Complementary spectroscopic and electrochemical analysis of the sealing of micropores in hexamethyldisilazane plasma polymer films by Al2O3 atomic layer deposition

In the present study, the effects of oxygen plasma treatment and subsequent 2 nm thin Al2O3 film deposition by atomic layer deposition on about 30 nm thick hexamethyldisilazane polymer layers are investigated by using a combination of spectroscopic and electrochemical analysis. The investigations focus on the microporosity of the corresponding films and their structural changes. Upon oxygen plasma treatment, the surface near region of the films is converted into SiOx, and the microporosity is increased. Atomic layer deposition of Al2O3 on the plasma oxidized films leads to the decrease of pore sizes and an effective sealing. A correlation between the film microporosity and the change of hydroxyl groups of the films with the adsorption of water was established by ellipsometric porosimetry and in situ Fourier transform infrared (FTIR) spectroscopy. Moreover, electrochemical analysis provided complementary information on the electrolyte up‐take in the differently conditioned thin films.


| INTRODUCTION
The study of plasma polymer thin films has become one of the most attractive topics in recent years, owing to the widespread applications of these films in various fields, including biomedicine, 1 electrics, and optics. 2 A common method is to manufacture thin films by plasma enhanced chemical vapor deposition (PE-CVD) using organosilanes as precursors, of which hexamethyldisiloxane (HMDSO) and hexamethyldisilazane (HMDSN) are the predominantly used precursors.Depending on the deposition conditions and parameters, films with different properties can be obtained.Exemplarily, superhydrophobic coatings deposited from pure HMDSO monomers show resistance to microorganisms, 3 hydrophilic and low cross-linked films formed from a mixture of HMDSN monomer and oxygen can be used for humidity sensors, 4 while nanoporous SiO x films based on PE-CVD act as carrier layer for liquid infusion and also as low dielectric constant materials for electronic devices. 5,6 a recent publication, 7,8 de Freitas et al. summarized in detail the correlations between the deposition mechanisms and the chemical structure of HMDSO-based films, outlining the prospective modifications of properties.Charifou et al. pointed out that SiO x and SiOCH films deposited on polymer substrates (such as polyethylene terephthalate and polypropylene) can improve the barrier properties (i.e., these deposited layers can improve polymer properties to prevent oxygen and water from passing through the polymer substrate); however, these effects are limited for enhancing the barrier to water.0][11] In this context, the ALD technique is relied on due to its precise controllability of the film thickness on the sub-nanometer level and its ability to provide highly conformal and uniform surface coatings. 12Exemplarily, it was demonstrated that the water vapor transmission rate of polyethylene substrates coated with an ultrathin Al 2 O 3 barrier layer (about 2 nm) could effectively be reduced from around 1.0 gÁm À2 Áday À1 to around 10 À3 gÁm À2 Áday À1 . 135][16][17] Based on these studies, water vapor transmission rat measurements have been adopted as a valid method for the evaluation of the barrier film permeability 18,19 ; however, this approach only allows for the identification of macroscopic defect densities.For determining pores and defects on the nanoscale, other methods need to be employed.In this regard, Hoppe et al. quantitatively studied the pore size and distribution in the SiO x films by means of positron annihilation lifetime spectroscopy.The authors demonstrated a correlation between the applied plasma treatment parameters and the features of the pores, establishing the possibility to adjust the porosity of the films. 20,21other more common and convenient tool used to define porosity is ellipsometric porosimetry (EP) measurements, [22][23][24] which is based on the changes in the optical properties of the film that occur when probing molecules (such as N 2 at T = 77 K, CO 2 at T = 273 K, water vapor and organic vapors) adsorb on the surface.The corresponding porosity of the films is then identified by the adsorption isotherm according to the International Union of Pure and Applied Chemistry (IUPAC) classification. 25 Perrotta et al. reported   that the corresponding studies on PE-CVD deposited SiO 2 films by using water as an adsorptive for EP measurements.The results showed that the nanoporosity of the films can be classified and quantified by EP.In combination with electrochemical impedance measurements (EIS), additional information about macroscopic defects can be obtained. 26,279][30] Therefore, it is additionally important to study the kinetics of chemical structure changes (particularly for OH groups) when the films are exposed to a humid environment.
In this regard, quartz crystal microbalance (QCM) is considered as a sensitive method to study the water adsorption behavior on the films.Grundmeier et al. analyzed the water adsorption behavior in HMDSO-based films by combining QCM with dissipation monitoring (QCM-D) and infrared spectroscopy. 31,32They showed in detail the changes in hydroxyl groups with increasing relative humidity and the specific contribution of each hydroxyl component.

| Materials
The substrates employed in this study were QCM crystals with Cr/Au coating (6 MHz, 13.98 mm ± 0.02 mm) from Demaco Holland BV, 200 nm thick Au coated Si(100)-wafers (size 15 mm Â 15 mm) and 200 nm thick mica template stripped ultra-flat Au substrates. 33To remove surface contaminations from the substrates, they were cleaned with an argon plasma as a pretreatment prior to each plasma deposition process.HMDSN ((CH 3 ) 3 Si-N-Si(CH 3 ) 3 ) (≥99%, Sigma-Aldrich) was used as monomer for plasma deposition without further purification.

| HMDSN plasma polymer deposition
The HMDSN plasma polymer films were deposited by PE-CVD in a plasma bell jar reactor.The plasma was ignited by an audio frequency power supply (3.5 kHz) with a current of 1.2 mA and a voltage of 300 V.The base pressure of the reactor was below 10 À5 mbar.
A mixture of argon (0.2 mbar) and HMDSN monomer (0.05 mbar) was used in the plasma deposition process.The total flow rate was 1.0 mL min À1 .This reactor was also equipped with a Cr/Au coated 5 MHz-QCM crystal (Fil-Tech Inc.) that allowed simultaneous deposition of thin films as well as monitoring of the growth rates.More details about the setup and deposition parameters have been described in a previous literature. 34

| HMDSN plasma polymer oxidation
A separate capacitively coupled plasma chamber was used to accomplish the oxygen plasma treatment, where the samples can be characterized directly by discrete polarization modulated IRRAS (DPM-IRRAS) after plasma oxidation (see Figure 1).A radio frequency electrode provided by Advanced Energy Industries Inc.
(f = 13.56MHz) was used for power supply.Detailed description of the setup can be found in the previous literature. 34In this study, a power of 63 W was applied and the DC bias voltage was 106 V.The pressure of the chamber was kept at 10 À4 mbar prior to the experiments.Afterwards, the plasma polymer films were exposed to an oxygen plasma with a mixture of Ar (2 sccm) and O 2 (4 sccm) at a working pressure of 10 À1 mbar for 10 min.

| Thermal ALD of Al 2 O 3
Within this study, a custom built ALD reactor (modularflow) was used to apply 2 nm thick Al 2 O 3 top coatings.Trimethyl aluminum (TMA) was employed as Al precursor and maintained at 30 C while distilled water (H 2 O) was used as co-reagent and kept at room temperature.
All depositions were carried out at a base pressure of 1.7 mbar and a deposition temperature of 120 C. Carrier gas flows for precursor and co-reagent were adjusted to 10 sccm each.All substrates coated in this study were introduced to the reactor chamber as received and kept in the heated chamber under active vacuum for 20 min prior to deposition to allow the temperature to equilibrate.The optimized Al 2 O 3 ALD pulse/purge process sequence at 120 C comprised an 8 ms TMA pulse followed by a 6 s purge, a 100 ms H 2 O pulse and then again, a 40 s purge.Given that the growth per ALD cycle amounted to 1.2 Å, 17 ALD cycles were applied to reach the targeted Al 2 O 3 coating thickness of 2 nm.

| DPM-IRRAS measurement
Discrete polarization modulated IRRAS (DPM-IRRAS) measurements were carried out in the plasma oxidation chamber, as shown in Figure 1.The spectrometer used for the DPM-IRRAS measurements was a Bruker Vertex 70 having a mercury cadmium telluride detector cooled with liquid nitrogen.The modulation of p-and s-polarization was performed by a rotatable KRS-5 polarizer, the details of which can be found elsewhere. 34The spectra were recorded in the wavenumber range of 3800-700 cm À1 at an incidence angle of 75 .The scans number was set to 256, and energy resolution was 4 cm À1 .All the measurements were performed under the same vacuum conditions of 10 À4 mbar.Au-coated Si wafer was used as a reference and as a substrate.

| In situ QCM/DPM-IRRAS measurement
In situ QCM/FT-IRRAS measurements were carried out in the same plasma oxidation chamber as described above, which allowed control of the relative humidity and simultaneous characterization of the surface chemical changes as well as mass changes of the samples by DPM-IRRAS and QCM measurements.A mixture of dry and humid with water content streams was used to provide a humidified atmosphere.The gas flow rate was adjusted by mass flow controllers to control the relative humidity while using nitrogen as carrier gas.
The chamber was evacuated to a base pressure (10 À4 mbar) for 60 min prior to the water adsorption measurements, followed by a purge with dry nitrogen at a flow rate of 1.5 L min À1 for 120 min.The humidity was increased stepwise and monitored by a humidity sensor (Rotronic GmbH), and after stabilizing the frequency of the QCM at each step, DPM-IRRAS measurements with an incidence angle of 75 were performed.The number of scans was 256, and the energy resolution was 4 cm À1 .A 1-octadecanethiol self-assembled monolayer modified QCM substrate was measured at 0% humidity as a reference to avoid water adsorption on the bare QCM surface.All the measurements were carried out at the same temperature, and QCM crystals with Cr/Au coating were used as substrates for in situ QCM/FT-IRRAS analysis.

| Ellipsometric porosimetry (EP)
The EP measurements were carried out in an ellipsometer (auto nulling ellipsometer, Ep3, Accurion GmbH), equipped with a vacuum F I G U R E 1 Schematic of experimental setup for oxygen plasma oxidation and in situ QCM/IRRAS measurements.
cell and a controllable humidity setup (Controlled Evaporator Mixer, Bronkhorst High-Tech B.V.) adjusted by a water mass flow controller and a N 2 mass flow controller.This system allows for the determination of the change in refractive index with relative humidity at an incidence angle of 60 .Prior to EP measurements, the thickness and refractive index of dry films were measured in the wavelength range of 363.7-905.9nm using Xenon lamp as the incident light source.The measurements were performed at a base pressure of 10 À5 mbar.The data were fitted by a Cauchy optical model with the absorption coefficient k(λ) = 0, which is suitable for transparent films. 34e EP measurements were done using four-zone nulling at a base pressure of 10 À5 mbar.The incident light source was laser with a wavelength of 658 nm.The adsorptive is ultra-pure water with a carrier gas N 2 .The relative humidity was increased stepwise from 0% to 92% (8% as a step value), and the refractive index of the film was recorded at each relative humidity.All measurements were performed at a temperature of 25 C.The data were determined by a n_k fix from "EP4Model" software (Accurion GmbH), which is used for the single wavelength ellipsometry.In all cases, the error values of the refractive indexes were about ±0.003.The measurements were performed using Au-coated Si wafers as substrates.
The adsorptive volume fractions are related to the refractive indices as function of relative humidity.Based on the Lorentz-Lorenz equation, the volume fraction of open pores that are filled with adsorptive V filled at each relative humidity can be calculated by the following equation: where n 0 is the refractive index value of the film with empty pores, n fill is the refractive index value of the film with filled pores, and n vapor is the refractive index value of the adsorptive.In the present study, the value of n 0 was measured at vacuum, and n vapor is the refractive index value of water, 1.331.When all the nanopores are filled with the adsorptive, the open porosity of the films can be determined.
According to the IUPAC classification, 25 the adsorption isotherm can be identified from the shape of the relative adsorptive volume as a function of relative humidity.Type I isotherm is associated with the microporous materials with pore size less than 2 nm.Type II isotherm is observed in the nonporous or macroporous materials with pore size larger than 50 nm.Type III isotherm describes the weak adsorbentadsorbate interactions of the nonporous or macroporous materials.In the case of Type IV isotherm, the mesoporous materials with pore size between 2 and 50 nm can be determined.

| Atomic force microscopy
The equipment used for the AFM analysis was a JPK Nanowizard II from JPK Instruments AG, which was equipped with an antivibration box.In this study, an NSC-15 type beam silicon cantilever (Mikromasch) with a resonance frequency of 325 kHz was used to record the topographic images.The scan resolution was 512 Â 512 pixels, and the scan size was 2 Â 2 μm 2 .Two hundred nanometers of thick mica template stripped Au substrates were used as substrates because of their low surface roughness.utilizing a dual-beam mode.A primary beam of 30 kV Bi 3 + ions with a sputter area of 100 Â 100 μm 2 was used, while a secondary beam of 1 kV Xe + ions was applied to a sample analysis area of 400 -Â 400 μm 2 .In this study, Al 2 O 3 ALD samples were measured for depth profiling analysis, using Au-coated Si wafers as substrates.

| X-ray photoelectron spectroscopy
Table 1 shows the surface composition of the PP films, obtained from the area of the core level XPS spectra.The XPS core level peaks with fitted components are shown in Figure 3.The Si2p peaks were fitted with the spin-orbit doublet (Si2p 3/2 and Si2p 1/2 ) with a peak separation of 0.6 eV and an intensity ratio of Si2p 3/2 :Si2p 1/2 = 2:1. 34For the fitting of Al2p peak, the spin-orbit doublet (Al2p 3/2 and Al2p 1/2 ) was considered with a peak separation of 0.44 eV and an intensity ratio of Al2p 3/2 :Al2p 1/2 = 2:1. 35e as-deposited HMDSN PP film shows a typical SiO x N y C z H composition with atomic ratios of 1.6 for C/Si, 0.3 for N/Si, and 0.4 for O/Si(SiO 0.4 N 0.3 C 1.6 H).The corresponding Si2p peak shows the formation of (-C-) 2 Si(-O-) 2 /Si-N and (-C-)Si(-O-) 3 components at 101.9 and 103.0 eV, respectively (see Figure 3). 34,36A main component of O1s peak was fitted at about 532.8 eV, assigned to Si-O groups. 34ter oxygen plasma treatment, the films show a decrease in nitrogen content, and a partially oxidized surface near region with an increased O/Si atomic ratio of 1.5 and a decreased C/Si atomic ratio of 0.5 in comparison with the as-deposited PP film.
The Si2p peak showed two fitted components at 101.9 and 103.5 eV, corresponding to (-C-) 2 Si(-O-) 2 and SiO x , respectively. 20,36ter the ALD of Al    surface near region to a SiO x -like structure after oxygen plasma treatment of the HMDSN PP film.The significant increase at the beginning of the sputtering (see the signals of Si-CH 3 + and SiO + ) could be due to the saturation of the detector caused by species on the outer surface, which has influence on the total ion intensity in the surface near region. 39,40cusing on the distribution of AlO + (mass: 43), except for the detected signal on the surface due to the deposition of Al 2 O 3 , a large intermediate phase of the AlO + fragment (down to about 15 nm depth) also can be observed, accompanied by an intermediate phase of SiO + fragment.This could be explained by the formation of porous structure after plasma oxidation, leading to the partial diffusion of ALD precursor (TMA) into the inner SiO x -like film.Also, note that the signal of AlO + was not detected after 15 nm depth, while the Si-CH 3 + fragment appeared, indicating the inner film with residual Si-CH 3 is not coated by Al 2 O 3 .

| DPM-IRRAS
Figure 5 shows the IRRAS data.For the as-deposited HMDSN PP film, the typical bands appear at around 935 and 1180 cm À1 corresponding to the Si-NH-Si stretching vibration 41 and N-H bending, 42 respectively.The peak centered at 1050 cm À1 is associated with a mixture of Si-CH 2 -Si wagging and Si-O stretching vibration. 43,44The Si-H stretching vibration is located at around 2140 cm À1 . 42The peaks at 846 and 1257 cm À1 correspond to the Si(CH 3 ) x rocking mode and Si-CH 3 bending mode, 44 while the peaks at 2958 and 2900 cm À1 are associated with the asymmetric and the symmetric CH 3 stretching vibrations. 42The corresponding IR peaks were shown in Table 2.After ALD with Al 2 O 3 , IRRAS showed an increase of the peak between 850 and 1000 cm À1 compared with the plasma oxidized HMDSN PP films (see Figure 5, red and blue spectrum).The broad band between 3800 and 3000 cm À1 can be observed, corresponding to the OH stretching vibration of the Si-OH and the Al-OH groups.
Compared with the IR spectra of the oxidized HMDSN PP film before ALD, the corresponding OH peaks decreased slightly, indicating the reaction of TMA with hydroxyl groups, as discussed in the ToF-SIMS results above.
For the as-deposited 2 nm ALD-Al 2 O 3 film (see Figure 5, brown spectrum), a significant Al-O bond centered at 938 cm À1 is observed. 45,46However, this peak is not as dominant for Al 2 O 3 deposited on the plasma oxidized PP film (see Figure 5, blue spectrum).
Instead, in this case, an increase in the peak between 850 and 1000 cm À1 was observed.This finding might be explained by the formation of Si-O-Al bonds in the porous structure. 47,48Additionally, the Al-OH bending mode at 1156 cm À146 and the C-H x bending mode in Al-CH x at 1234 cm À149 were not observed in the case of Al 2 O 3 deposited on the oxidized plasma polymer film (blue spectrum).
This could be due to the overlap of Si-O-Si stretching vibration at 1230 cm À1 .

| Ellipsometric porosimetry measurements
The EP isotherms of the films are shown in Figure 6.The open porosity in the investigated films was obtained using Equation ( 1) and averaging over three measurements performed on three identially prepared samples.The error bars represent the dispersion values between the three measurements.
For the as-deposited HMDSN PP film (see Figure 6A), the shape of the adsorption isotherm is close to type III isotherm (non-porous film), and no apparent hysteresis loop was observed in the desorption isotherm.This adsorption behavior is caused by the nonpolar Si-CH 3 groups of the surface, which reduces the interaction between the surface and the adsorptive water.Therefore, the adsorption of water is small at lower relative humidity; however, once a monolayer of molecular water is adsorbed on the surface, the further adsorption of water will be promoted. 50In the case of HMDSN PP film, the monolayer of adsorbed water is formed at around 45% of relative humidity with open porosity of about 0.31%.
After plasma oxidation of the HMDSN PP film (see Figure 6B), the surface becomes more hydrophilic, due to the formation of hydroxyl groups in the film (as shown in the FTIR data).The adsorption isotherm showed a higher amount of adsorbed water in comparison with the PP film before oxygen plasma.Also, note that a significant hysteresis loop caused by capillary condensation was observed at higher relative humidity in the desorption process. 25,50us, the isotherm shape after oxygen treatment converted toward a type IV isotherm for a mesoporous film.However, a steep portion was observed in the initial adsorption process; this could be explained by the presence of micropores (a type I isotherm).Therefore, the water adsorptions at around 25% and 50% relative humidity are associated with the filling of micro-and mesopores, respectively, and the open porosity of the film is around 2.1%.At higher relative humidity, the uptake of water is increased, indicating the formation of multilayer adsorption.
After ALD with Al 2 O 3 (see Figure 6C), a decrease in the water adsorption was observed.The adsorption isotherm changes toward a composite of type I and II isotherms, which is represented by the increase of water adsorption at the very low relative humidity.The open porosity is reduced to 0.5% compared with the film before ALD.
In summary, the EP results show the significant differences in pores of the investigated films.Although the HMDSN PP film showed a type III isotherm for a nonporous film, this does not mean that the film is free of micropores.The results also show a low open porosity   this change is obvious in the case of oxidized PP film (see Figure 7B), indicating that the formation of hydroxyl groups and pores after oxygen plasma treatment leads to water up-take in the film.However, after the deposition of Al 2 O 3 (see Figure 7C), the change in the intensity of the hydroxyl groups is not as pronounced compared with the oxidized PP film at 90% relative humidity.This could be explained by the reduction of OH groups and pore size of the film after Al 2 O 3 deposition.The significantly high hydroxyl peak area for the PP film after oxygen plasma conditioning indicates a high amount of adsorbed water on the surface, while the change is not obvious in the case of the as-deposited HMDSN PP film.After Al 2 O 3 coating, the variation of the hydroxyl group with relative humidity is overall lower than that in the oxidized film as already shown in the QCM analysis.
The combination of in situ QCM and FTIR measurements indicates that the deposition of Al 2 O 3 by ALD reduces the adsorption of water on the surface; however, it still shows a higher amount of water uptake in comparison with the as-deposited HMDSN PP film.These differences are even significant at higher relative humidity (>50%).
The reason could be due to the partial infiltration of the ALD precursor in the oxidized PP film, leading to a decrease of pore size.
Figure 9 shows the difference spectra of OH groups for the HMDSN based films in the range of 3800-2500 cm À1 at 90% humidity.As described in previous works, [30][31][32]51,52 the OH group consists of different components: the "dangling" OH bonds at around 3600-3700 cm À1 correspond to isolated silanol groups of the surface; "liquid-like" component at around 3400 cm À1 , and "ice-like" component at around 3200-3300 cm À1 are associated with the hydrogen-bonded OH stretching vibrations, indicating the presence of adsorbed water on the surface.
It can be seen that the change of the "dangling" OH bonds are not remarkable in all types of films.The appearance of the "ice-like" components at high humidity indicates the formation of a first layer of adsorbed water at the surface.In all cases, the dominant OH stretching vibration is contributing to the "liquid-like" component; this is due to the water molecules in the second layer bonding with more bulk water.In comparison with the oxidized PP film without Al 2 O 3 top coating, a significant decrease of the "liquid-like" component is observed in the Al 2 O 3 -coated film, and the amount of water adsorbed on the surface is only slightly higher than in the as-deposited HMDSN film.These results are in a good agreement with the EP data.

| Electrochemical impedance measurements
EIS measurements allow for the analysis of changes in impedance values of various porous films caused by the penetration of electrolyte through defects and/or pores from the outer surface down to the gold substrate.Moreover, the film capacitance and uptake of electrolyte could be determined.
Figure 10 shows the Bode plots as a function of immersion time for the investigated films.In the case of the as-deposited HMDSN PP film (see Figure 10A), a clear capacitive behavior was found, represented by the phase shift closed to 90 .After oxygen plasma treatment (see Figure 10B), the film shows a drop in impedance and a lower phase value in the low frequency region.This is attributed to the effect of electrolyte/substrate interfacial charge transfer, indicating the electrolytes penetration through the formed pores, which is also confirmed by the EP and FT-IRRAS results.
Compared to the oxidized PP film, the film after Al 2 O 3 by ALD shows a similar behavior (see Figure 10C); however, a slight increase in phase shift at lower frequencies, and this could be due to the decrease of pores in the film after ALD.
The penetration of electrolyte into the film has an influence on its dielectric constant, leading to an increase in the capacitance of the film.Using the following equation, the film capacitance can be obtained 53 : where C p is the capacitance of the investigated films; f is the frequency, in the present study, a constant frequency at the maximum of the phase shift was used; Z " is the imaginary part of the impedance.
Based on the changes in the film capacitances, the volume of electrolyte uptake φ can be determined using the Brasher-Kingsburg equation 53,54 : where C 0 and C layer are the capacitances for "dry" film (fitted with a polynomial function and obtained by extrapolating t = 0) and the film In situ DPM-IRRAS difference spectra of HMDSN based films in the range of 3800-2500 cm À1 at 90% relative humidity.In all cases, the spectrum at 0% RH was subtracted from each corresponding spectrum.In situ QCM/FTIRRAS analysis indicated that the water adsorption is dependent on the nanoscopic structure on the surface.The hydroxyl-rich oxidized HMDSN SiO x -like plasma polymer films showed a significant increase of adsorbed water in comparison with the as-deposited HMDSN film.However, ALD of ultrathin Al 2 O 3 led to a reduced water adsorption based on the sealing of micropores.
The QCM/FTIRRAS and EP results were in good agreement with electrochemical impedance studies which furthermore provide information on the presence of defects in the films.
Based on previous investigations, it is meaningful to study the effect of ALD coatings on porous films, especially to study the variation of surface pores after ALD as it directly affects the barrier properties of the films.The aim of the presented work here was to analyze the adsorption of water as a function of film microporosity and pore surface chemistry.A combination of in situ QCM/FTIR reflection absorption spectroscopy (FT-IRRAS) and EP measurements in combination with electrochemical analysis was utilized to evaluate the microporosity and water as well as electrolyte adsorption characteristics of HMDSN-based films.Moreover, a post plasma oxidation of plasma polymer films and a 2 nm coating of Al 2 O 3 by thermal ALD were employed to modify the surface chemistry and the pore size.Ex situ X-ray photoelectron spectroscopy (XPS) measurements and atomic force microscopy (AFM) topographic measurements were performed to analyze the chemical composition of the thin films in the surface near region and the surface roughness, respectively.ToF-SIMS sputter profiling as a complementary method was used to analyze the depth profile and show the effect of ALD on the films.

2. 3
.5 | X-ray photoelectron spectroscopy XPS measurements were performed with an Omicron ESCA+ system (Omicron NanoTechnology GmbH) at a base pressure of 10 À10 mbar.The X-ray source was a monochromatic Al Kα (1486.7 eV).All the measurements were done at a take-off angle of 30 with respect to the surface plane, without neutralization since the investigated films were approximately 35 nm thick in all cases.Pass energies of 100 eV with a step size of 0.5 and 10 eV with a step size of 0.1 eV were set for recording survey and core level spectra, respectively.For the plasma oxidized HMDSN sample and the Al 2 O 3 deposited sample, the binding energy scale was calibrated by fixing the adventitious C1s peak at 284.8 eV.For the HMDSN plasma polymer sample, the main component of the C1s peak at 285.0 eV was used for calibration.A Shirley-type background was utilized for quantification.By integrating each peak area after subtraction of the background, the atomic composition of the film can be determined, where the corresponding relative sensitivity factors were also considered.Au-coated Si wafers were used as substrates.2.3.6 | Time-of-flight secondary ion mass spectroscopy (ToF-SIMS) ToF-SIMS measurements were carried out in a ToF-SIMS 5-100 (IONTOF GmbH) (at the ICAN of the University of Duisburg-Essen), Electrochemical impedance measurements were performed to analyze the electrolyte (an aqueous solution of 10 mM K 4 [Fe (CN) 6 ] and 0.1 M KCl) up-take of the studied films, for this a three-electrode cell with a Reference 600 potentiostat (Gamry Instruments) was utilized.The working electrode was the investigated film deposited on Aucoated Si wafer with an exposed area of 0.196 cm 2 .A gold wire and a saturated Ag/AgCl electrode were used as the counter and the reference electrode, respectively.The frequency range was from 10 À1 Hz to 10 5 Hz (10 points per decade), and the applied amplitude of the sine-wave voltage signal was 20 mV (rms).Every 5 min the impedance plots were recorded.

3 | RESULTS AND DISCUSSION 3 . 1 |
Surface characterization of HMDSN based plasma polymer (PP) films3.1.1 | Atomic force microscopyAFM measurements were carried out to characterize the morphology of the thin films (see Figure2).In each case, three identically prepared samples were measured.All types of films show smooth and homogenous surfaces with similar morphologies.The surface roughness of the PP film decreased after oxygen plasma treatment, but slightly increased after the Al 2 O 3 ALD film was applied.

Figure 4
Figure 4 shows the ToF-SIMS depth profiles of ion fragments (AlO + , SiO + , and Si-CH 3 + ) for the oxygen plasma treated HMDSN PP film followed by the deposition of 2 nm Al 2 O 3 by means of thermal ALD.The thickness of the measured film determined by ellipsometry was about 35 nm, and the sputter depth can be calculated from the sputtering time and the film thickness.It can be observed that the intensity of SiO + (mass: 44) reaches a maximum value at a depth of about 10 nm, while the signal of Si-CH 3 + (mass: 43) was not detected, indicating the conversion of the F I G U R E 3 C1s (A), O1s (B), and Si2p (C) XPS core level spectra of the HMDSN based films.(D) Al2p XPS core level spectra of HMDSN PP film after 10 min O 2 plasma and deposition of 2 nm Al 2 O 3 by thermal ALD.

F I G U R E 4 3 +
ToF-SIMS depth profiles of AlO + , SiO + , and Si-CH positive ions for the HMDSN PP film after 10 min O 2 plasma and deposition of 2 nm Al 2 O 3 by thermal ALD.F I G U R E 5 DPM-IRRAS data in the range of 3800-700 cm À1 for the as-deposited HMDSN PP film (black spectrum), 10 min O 2 plasma treated HMDSN PP film (red spectrum), the HMDSN PP film after 10 min O 2 plasma and deposition of 2 nm Al 2 O 3 by thermal ALD (blue spectrum) and as-deposited 2 nm Al 2 O 3 film by thermal ALD (brown spectrum).Au-coated Si wafer were used as substrates and served as a reference in all cases.T A B L E 2 Assignment of the IR peaks of the investigated films.Assignment Peak position (cm À1 ) References ν of HMDSN PP film, the cleavage of Si-C and Si-N bonds is accompanied by the formation of Si-O bonds, represented by the decrease of alkyl and nitrogen groups as well as the increase and shift of the peak at 1050 cm À1 compared with the PP film before oxygen plasma.The appearance of the asymmetric stretching Si-O-Si band at 1230 cm À1 also illustrates the formation of the SiO x structure.Additionally, the intensity of the Si-H band decreases after plasma oxidation.The increase of dangling hydroxyl group between 3800 and 3000 cm À1 promotes the appearance of nanoscopic pores.

F I G U R E 6
Adsorption isotherms for HMDSN based films as function of the relative humidity (filled symbols: adsorption, empty symbols: desorption).Water was used as an adsorptive in all cases.value of about 0.31%.Compared to the as-deposited PP film, the oxygen plasma treatmsent leads to the formation of dangling hydroxyl groups in the film, resulting in the presence of nanoscopic pores.The open porosity is increased to around 2.1%.With the ALD-Al 2 O 3 samples, the pores of the film are effectively reduced, and while the results indicate that micropores are still present in the film, the open porosity is decreased.

Figure 7
Figure7shows the FTIR data of the investigated films at 0% and 90% relative humidity.The structural changes of the films at high relative humidity are mainly represented by the increase of hydroxyl group intensity, indicating the adsorption of water on the surface.It is clear to see that the change of OH group is not significant at 90% relative humidity for the as-deposited HMDSN PP film (see Figure7A), while

Figure
Figure 8A shows the frequency change of QCM crystals with relative humidity.The frequency decrease of the crystal reflects the water adsorption for different relative humidity values.For the as-deposited HMDSN PP film, the change in frequency is initially fast and then gets slower for relative humidity values between 20% and 50%.After 50% humidity, a rapid shifting of frequency was observed.Similar behavior also can be found in the frequency change of the oxidized PP film and the Al 2 O 3 coated oxidized film.By comparison, the analysis shows a significantly higher shifting of frequency after plasma oxidation, indicating that the formation of hydrophilic pores promotes the adsorption of water on the surface.In the case of at any time t, respectively; ε w is the permittivity value of pure water (78.4 at 25 C).As shown in Figure11A, it is observed that the film capacitance increases after oxygen plasma treatment, while the film capacitance values after ALD show lower values.Similar behavior is present in the changes of electrolyte uptake (see Figure11B), the PP film after plasma treatment shows a significant higher φ value in comparison with the as-deposited PP film, indicating that the plasma oxidation promotes the uptake of aqueous electrolytes because of the formation of hydrophilic hydroxyl groups in the film.The reduction of electrolyte uptake after ALD can be due to the decrease of open pores, as already discussed in the EP and QCM/FT-IRRAS results, revealing the effective modification of the film by ultrathin ALD-Al 2 O 3 film.Additionally, as mentioned above, the determination of water uptake by EP measurement can provide information on nanoscopic pores based on the variation of the film optical properties, namely the refractive indexes of the films.Complementary to the EP measurements, EIS measurements have proven to be a method for detecting both nanoscopic pores and macroscopic defects.26,27Because the electrolyte uptake φ obtained by EIS is related to the open porosity in the film, it can be observed that the φ values of the studied films are all slightly higher than the open porosity obtained by EP, even at a short immersion time, illustrating the potential presence of defects in the films other than open pores.4 | CONCLUSION The combination of spectroscopic and electrochemical analyses provided information on the chemical structural changes in HMDSNbased plasma polymer films before and after deposition of Al 2 O 3 by ALD, as well as on the porosity and defects of these films.The results of XPS and ToF-SIMS depth profiles serve as complementary information showing that the surface near region of the HMDSN PP films are converted to a SiO x -like film after oxygen plasma treatment and that the ALD-deposited Al 2 O 3 ultrathin films formed in the surface near plasma oxidized region of the films.It could be demonstrated by DPM-IRRAS and EP measurements that dangling hydroxyl bonds were formed in the HMDSN plasma polymer after oxygen plasma treatment and increased the open porosity of the films.By comparison, the porosity of the oxidized PP films was effectively reduced after ultrathin Al 2 O 3 sealing by ALD.

F I G U R E 1 1
Evolution of (A) the film capacitance at a fixed frequency and (B) the electrolyte uptake (φ in %) for the HMDSN based films within 20 min of immersion in K 4 [Fe (CN) 6 ] (10 mM) electrolyte.The corresponding values were averaged over three measurements performed on three identically prepared samples.F I G U R E 1 0 Bode plots of (A) HMDSN PP film, (B) 10 min O 2 plasma treated HMDSN PP film, and (C) the HMDSN PP film after 10 min O 2 plasma and deposition of 2 nm Al 2 O 3 by thermal ALD within 20 min of immersion in K 4 [Fe (CN) 6 ] (10 mM) electrolyte.
2 O 3 , Al is visible in the XPS spectrum, showing two components at 74.2 and 75.2 eV, assigned to Al-O and Al-OH groups, respectively. 37,38The component of the O1s peak fitted at 531.0 eV is assigned to Al-O bond in Al 2 O 3 .Additionally, the Si2p and O1s peaks shifted to lower binding energies after Al 2 O 3 deposition compared with the plasma oxidized HMDSN PP films, which could be explained by the involvement of Al 2 O 3 in the silicon oxide environment.