1 Introduction

Thermal processing of plastics by injection molding is a well-established process technology for the mass production of plastic products of complex shape [1]. Based on this process polymeric parts e.g. micro-fluidic devices (“lab-on-a-chip”) and lenses with high surface quality are produced [2].

However, the injection molding process is affected and destabilized by reactions at the polymer melt/mold interface [3]. Strong interactions between the polymeric phase and the mold alloy surfaces can result in increased demolding forces and thereby limit the accessible dimensions and aspect ratios [4]. In addition, the formation of insoluble interfacial deposits has been observed. Such deposits may change their composition over time, de-adhere from the interface and finally enter the polymer melt by shear-induced stripping [5]. Thereby, such deposits form discolored speckles in product surfaces [5]. As an important example, polycarbonate (PC) tends to thermally degrade and adhere on FeCr- alloy surfaces during melt processing [6].

In principle, the degradation can occur via various routes depending on external factors. Exposing the polymer to heat leads to thermal degradation, while the addition of oxygen to the process causes thermal oxidation. If the polymer is in contact with chemicals, especially water, hydrolysis can also affect the degradation process. Moreover, irradiation of the polymer with UV light, blue light or γ-rays can lead to different kinds of photo- or γ-rays degradation [7]. In this work these irradiation-based degradation pathways should however not play a major role in the degradation process of the polymer.

The degradation mechanisms of PC have been discussed many times in literature [7,8,9,10,11,12,13,14,15,16,17,18]. Since there are many reaction pathways that overlap and influence each other, not one singular degradation mechanism, which might lead to a singular degradation product or at least a small selection of products, has been established as the dominant one. However, there are a number of possible routes in the degradation process, that have been agreed upon, and which are briefly discussed below in connection to the expected results of this work.

The binding energies in a bisphenol A polycarbonate unit suggest, that the C–C bond between the methyl group and the isopropylidene bridge is the weakest bond [7, 16, 19]. During thermal degradation this bond can break resulting in a radical, which then can undergo a multitude of further reactions, including branching and recombination [7]. In the case of thermal oxidation the presence of oxygen can lead to the formation of peroxide radicals from the reaction between oxygen and the original radicals [7, 11, 16]. Those peroxide radicals can then follow similar degradation pathways as the radicals in oxygen-free thermal degradation [7, 16]. Since the polymer is heated under nitrogen atmosphere in our case, we expect that the degradation mostly follows an oxygen-free mechanism. It is noted however, that the presence of oxygen mainly affects the initiation of the degradation, the following reaction pathways are fairly similar in both cases, leading to similar products [16].

An additional reaction pathway involving rearrangement of the carboxyl group during the thermal degradation of PC has been described [7, 8, 12]. It involves the release of CO2 and the formation of ethers. Multiple different reactions can follow and lead to a selection of intermediate species and degradation products. For this study, the most relevant of them are phenyl-2-phenoxybenzoate and dibenzofuran, because they cause a fluorescence signal which is measurable with Raman-fluorescence microscopy [8, 20].

The second weakest bond in the monomer unit is the C–O bond in the carbonate group [7, 16, 19]. It can break during the already mentioned rearrangement reactions, but it is also susceptible to hydrolysis reactions, which yield phenol end groups [7, 12]. The polycarbonate in this study is dried before it is heated onto substrates. However, it should be noted, that even if there is no water present in the beginning of the process, H2O can be generated during thermal oxidation processes and then lead to hydrolysis reactions [16].

In addition to factors such as temperature or humidity [8, 21], the type and oxidation state of metal ions at the interface to the melt also play a role for the degradation kinetics [6, 20, 22]. This is particularly important with regard to the use of high-alloy steels. Wiesing et al. showed that, in the case of tools made of steels, mainly Fe(II) species act as catalytic sites for the degradation of polycarbonate [20]. Moreover, the authors illustrated the influence of thin TiAlN films [20] and adsorbed perfluorinated monolayers on the deposit formation and adhesion of PC [23]. Chromium-containing coatings have also been studied [24, 25]. FeCl3 and γ-Fe2O3 have also been shown to increase the degradation of PC [26, 27]. Blazsó [28] suggested that iron chlorides might promote the fission of the chain at the isopropylidene-phenylene bond. Gupta and Viswanath [6] argued that the peroxide radicals might interact with metal oxides during thermal oxidation. Depending on the metal this could increase or decrease the activation energy of following reaction steps and influence the rate of the degradation.

In addition to such interfacial polymeric degradation processes, corrosion of the alloy can occur when a critical concentration of water in the melt is exceeded [3]. Such interfacial corrosion processes lead to roughening of the interface and an increase in the metal release rate. Thereby, the formation of detrimental deposits is further promoted.

In this work, a combinatorial thin film approach [29,30,31,32,33,34,35,36] is applied for the investigation of the influence of the composition of Fe–Cr alloys on the interfacial degradation behavior of polycarbonate melt. For this purpose, Fe–Cr alloys films with a variable Fe:Cr ratio were synthesized and used for combinatorial analysis of the related interfacial degradation of PC. Corresponding gradient alloy films were co-sputtered on reference substrates and the formed insoluble interfacial PC films after melt contact were analyzed by means of spectroscopic (FTIR-, XPS-, fluorescence spectroscopy) and microscopic (imaging ellipsometry and AFM) as a function of the alloy surface chemical composition. Thereby, the influence of the surface concentration of Fe in the FeCr-alloy on the formation of PC residues during melt contact could be revealed.

2 Experimental

2.1 Materials and chemicals

Si wafers (Siegert Wafer, Type/Dopant: P/B, Orientation: <100 > ± 0.5°, resistivity: 1–10 Ωcm, thickness: 675 ± 25 μm, single side polished) and glass slides (Menzel-Gläser, white hydrolytic class 1 glass, thickness: 0.15 mm) were used as substrates. Polycarbonate (PC, Makrolon LED2247), ultrapure water (VWR Chemicals, ≤ 1µS/cm), ethanol, acetone and tetrahydrofurane (THF) all in p.A. quality, as well as RCA1 solution (H2O (VE) : H2O2 (35%) : NH3 (25%) = 5 : 1 : 1) were used.

2.2 Sputter deposition of gradient films

Si wafer and glass substrates were used as substrates for the metal film deposition. The wafers were ultrasonically cleaned first in ethanol and then in acetone for 10 min each at room temperature and rinsed with ultrapure water. The glass substrates were cleaned in RCA1 solution at 70 °C for 1 h, then rinsed with ultrapure water and ethanol.

An RF-magnetron co-sputtering setup (Angstrom Engineering) composed of two circular planar magnetrons with 2″ diameter targets was used to deposit the films. The Fe target had a thickness of 0.25 mm (Evochem Advanced Materials, 99.95%) and the Cr target a thickness of 6.35 mm (Evochem Advanced Materials, 99.95%). The operating pressure was 4 ∙ 10−3 Torr and a gas flow of 30 sccm Ar was used. Fe, Fe/Cr and Cr films of approximately 8 nm thickness each were deposited according to the parameters in Table 1. In order to obtain Fe/Cr films with different Fe:Cr ratios, the corresponding substrates were mounted onto the sample holder in different positions (see Fig. 1).

Table 1 Deposition parameters for Fe, Cr and FeCr gradient films by means of magnetron co-sputtering
Full size table

2.3 PC application

Polycarbonate (PC; Macrolon LED2247) was dried for 2 h at 120 °C and then melted onto the alloy surfaces in N2 atmosphere. Two different application conditions were used. For one set of samples PC was melted onto the alloy film just for 10 min at 300 °C. For the other set of samples, the annealing time was increased to 60 min and the temperature was kept at 320 °C, which simulates the conditions during PC processing.

Afterwards excess PC was removed from the alloy films by immersion of the samples in stirred THF at room temperature for 2 h. The samples were then ultrasonically cleaned in ethanol first, then acetone and finally rinsed with ultrapure water. The remaining PC residue was analyzed.

2.4 Confocal fluorescence microscopy

For the fluorescence microscopy, glass substrates coated with metal films, an additional glass substrate, and thicker glass spacers were arranged according to the schema shown in Fig. 1. Dried PC was melted onto the substrate at 320 °C for 20 h in N2 atmosphere. The samples were cooled to room temperature before the measurements were performed.

A confocal Raman microscope (InVia, Renishaw) with a 50 Lx objective and a 532 nm Laser (50 mW) was used. The laser intensity was set to 5 mW (10%) during the measurements. 40 accumulations and an exposure time of 20 s were used. Measurements were taken at different in-depth positions between the metal/PC interface and into the PC bulk. The intensity of the fluorescence signal at one position was determined by the integrated intensity measured in the range of 1750–2250 cm−1 using 391 channels. The signal was then normalized to the measured bulk intensity (measurement in the bulk region of the polymer) of each profile [20].

Fig. 1
figure 1

(a) Schematic of the spatial arrangement of magnetron and samples for the deposition of Fe/Cr gradient films. The numbers at the bottom show the Fe:Cr ratios as function of substrate position for the case of 20 W applied to the Fe target and 23 W applied to the Cr target (see Table 1 for details). (b) Sample set-up for the fluorescence microscopy measurements shown in Fig. 3. A depth of 0 μm corresponds to a fluorescence microscopy measurement performed at the FeCr/PC interface

Full size image

2.5 XPS characterization

An ESCA UHV system (ScientaOmicron) with monochromatized Al Kα irradiation (1486.7 eV) was used. If measured normal to the surface, the X-ray spot area in the experimental setup is 0.75 mm in diameter. The background pressure during measurements was ca. 6 ∙ 10−10 mbar. The spectra were measured at a 30° take off angle relative to the surface. For the survey spectra a pass energy of 100 eV was used, while the core level spectra were measured with a pass energy of 20 eV and a step size 0.1 eV. Under these conditions, the Au4f7/2 peak of a clean Au sample was found at a binding energy of 83.8 eV and had a full width at half maximum (FWHM) of 0.9 eV.

No neutralization procedure was used during the measurement. The as-deposited metal films showed no charging and the spectra are shown without any correction to the binding energy (BE) scale (the C1s peak corresponding to adventitious carbon appeared in those cases at 284.8 ± 0.2 eV [37]). For the samples covered with PC residue some charging was observed depending on the thickness of the residue; in those cases, the C–H component of the C1s peak of PC at 284.5 eV was used to adjust the binding energy (BE) scale [38].

CASA XPS software was used for the analysis. For the analysis of the C1s and O1s core level spectra, a Shirley background and a Gaussian-Lorentzian product line shape with 30% Lorentzian character for peak fitting were used. The decomposition of the Cr2p3/2 and Fe2p3/2 peaks was done according to the procedure described by Biesinger et al. [39].

2.6 FT-IR spectroscopy under grazing incidence (FT-IRRAS)

A Bruker Vertex 70 was used to analyze the PC residue on the metal films. The measurements were performed in dry air atmosphere at room temperature with un-polarized light, an energy resolution of 4 cm−1, and an aperture of 0.5 mm. The incidence angle was 70° to the surface normal. 256 scans were measured for each spectrum to achieve a high signal to noise ratio.

OPUS software was used for the analysis. The signals corresponding to water and CO2 from the air were subtracted from the absorbance spectra. For this procedure, first a spectrum was measured on a sample in air. Then the same sample was measured again in dry air and by division of these two measurements a spectrum of the atmospheric components alone was obtained. This atmospheric spectrum was converted to absorbance, providing a measurement of water and CO2 lines that was used to subtract these signals from the absorbance spectra of the different measurements, after multiplication with the appropriate factors.

For each sample the corresponding metal film without contact to PC was used as a reference.

2.7 AFM studies

The measurements were performed on a Nanowizard II (JPK Instruments) in Intermittent Contact (Air) modus with a Line rate of 1.0 Hz. NSC15 cantilevers (MikroMasch, 40 N/m) were used. The topography of each sample was measured in at least 4 different positions.

2.8 Ellipsometric imaging

A nulling ellipsometer (Nanofilm EP3, Accurion) was used. The spectroscopic measurements were performed in the range of 363.7 to 995.9 nm with an angle of incidence of 55°.

An optical model consisting of effective substrate and polycarbonate layers was used in order to determine the thickness of the PC residue. The dispersion function for PC was obtained from databases, and it is given in the supporting information (figure S7). The effective substrate for each case is composed of the thin (8 nm) metal film deposited onto Si. The ∆ and Ψ values of the corresponding thin metal film on the Si substrate were measured first (prior to contact to the PC melt), and the n and k dispersions of the effective substrate were calculated from this measurement. These values of n and k are then used in order to simulate the optical model used for the determination of PC thickness, in a measurement done after the substrate has been exposed to the PC melt.

3 Results and discussion

3.1 Characterization of gradient films before exposure to PC

The XPS data of the as-deposited films are summarized in Table 2 for three different Fe:Cr ratios as well as pure iron and chromium films. The corresponding survey spectra are shown in the supporting information (figure S1).

The determination of sample stoichiometry by XPS in samples containing transition metals can be subject to large errors, depending on the nature of the background used. In this work, we will base the discussion of our results on the stoichiometry obtained from evaluation of the areas of the Cr2p3/2 and Fe2p3/2 peaks done following the approach of Biesinger et al. [38], as stated above. In this procedure, only the 3/2 component of the spin-orbit doublet is considered, together with a Shirley background. The results of these analysis are shown in Table 2. The three films with a mixed composition of Fe and Cr showed Fe:Cr ratios of 1:0.5, 1:1.2 and 1:3.5, thus providing in one single deposition three well differentiated samples in terms of Fe:Cr composition.

However, it is relevant to note that the stoichiometry calculation done in this way can seriously underestimate the absolute amount of Fe and Cr present at the surface. In order to estimate the range of possible error, we calculated the stoichiometry using a different approach, using the low resolution survey spectra. In this case, a Shirley background was used for the C1s and O1s peaks, while a U2 Tougaard background was used for the evaluation of the complete Fe2p and Cr2p peaks [40, 41]. Due to the shape of the Tougaard background this method yields higher metal concentrations, as you can see in the lower part of Table 2. The results of the two methods can be seen as an estimation of the upper and lower limit of the actual metal concentrations. When comparing the Fe:Cr ratios obtained by this second method we get values of 1:0.6, 1:1.7 and 1:2.5. Although differing slightly from the ratios obtained by the first analysis method, the trend is clear, and the three samples can in any case be considered as good examples of a Fe-rich surface, a Fe-poor surface, and a surface with similar Fe:Cr concentrations. For the rest of the paper wherever a quantification is mentioned, it will refer to the analysis obtained from the first method.

Table 2 Overview of the composition of Fe, Cr and FeCr gradient films in at-%. For evaluation of the survey spectra, a U2 Tougaard background for Fe2p and Cr2p and a Shirley Background for O1s and C1s was used
Full size table

The core levels of Fe2p3/2, Cr2p3/2 and O1s are shown in Fig. 2. The corresponding C1s core levels are typical for adventitious carbon, and they are shown in the supplementary information (see figure S2).

The O1s peaks are decomposed into three components at 530.2 ± 0.2 eV, 531.7 ± 0.2 eV and 534.1 ± 0.1 eV, corresponding to metal oxides, hydroxides and water respectively [39, 42].

The Fe2p peak has been fitted according to Biesinger et al. [39]: Metallic Fe0 appears at 707.2 eV ± 0.9 eV. Fe2O3 components are modelled with a series of peaks. The one with the lowest binding energy is situated at 709.5 ± 0.9 eV and additional peaks appear at + 1.0 eV, + 1.8 eV, + 2.9 eV, + 4.0 eV and + 9.6 eV in relation to that first peak. Similarly, a FeO component is fitted at 708.3 ± 0.2 eV and additional peaks at + 1.3 eV, + 2.5 eV, + 3.7 eV, and + 7.0 eV. A FeOOH component appears with a peak at 710.2 ± 0.1 eV and additional peaks at + 1.0 eV, + 1.9 eV, + 3.0 eV, + 4.1 eV and + 9.4 eV. Interestingly, the contribution of metallic Fe0 in comparison to FeII, III in the surface near region increases with increasing Cr-concentrations. The contributions of FeIII oxides and hydroxides as well as FeII oxides show no correlation with the Cr-concentration of the surfaces. However, in each surface the contribution of FeIII species is much higher than that of FeII.

The Cr2p peaks are decomposed into three components, which were also fitted according to Biesinger et al. [39]: The first one is Cr0 which appears at 574.2 ± 0.05 eV. The second component can be found at 577.3 ± 0.05 eV and is assigned to CrIII hydroxides. Another component is assigned to CrIII oxide, consisting of a first peak at 575.8 ± 0.05 eV and additional peaks at + 1.5 eV, + 2.5 eV, + 3.3 eV and + 4.3 eV. The composition of the Cr2p spectra does not show clear trends correlating to the overall Cr-concentration of the surfaces. However, the main component in all spectra are CrIII oxides, which could inhibit the oxidation of the metal surface. These findings, in addition to the increase of Fe0 in relation to FeII,III with higher Cr contents in the surface, are in line with the observed increased corrosion resistance of chromium-rich steels [43, 44].

The calculated contributions of the different oxidation states are summarized in Table 3.

Fig. 2
figure 2

XPS core levels of the as-deposited FeCr gradient films: (a) Fe, (b) Fe:Cr = 1 : 0.5, (c) Fe:Cr = 1 : 1.2, (d) Fe:Cr = 1 : 3.5, (e) Cr. In the Cr2p and Fe2p spectra the individual peaks of the Biesinger-fits were added up and only the envelope is shown. The spectra were normalized and shifted vertically for better visualization

Full size image
Table 3 Detailed decomposition of the Fe2p and Cr2p XPS peaks of the different films in % of the total concentration of M (M = Fe + Cr)
Full size table

3.2 Studies of interfacial PC degradation

Confocal fluorescence microscopy was performed in the PC interfacial region down to 300 μm (see Fig. 2) in order to evaluate qualitatively the amount of PC degradation products. Figure 3 shows the intensity of the fluorescence signal within the PC as function of the distance to the metal interface. The fluoresce intensity is correlated to the degree of PC degradation. The figure shows clearly that the intensity (i.e. degradation degree) increases with the percentage of Fe at the alloy surface. Moreover, the confocal measurements clearly show that the degradation occurs preferentially at the interface with the alloy. Degradation products are then transported into the bulk of the melt. It should also be noted that metal ions can diffuse into the PC bulk and that Fe ions diffuse at a higher rate than Cr ions [45,46,47]. A degradation of the polymer in the bulk of the melt is therefore also possible, but most likely only plays a minor role.

No interfacial degradation could be observed for the pure Cr-film. Moreover, even for the smallest Cr-concentration achieved in this work (Fe:Cr = 1:0.5) a significant inhibition of the degradation process could be observed. The inhibition was in this case larger than expected based on a pure effect of the surface concentration of Fe-atoms, which hints at a limitation of the Fe-atom release by the percolated Cr-oxide [48].

Fig. 3
figure 3

Fluorescence Microscopy of PC on Fe, Cr and FeCr gradient films. The intensity is normalized to bulk intensity

Full size image

Ellipsometric and AFM measurements were performed to characterize the film thickness and morphology of the insoluble residue on the respective alloy surface. Table 4 shows the thickness of the residual PC film as function of the alloy composition. In all cases, the residual film thickness (between 3 and 16 nm) is much smaller than the depth at which fluorescence is observed within the PC (down to 300 μm). The formation of insoluble polymeric products is limited to a thickness of about 16 nm even for the pure iron film after thermal degradation of 1 h. After 1 h of melt contact the interfacial residue clearly showed a dependence on the interfacial composition of the alloy. However, even in the case of the pure Cr-film for which no increase in fluorescence was detected, an insoluble interfacial layer with a thickness of 4.4 nm was observed. Additionally, the ellipsometric imaging showed that the PC residue layer formation was homogeneous on a mesoscopic scale, as shown in Fig. 4.

Table 4 Thickness of the PC residue on Fe, Cr and FeCr gradient films as determined by ellipsometry
Full size table
Fig. 4
figure 4

Thickness maps obtained from the ellipsometry characterization of insoluble PC residue films on pure Fe and pure Cr after 60 min of melt contact at 320 °C. The thick blue line marks the outer perimeter of the PC residue on the sample

Full size image
Fig. 5
figure 5

AFM measurements of insoluble PC residue films on pure Fe and pure Cr after 60 min of melt contact at 320 °C

Full size image

Figure 5 shows the AFM characterization of the PC residue on pure Fe and pure Cr. PC crystallites are clearly visible on both surfaces [49, 50]. Overall, the microscopic analysis showed that although the residual film was thicker in the case of iron, no significant morphological differences were observed. The ellipsometry and AFM data for other Cr/Fe film compositions can be seen in the supporting information in figures S6 and S8.

Figure 6 shows FT-IRRAS measurements of the insoluble PC residues. The absorbance for all characteristic groups increases with increased time of melt/surface contact. Thicker films form when the surface concentration of iron increases (in agreement with the ellipsometry analysis). However, no clear trend in the relative intensity of the peaks was observed. The peak assignments were done according to the literature considering PC films [11, 20]. The FTIRRAS spectra are shown in Figure S9 of the supporting information.

Fig. 6
figure 6

Area of the peaks of the FT-IRRAS spectra for Fe, Cr and FeCr gradient films after 10 min of melt contact at 300 °C and after 60 min of melt contact at 320 °C

Full size image

The chemical analysis of the interface region of the PC-residue was done with XPS. The complete elemental analysis for all samples is shown in Table 5. Note, however, that in this analysis the attenuation of the metal core level peaks due to the presence of a PC overlayer has not been taken into consideration. The modelling of the exact metal concentrations under an overlayer has not been attempted, because the PC residue is not homogeneously distributed (see Figs. 4 and 5).

There are, however, clear differences in the metal:oxide ratios, as can be seen in the Fe2p and Cr2p peak shapes (data shown in the supporting information, figure S5). For example, the relative intensity of the Fe0 peak (ca. 707 eV) [39] decreases strongly after the metal surfaces have been in contact with molten PC. Likely this is indicative of the oxidation of Fe due to heat and/or the contact to PC as well as humidity. In the Cr2p spectra the metallic component (ca. 574 eV) [39] also decreases in the Cr/Fe gradient films after contact with the PC melt. However, for the pure Cr films, there is an increase in the Cr0 component after contact to PC. Sonnenberg et al. [45] stated that it is possible for metal ions to migrate into a polymer melt, although it should be stated that their method of applying the PC melt differed from the one used in our approach. It is suspected that for the pure Cr films a similar transport of Cr ions into the PC occurs, although they do not seem to affect the degradation process of the polymer.

Table 5 Overview of the composition of the measured PC residue on Fe, Cr and FeCr gradient films in at-%
Full size table

In Fig. 7 we show the C1s and O1s peaks corresponding to the PC residue. In accordance with the FTIR data, the XPS analysis of the PC shows no significant differences in the composition of the insoluble surface as function of the alloy composition. The C1s peaks are decomposed into five components: The first one is fixed to 284.5 eV and assigned to aromatic CH. Further peaks can be found at 285 eV, 286.2 ± 0.04 eV and 290.4 ± 0.04 eV and are assigned to aliphatic CH, C–O and C–C(=O)–C respectively [51]. The last peak appears at 291.1 ± 0.1 eV for the samples that were heated to 300 °C for 10 min and at 291.3 ± 0.07 eV for the ones that were heated to 320 °C for 60 min. This peak is described as a shake up signal. Overall, these peaks are typical for polycarbonate [38].

The O1s data consist of three components, which are located at 530.2 ± 0.3 eV, 532.1 ± 0.2 eV and 534 ± 0.1 eV. They are assigned respectively to metal oxides, C=O and C–O species [39, 52]. The relative fractions of each component as function of the substrate composition are shown in Table 6. The increase in C–O/C=O contribution and the associated decrease in metal oxides mainly reflect the increased thickness of the residue layer after the extended melt contact.

Fig. 7
figure 7

High resolution XPS spectra of PC residue on gradient films: (a) Fe, (b) Fe:Cr = 1 : 0.5, (c) Fe:Cr = 1 : 1.2, (d) Fe:Cr = 1 : 3.5, (e) Cr. The spectra were normalized and shifted vertically for better visualization. The inserts show the show the area of 294–288 eV with a magnification of 5

Full size image
Table 6 Detailed composition of the O1s XPS peak of the measured PC residue on Fe, Cr and FeCr gradient films in at-% of the total peak area
Full size table

4 Conclusion

The analysis of the interfacial formation of sparingly soluble PC residue films during melt contact for Fe/Cr alloy films shows that the degradation process (as analyzed by fluorescence microscopy) is strongly inhibited by the formation of interfacial Cr-rich oxide films. The ellipsometric and AFM data show that the interfacial deposit formation is kinetically inhibited by the Cr-rich passive film as well. However, a strong reduction in the residual PC film thickness was observed solely in pure chromium films. XPS and FT-IRRAS show that the changes in the chemical composition of formed residue films are rather minor and mainly confirm the ellipsometric data on the thickness of the formed layers. The combinatorial approach based on magnetron co-sputtering helped to reveal the influence of the Cr-concentration in the passive film on the resulting interfacial reactivity. It can be followed from the complementary data, that even minor surface concentrations of iron promote the formation of sparingly soluble PC deposits on the alloy surface in contact with the melt, most probably based on Fe-Polymer complexes. However, the increase of the fluorescence during melt/alloy contact can be strongly inhibited by Cr-rich passive films.