflection is measured along the +z direction. Here, the lateral swelling (red arrow) of polymer 459 film deposited on top of the MCS along the beam direction results in a bending of MCS. 460   461 DeflectionReference cantileverppAA filmSi chip Vertical swellingLateral swellingSensing cantileverOptical reflectionGas flowFluidic chamberW indow(Two cantilever model)PumpFour way switching valveDry N 2W et N2Mass flow controller      - 18 -   462  463 Figure 2. Calculated Young’s modulus of a thin uniform polymer film (tfilm = 100 nm) that is 464 located on a 500 µm long MCS made from silicon (tSi = 2 µm). The deflection is induced by a 465 humidity change H of 10 % while a constant expansion coefficient film is assumed.  466   467 107 108 109 1010 101110-910-810-710-610-510-4  tfilm / tSi = 0.05lMCS= 500 μmαfilm = 0.0001ΔH = 10%Young’s modulus of film, 𝐸film [Pa]Calculated deflection, [m]Si1 nm∆  (𝛼film,∆𝐻)    - 19 -  468 Figure 3 (a). The relative film thickness change, i.e. (d/d0), of ppAA films prepared with 5 W 469 and 90 W plasma power levels as a function of humidity. Δd and d0 are the changes in the 470 thickness and the thickness at 0 % humidity, respectively.  471   472 0 20 40 60 80 10002468  PPAA 5 W PPAA 90 W  Relative thickness change  [%]Humidity [%]    - 20 -  473  474 Figure 3 (b). Calculated expansion coefficients of ppAA films. Ellipsometry 475 measurements indicated that the expansion coefficient at H < 60 % is constant (blue dashed 476 line). This can indicate isotropic swelling (yellow arrows). At H > 70 %, the measured film ⊥ 477 increased up to six times for the 5 W ppAA film and two times for the 90 W ppAA film (red 478 dashed line). Here it possibly indicates hindrance in lateral expansions while vertical 479 expansion is promoted. This is depicted by different sized yellow arrows.  480  481   482 0 20 40 60 80 PPAA 5 W PPAA 90 W  Humidity [%]0.0010.0020.0030Expansion coefficient to humidity    - 21 -  483  484 Figure 4. Series of exposures of the MCSs coated with ppAA films at  different humidities. The 485 MCSs were exposed alternatively to each defined humidity value followed by exposure to dry 486 nitrogen. Switching the 4-way solenoid valve allowed rapid changes (< 100 msec) of flow in 487 the sample cell between 0 % humidity and different levels of humidity. One complete cycle of 488 increasing and decreasing humidity was performed in 18 hours. (a, c) ppAA films deposited 489 with 5 W and (b, d) 90 W plasma power.  490  491   492 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0200010000 Deflection [nm]Time [hour] PPAA 90W Reference1.0% 2.5% 5.0% 10%   0 5 10 15 20 25150010005000 Deflection [nm]Time [hour]  PPAA 5W0 5 106000400020000Deflection [nm] Time [hour]  Humidity control85%0% 0% 85%Humidity control85%0% 0%20%0.0 0.5 1.0 1.5 2.0 2.5600400200010% 5.0% 2.5% 1.0%  Deflection [nm]Time [hour] PPAA 5W Reference  PPAA 90W(a) (b)(c) (d)    - 22 -   493 Figure 5. Saturated surface stress in compressive responses at each stage of humid environment 494 for ppAA films. The black arrows indicate increasing and decreasing humidity cycles. The solid 495 green and pink lines correspond to the average stress changes calculated from increasing and 496 decreasing humidity cycles.  497   498 0 10 20 30 40 50 60 70 80 900510152025303540  Stress change [MPa]Humidity [%] ppAA 90W ppAA 5W    - 23 -      499 Figure 6. The calculated Young’s modulus using the MCS (circles and triangles) and the LA-500 SAW (squares) techniques. The uncertaininty in Efilm values for the MCS technique is dominated 501 by the error in film, which was calculated from the thickness measurements by ellipsometry at 502 different humidities. The error bars for the LA-SAW measurement correspond to measurements 503 at different points on the sample and different distances ∆l. An  error of ±3 % in the humidity 504 values based on the stability of the humidity of the measurement system during the experiments 505 was calculated in. In total, we obtained error bars of 0.07 GPa in the value of E’film. 506  507 0 20 40 60 801001000  ppAA 90W (LA-SAW)ppAA 5W (LA-SAW)ppAA 90 WppAA 5 WYoung’s modulusE[MPa]Humidity [%]ppAA 90W (MCS)ppAA 5W (MCS)ppAA 90W (a=0.0005)ppAA 5W (a=0.0005)