Effect of Isocyanate Absorption on the Mechanical Properties of Silicone Elastomers in Polyurethane Vacuum Casting.

Polyurethane vacuum casting with silicone molds is a widely used industrial process for the production of prototypes and small batches. Since the silicone casting molds absorb the isocyanate component of the curing PUR casting resin at the cavity surface, the service life of the molds is typically restricted to very few casting cycles. The successive deterioration of the material properties results from the polymerization of the absorbed isocyanate with moisture to polyurea derivatives within the silicone matrix. In this study, we show for the first time the influence of isocyanate absorption on the mechanical properties of silicone elastomers as well as quantitative differences between commercial materials. The changes in mechanical properties were quantified in terms of Shore A hardness, Young's modulus, tensile strength, elongation at break, and complex shear modulus. It was found that the influence of the isocyanate type on the relative property changes of the silicone was significantly greater than that of the silicone used. The results show that, regardless of its hardness, the silicone absorbs considerably less methylene diphenyl diisocyanate (MDI) than hydrogenated MDI, although the latter causes less deterioration of the mechanical properties and achieves a longer mold service life.

Introduction

Vacuum casting is widely used both in noncommercial areas and in industrial applications. Usually, the term refers to the use of silicone casting molds for the production of plastic products for a prototype or small batch production.[1−3] For this purpose, the master model to be replicated is usually produced in a 3D printing process and embedded in a two-part room-temperature vulcanizing (RTV-2) addition-curing cast silicone based on poly(dimethylsiloxane) (PDMS). The prepared casting mold is then filled with a two-component resin in a vacuum chamber, which is then cured in an oven.[4−6] Depending on the application, a variety of different resins can be used for this purpose, but for industrial applications—such as the production of prototype series—polyurethane (PUR) is used most frequently.[7−10] The main reason why the process is hardly ever used for the production of serial products is that all industrially relevant casting materials, that is, synthetic polymers, cause deterioration of the casting molds, causing damage to the silicone mold cavity after only few casting cycles.[11,12] The reason for this is the diffusion of cast resin components into the silicone.[12] Especially in the context of PUR vacuum casting, the chemical mechanisms of silicone deterioration have been studied extensively. Isocyanate, which reacts with polyol in a polyaddition reaction to form PUR, diffuses into the silicone during the course of the reaction. Within the silicone, it reacts with moisture in a two-step reaction first to form a diamine and then polyurea. This leads to the formation of polyurea clusters on the silicone surface.[12,13] This process becomes optically and haptically noticeable and evidently causes the breakdown of the casting molds. Although the processes are well understood at the molecular level, little is known about how they influence macroscopic material properties and how they affect the mold service life. During demolding of the cured PUR product, geometry-dependent stresses act in the silicone, which at some point lead to the rupture of the deteriorated silicone. This is presumably due to the reduced ductility on the one hand and the reduced resistance to tearing on the other hand. Usually, only a vague description of the deterioration of the mechanical properties is reported.

Depending on the required material properties, different polyols and isocyanates are used, whereby the latter have a significant influence on the service life of the mold. In PUR vacuum casting, isocyanates based on 4,4′-methylene diphenyl diisocyanate (MDI) and 4,4′-methylene dicyclohexyl diisocyanate (hydrogenated MDI; H12MDI) are used almost exclusively. Different chemical characteristics of silicone deterioration by these two types of isocyanate have been discussed in earlier studies.[3,5,6,8,12,13] However, nothing is known about the effect of isocyanate absorption on mechanical properties in general, nor about the influence of the materials used. In this work, we thus investigate the effect of both MDI and H12MDI absorption on the characteristic material parameters Shore A hardness, Young’s modulus, tensile strength, elongation at break, and complex shear modulus of two RTV-2 silicone elastomers, similar to those examined in an earlier study.[13] The fracture sites were examined using helium ion microscopy (HIM). On the one hand, these investigations serve to verify previously proposed theories about the chemical mechanisms underlying the mold deterioration. On the other hand, the results and methods presented can contribute to a systematic development of durable materials. Since similar deterioration processes occur in many other application areas of silicone molding with a variety of resin materials, these findings may also be applicable to other research areas as well.

Results and Discussion

Isocyanate Absorption

A total of four material pairings were examined using two RTV-2 silicone elastomers (SE) and two isocyanates/PUR resins. A soft silicone and a hard silicone were examined, which, according to their Shore A hardness, will be referred to as SE13 and SE38, respectively. Both isocyanate components of the PUR resins are mixtures of the respective base monomers (MDI and H12MDI, respectively) with higher molecular weight polyisocyanates (structural formulas are shown in Figure S1 in the Supporting Information). Due to their significantly higher molecular weight/size, the latter are assumed negligible in terms of both absorbed mass and contribution to property changes. The applied isocyanate mixtures will be referred to as MDI-iso and H12MDI-iso, respectively. The reactive mixtures with their corresponding polyol component will be referred to as MDI-resin and H12MDI-resin.

First, quantitative differences between the material pairings with respect to isocyanate absorption were investigated. For this purpose, PUR castings were performed on round silicone samples to provide information about the isocyanate absorption in the actual vacuum casting process. In addition, absorption experiments with the pure isocyanate components of the casting resins, that is, MDI-iso and H12MDI-iso, were also carried out using a similar test setup for better reproducibility and comparability with other studies.

Figure shows the average mass uptake by isocyanate absorption per cast for MDI- and H12MDI-based PUR-resins as well as soft SE13 and hard SE38 silicones. A clear deviation from a linear regression in the absorption curve can be interpreted as damage to the silicone surface. The H12MDI-resin causes, in addition to silicone tears, another type of damage: the deposition of resin residues that cannot be removed without damaging the surface. The extent of this type of damage in this experiment is probably because the resin components were not sufficiently degassed, which is why the diagrams do not allow any direct conclusions to be drawn about the achievable service life of the mold. However, it is clear that the casting resin has a much greater influence on the mass increase than the silicone, which can be deduced from the slope of the linear regression. Although the H12MDI-resin is known (and now confirmed) to allow a longer service life, this experiment clearly shows that, compared to MDI, about 3 times as much H12MDI is absorbed per casting cycle. An important finding from this is that the damage to the silicone depends not only on the pure quantity of the absorbed isocyanate but also on the chemical mechanisms inside the silicone.

Figure 1

Accumulated mass increase per area over the number of casting cycles, measured using the specimen geometry described in the Experimental Section. The initial mass decrease results from the desorption of non-cross-linked sample components. All curves have been shifted so that the first minimum (cast no. 1 or 2) is given the y-value 0. The following approximately linear region was fitted as a common data set by linear regression and the blank circles indicate data points used for the regression. Subsequent deviations from the linear regression indicate surface damage: mass losses are caused by silicone ruptures and mass increases by adherent PUR residues on the sample surface. The inserted photographs show one of the three samples used after the 15th casting cycle.

The differences in iscoyanate absorption are probably mainly due to two factors: first the concentration of monomeric (i.e., difunctional) isocyanate and second the molecular structure of the isocyanate. Regarding the latter, different characteristics of the isocyanate determine the diffusion coefficient and thus the absorption. Among the most important ones are size, shape, and chemical interactions with the medium. Another important factor is that aromatic hydrocarbons (phenyl groups) are generally less flexible than aliphatic ones (cyclohexane groups), whereby the influence of the structure (features such as symmetry, degrees of freedom, and linearity) with respect to the diffusion coefficient is greater than that of the molecular size.[14,15] Thus, a lower diffusion coefficient is expected for aromatic MDI than for the aliphatic H12MDI.

How much isocyanate per cast is absorbed by the silicone further depends on the composition of the isocyanate as well as the duration of exposure. The latter depends on the course of the polyaddition reaction of the PUR and the released heat of reaction. Slower polymerization means more time for the monomeric components to be absorbed by the silicone. More heat of reaction results in a higher diffusion coefficient. One contributing factor to the higher H12MDI absorption is the fact that the H12MDI-resin reacts slower than the MDI-resin, which leads to a longer exposure time per casting cycle.[13]

The deterioration of the silicone becomes noticeable by a progressive white coloration of the surface, due to the polyurea formation, as can be seen in the inserted photographs in Figure . MDI-iso shows a stronger white coloration, as H12MDI-iso is transparent and reacts to form transparent PUR or transparent polyurea. The fact that the material nevertheless turns opaque can be explained by diffuse light scattering at phase boundaries between the PDMS and polyurea.[13] The polyurea polymerization of the difunctional isocyanates with water is shown in Figure S1 in the Supporting Information.

Figure shows some HIM images of fractures on the silicone surface that occur when pieces of the silicone are torn out when removing the cured PUR. It can be seen that the sample surface, which was very smooth in its original condition (as seen in Figure S2 in the Supporting Information), has become much coarser after the 15th casting cycle due to MDI/polyurea incorporation. The mainly smooth fracture surfaces show topological elevations immediately below the sample surface, which can be interpreted as polyurea clusters. The fact that pieces of the silicone are torn out of the surface without shear stress illustrates that the roughening of the surface causes mechanical adhesion between the silicone and PUR in some regions because of microscopic undercuts.

Figure 2

HIM images of fractures at the surface of the samples shown in Figure after the 15th casting: (a) SE38 and MDI-resin and (b) SE38 and H12MDI-resin. The smooth sample surface before the casting experiments is shown in Figure S2 in the Supporting Information for reference.

In order to investigate the interaction between silicone and isocyanate independently of the PUR polymerization, the time-dependent absorption of the pure isocyanate components was investigated using the same experimental setup. The resulting absorption isotherms for all material pairings are shown in Figure .

Figure 3

Isocyanate absorption per area over the duration of exposure (three samples per data point with an average standard deviation of 0.010 mg/mm2).

The results confirm that first the type of isocyanate is decisive for the absorption rate and second that more H12MDI than MDI is absorbed per unit time. The difference between SE38 and SE13 is not significant. It should be noted that a linear absorption isotherm over time is not explained by the classical Fickian diffusion, for which a linear absorption isotherm over the square root of time is to be expected.[16] This diffusion anomaly similar to the so-called case II diffusion results from the polymerization of the diffusing isocyanate. However, this will be covered in an upcoming publication. Since both the absorption in actual casting cycles and the absorption of the pure isocyanate component are known, it is now also possible to assign a corresponding equivalent duration of pure isocyanate exposure to a distinct number of casting cycles. One hour of exposure to H12MDI-iso simulates approximately 3 casting cycles with H12MDI-resin and 1 h of exposure to MDI-iso simulates approximately 10 casting cycles with MDI-resin. The different absorption rates upon exposure to the pure isocyanates are due to the different molecular characteristics of monomeric MDI and H12MDI. However, the different absorption rates during PUR casting cycles are largely explained by different reaction rates.

Due to the isocyanate absorption, the deterioration of the silicone is accompanied by a change of mechanical properties. These were quantified by means of the material parameters Shore A hardness, Young’s modulus, tensile strength, elongation at break, and complex shear modulus. The inhomogeneous distribution of the isocyanate or polyurea in the respective silicone samples poses a particular challenge when measuring these parameters. Since the isocyanate is only absorbed by the exposed surface but does not homogenize over the sample cross-section within the relevant exposure time, no generally valid characteristic values can be determined. These macroscopic characteristics are inherently anisotropic and depend on the geometry-dependent concentration distribution in the specimen. The results given below are therefore only valid within the scope of the respective experiment and are mainly used for the relative comparison of the investigated material pairings.

Hardness

The hardness of the silicone affects the force required to demold the cured product. In general, softer silicones are preferred for delicate structures. Hard silicones, on the other hand, are preferred for large casting molds to avoid cavity deformation under dead weight. In order to examine the influence of isocyanate absorption on the hardness, the same samples as for the absorption measurements were used. The hardness measurement was then performed on the exposed surface (the measurement did not cause any apparent damage to the surface). In principle, the embedded polyurea clusters that form during isocyanate absorption are harder than the surrounding silicone matrix.[13] The aromatic main chain of polyurea indicates a rigid mechanical behavior and a high glass transition temperature (Tg > 140 °C), which is also reflected in the brittle nature of the in vitro synthesized MDI- and H12MDI-based polyurea reported in an earlier publication.[13] Since PDMS has virtually no intermolecular interactions anyway, the presence of MDI and polyurea in the PDMS matrix between the clusters is not expected to have a plasticizing effect on the silicone.

The Shore A hardness measurements summarized in Figure confirm that the hardness of the exposed silicone surface increases with the duration of isocyanate exposure for all material pairings examined. Furthermore, the effect is more pronounced for MDI-iso than for H12MDI-iso, which is also consistent with first-hand experience regarding the service life. The reason for this is probably the more pronounced clustering of the MDI-iso, which has been discussed before.[13] Since clustering is limited to the near-surface region, softer regions below the surface affect the measurement result, so the overall effect appears relatively mild. However, since damage occurs particularly on small, very delicate structures of the mold’s cavity surface, the change in hardness could play a greater role in the mold deterioration than the change in Shore A value would suggest.

Figure 4

Shore A hardness over time of isocyanate exposure.

Complex Shear Modulus

In areas where the cured product’s contour is parallel to the direction of demolding, shear stresses can cause severe deformation of the silicone. To examine the influence of isocyanate absorption on the mechanical behavior of silicone under shear stress, dynamic mechanical analysis (DMA) was used to measure the complex shear modulus of the silicone as a function of temperature and duration of isocyanate exposure. For this purpose, a new sample geometry shown in Figure was developed, which is suitable for the shear-sandwich specimen holder of the DMA. The measured thickness of the cross-linked samples was (3.12 ± 0.16) mm with a measured volume of (313.2 ± 15.2) mm3 and a calculated surface area of 312 mm2, resulting in an average specific surface area of 1.04 mm–1. Using the absorption curves shown in Figure , the duration of isocyanate exposure can be converted to a concentration if required. The samples were completely immersed in isocyanate so that the absorption occurs through the entire sample surface. The total average concentration C̅ can thus be approximated as followswhere MIso is the total amount of isocyanate absorbed per unit area and unit time, t is the exposure time, and As is the specific surface area. All DMA measurements as seen in Figure show a slight increase in the complex shear modulus with increasing temperature, probably due to the entropy elasticity commonly observed in elastomers.[17,18]

Figure 5

Complex shear modulus over temperature measured by DMA in shear-sandwich mode for 0–4 h of isocyanate exposure. Diagrams of the SE38 samples and the SE13 samples have the same scaling to illustrate the relative influence of the isocyanate. The inserted illustration shows the sample geometry.

Figure shows an overview of the DMA results at room temperature. The results confirm the already known (but never before quantified) decrease of elasticity[5] by an approximately linear increase of the complex shear modulus with increasing duration of isocyanate exposure. It is interesting to note that the slope of the linear regression depends essentially on the type of isocyanate and not on the type of silicone. While the MDI-contaminated silicone has hardened significantly, the H12MDI-contaminated silicone shows practically no change, which supports the results of the hardness measurement. This is most likely the explanation for the significantly longer service life achieved by H12MDI-based resins. This effect can also be explained by the formation of the hard polyurea phase.

Figure 6

Complex shear modulus at 24 °C measured by DMA over the duration of isocyanate exposure. The values are based on the results shown in Figure .

Since the chemical mechanisms of deterioration caused by H12MDI-iso have received considerably less attention in previous studies, its relatively minor effects on the mechanical properties of the silicone cannot be satisfactorily explained here. Presumably, this behavior and the associated longer service life are due to the significantly slower polyurea reaction that has previously been reported.[13] In general, H12MDI reacts relatively slowly compared to MDI. Typically, catalysts such as dibutyltin dilaurate are used in H12MDI casting resins to accelerate the PUR reaction. Further inhibition of the MDI-iso reaction to polyurea inside the silicone could be an effective method to reduce deterioration effects.

Tensile Testing

During demolding of the cured PUR product, the silicone can be subject to severe elongation, especially during the demolding of undercuts. The change in the mechanical behavior under tensile stress thus plays a significant role in the deterioration of casting molds. Figure shows all measurement curves of the tensile tests. The average isocyanate concentration in the measurement section of the tensile bars can be approximated with eq using the specific surface area given in Section . In principle, none of the measurements show signs of plastic deformation even after isocyanate exposure. As expected, SE38 shows a significantly higher Young’s modulus in the low elongation region and a lower elongation at break than SE13. The higher elongation at break is accompanied by a higher standard deviation. The color-highlighted curves are averaged from the individual measurements. Although the influence of the isocyanate exposure on the Young’s modulus is generally low, a tendency toward higher moduli can be observed for SE38 and a tendency toward lower moduli for SE13, although the latter is not significant because of high variance. A clear difference between the isocyanates is not evident here. The low influence of the isocyanate exposure on the Young’s modulus can be explained by the fact that the majority of the polyurea is located in clusters, which hardly forms any intermolecular interactions with the PDMS matrix.[13] The elastic deformation of the blend thus continues to occur almost exclusively in the silicone phase.

Figure 7

Stress–strain diagrams determined by the tensile test after 0, 2, and 4 h of isocyanate exposure. The color-highlighted curves are averaged over seven individual measurements and are shown up to the tearing of the first sample. The enlarged diagram regions show linear regressions of the averaged curves up to an elongation of 120%, whose slope is the Young’s modulus E for this region.

A closer look at elongation at break and tensile strength, as shown in Figure , on the other hand, shows a significant decrease in the characteristic values with increasing exposure time. A reduction of tensile strength with increasing exposure time confirms the previously proposed hypothesis that the embedded polyurea clusters promote crack formation and propagation.[13] Probably due to the high standard deviation, however, no clear influence of the isocyanate can be identified.

Figure 8

(a) Elongation at break and (b) true tensile strength (see eq ) determined by tensile tests over the duration of isocyanate exposure. The values are based on the results shown in Figure .

The chemical mechanisms underlying the material deterioration and the inhomogeneous concentration distribution of isocyanate and polyurea within the sample suggest that the correlation between all measured mechanical properties and the duration of isocyanate exposure can only be approximated by linear functions. The actual correlations are much more complex and cannot be interpreted more accurately on the basis of the measured data.

Figure shows the fracture cross-sections of the tensile bars. It can be seen that, in contrast to the reference, the outer regions of the isocyanate-exposed tensile bars are significantly coarser than the regions further away from the surface. While in the MDI-iso exposed sample, only the near-surface areas appear coarse, the H12MDI-iso exposed sample shows clearly visible bulges even in deeper regions, which confirms that the polyurea tends to accumulate on the surface due to the faster MDI reaction.[13] From the fractography of Figure a,b, the hachures can be traced back to the origin of the fracture in the corner of the cross-section. The hachures seem to indicate discontinuous crack growth in mostly homogeneous regions of pure silicone. However, Figure c shows a mostly forced fracture rather than hachures.

Figure 9

HIM images of cross-sections of tensile bars at the fracture site showing the region where the initial tearing originated: (a) SE38 reference, (b) SE38 after 4 h of MDI-iso exposure, and (c) SE38 after 4 h of H12MDI-iso exposure.

Conclusions

The influence of the frequently used MDI- and H12MDI-based isocyanates on the deterioration of the mechanical properties of two platinum-catalyzed RTV-2 silicones in polyurethane vacuum casting was investigated. The characteristic material parameters Shore A hardness (increasing), Young’s modulus (inconclusive), tensile strength (decreasing), elongation at break (decreasing only for MDI-iso), and complex shear modulus (increasing only for MDI-iso) were measured as a function of the duration of isocyanate exposure. It was found that the isocyanate used has a significantly greater influence on the deterioration process than the silicone. Both silicones absorbed more H12MDI both during exposure to pure isocyanate and per casting cycle regardless of their Shore A hardness. Nevertheless, the comparatively lower MDI absorption showed a significantly stronger influence on the mechanical properties of the silicone, thus confirming the reportedly higher mold service life for H12MDI-based resins. No influence of H12MDI-iso on complex shear modulus and elongation at break could be observed. The most probable reason for this is the slower polyurea polymerization within the silicone, which has been addressed in earlier studies. The results demonstrate the validity of previously proposed theories about the underlying chemical mechanisms and their effects on mold deterioration. A better understanding of the critical material parameters and the relevant influencing factors facilitates the systematic development of more durable materials. This will in the long term promote the economic viability of the process and its industrial applicability for series production.

Experimental Section

Materials

To investigate the effect of different technical-grade materials on the deterioration of the silicone, two PUR resins and two silicone elastomers were investigated. The PDMS-based RTV-2 silicones were used as provided and according to the manufacturer’s instructions: Essil 291 (platinum-catalyzed/addition-curing, Shore A hardness = 38, Sika Deutschland GmbH, Bad Urach, Germany) and SF13 (platinum-catalyzed/addition-curing, Shore A hardness = 13, Silikonfabrik, Ahrensburg, Germany), referred to as SE38 and SE13, respectively. The liquid silicones were degassed for 20 min at about 1 mbar before casting to remove air entrapments and subsequently cross-linked at 70 °C. For PUR-casting, two PUR resins, each consisting of polyol and isocyanate, were used: PX226 (Sika Deutschland GmbH, Bad Urach, Germany), which is based on polymeric MDI containing di-, tri-, tetra-, and pentafunctional phenyl isocyanates[19] (viscosity ≈ 250 mPa s), and PRC 1710 (Synthene, Ferme de l’Evêché, France), which is based on H12MDI and contains an unknown amount of PUR prepolymers (viscosity ≈ 320 mPa s). With the exception of SE38, these are the same materials that have been studied previously in terms of their chemical characteristics (SE38 is marginally softer than the previously studied silicone).[13] Chemical formulas of the materials used are shown in Figure S1 in the Supporting Information.

Absorption Measurements

Isocyanate absorption was determined both per PUR resin cast and, as in previous studies, per duration of exposure to pure isocyanate. For the casting experiments with PUR resins, the respective isocyanate and polyol components were first degassed and then mixed in a ratio specified by the manufacturer, and finally, 2 g of the mixed PUR resin was applied to the surface of dried round silicone samples (10 mm thick and 30 mm diameter), as detailed elsewhere.[13] The liquid resin is enclosed by a silicone sealing ring so that only the upper side of the sample is exposed. After complete curing of the PUR in an oven at 70 °C, the solidified PUR is removed and the procedure is repeated 15 times with three samples per material pairing. The absorption of pure isocyanate was examined with the same setup at 70 °C: three samples per exposure time were weighed with an analytical balance AG204 DeltaRange (Mettler Toledo, Gießen, Germany) before and after the isocyanate exposure. It should be noted that in all absorption measurements the samples were stored at 70 °C until their mass stabilized due to the evaporation of volatile components. Although the samples resorbed some moisture during the casting experiments, the deterioration of these samples may not be directly comparable with real casting molds due to the low moisture content.

Characterization Methods

For Shore A hardness measurements (for elastomers) according to ISO 48-5:2018-08, the penetration depth of a cone tip is measured at room temperature with 1 kg weight pressure after 15 s. The hardness value is indicated on a scale from 0 (2.5 mm) to 100 (0 mm) depending on the penetration depth. Five measurements per sample (as described in Section ), 15 in total, were performed with a Shore A hardness tester (Zwick Roell).

DMA was used to measure the complex shear modulus G*, which is the absolute value of a complex number whose real component (storage modulus G′) reflects the elastic portion and whose imaginary component (loss modulus G″) reflects the viscous portion of the deformation

Measurements were performed with a DMA Q800 (TA Instruments, New Castle (DE), USA) using a shear-sandwich specimen holder with a deformation amplitude of 20 μm (0.2% of the sample length) and a frequency of 1 Hz. After a holding time of 10 min at 0 °C, the samples were heated to 100 °C at a constant rate of 2 K/min. The mixed silicone components were poured into milled steel molds of 10 mm × 10 mm × 3 mm for sample preparation. Since the excess silicone is manually scraped off the open side of the mold, measurements are subject to some variance in the specimen geometry. The samples were then completely cross-linked, dried, and fully immersed in isocyanate for the duration of the experiment at 70 °C. Two samples are used for each measurement, and each measurement was repeated two times with different pairs of samples.

The tensile bars according to DIN 53504 (type S1) for elastomers used for the tensile tests are punched out of cast silicone plates. For this purpose, the noncross-linked silicone is completely degassed and homogeneously distributed in a casting frame with a scraper and then fully cross-linked. The bar thickness of (2 ± 0.2) mm required by the standard is difficult to achieve in some cases depending on the viscosity of the noncross-linked silicone. At (2.52 ± 0.35) mm, the tensile bars from the more viscous SE38 are correspondingly thicker than those from the SE13 at (2.15 ± 0.17) mm. These values result in an average specific area (surface per volume) of the middle section of the tensile bars of 1.12 mm–1 for SE38 and 1.25 mm–1 for SE13. The isocyanate exposure is carried out in 500 mm long, V-shaped steel troughs at 70 °C in an oven. The troughs are filled with the isocyanate until the entire measurement length of the tensile bars is immersed. This ensures that the entire test length (without contacting areas with the bottom of the trough) is homogeneously exposed (an exemplary photograph of a SE38 tensile bar after 4 h of exposure to MDI-iso is shown in Figure S3 in the Supporting Information). The test setup allows to use a minimum amount of toxic isocyanates and to dispose of them properly afterward. Seven tensile tests per trial were performed in accordance with DIN 53504 at room temperature with a Zwick Z020 (Zwick Roell, Ulm, Germany) tensile testing machine at a testing speed of 200 mm/min. The tensile force and elongation were measured from the point where the preload of 2 N was reached at a run-up speed of 20 mm/min. Each tensile force measurement was related to the respective individually measured thickness (cross-section) value. Since the tensile bars taper significantly during the tensile test, a distinction is made between the measured tensile strength σm (related to the initial cross-section A0) and the “true” tensile strength σt (related to the tapered cross-section A′ when the specimen tears)

Silicone samples examined for mechanical properties were, comparable to industrial conditions, not dried (stored in ambient conditions at room temperature and 40% RH) and measured within at least 48 h after isocyanate exposure.

For microscopic investigations, the HIM Orion Plus (Carl Zeiss, Jena, Germany) was used at an acceleration voltage of 35.3 kV with a spot control of 6.5 to obtain a beam current of 0.5–0.6 pA. For charging compensation during the secondary electron detection, an electron flood gun was used after each line scan. Since the results of the mechanical measurements indicated that the qualitative property changes depend primarily on the type of isocyanate and not on the type of silicone, only SE38 was investigated by HIM. Micrographs were taken of randomly selected SE38 tensile bars and SE38 samples used in casting experiments.