**4.3 Performance of diamond coatings on steel substrates**

A set of AISI P20 modified steel molding inserts was prepared and pre-coated with a 2 μm thick PVD chromium nitride (CrN) film, in order to block the mutual diffusion between the ferrous substrate and the diamond growth atmosphere. All the steel plates had the molding surface polished with silicon carbide paper till grit #2000. The samples to be diamond coated were ultrasonic abraded with a diamond solution and then cleaned with isopropyl alcohol. Diamond growth was performed in a hot-filament CVD reactor, using timemodulated CVD process.

In a first experiment, four different samples were prepared to evaluate their performance has molding surfaces. Sample AC1 was submitted to 4h30 of diamond deposition and sample AC2 to 9h00. The full deposition conditions used in steel substrates can be seen in Neto et al. (2008b). Sample AC3 had only the CrN film an AC4 was a bare steel substrate. The two last samples were used in order to compare their performance with the diamond coated inserts.

Figure 7 shows the SEM images of inserts AC1, AC2, AC3 and AC4 before injection molding. Both diamond coated samples exhibited diamond crystallites mainly displaying (111) crystal orientation, although (100) oriented crystals were also observed. These growth directions are typical for the processing temperatures employed in this investigation. The average crystal size of these films is 1.74 and 1.25 μm, and the measured average roughness is 0.18 and 0.16 μm, for samples AC1 and AC2, respectively. Inserts AC3 and AC4 displayed an average roughness of 0.10 and 0.11 μm, respectively. Raman spectroscopy was used to assess the diamond Raman quality, on the diamond coated samples, as proposed by Kulisch et al. (1996) and to estimate the residual stresses of the diamond film according to Ralchenko et al. (1995). Calculated quality factor values for the diamond coatings were 56.0 and 58.3%, for samples AC1 and AC2, respectively. Samples AC1 and AC2 presented diamond peak shifts (∆φ) of 11 and 13 cm−1, respectively. Calculated residual stress (σ) values for the diamond coatings were 6.2 and 7.4 GPa, for samples AC1 and AC2, respectively.

After this preliminary analysis using the molding inserts, they were placed in a mould tool specially designed to accommodate the 10 x 10 x 3 mm inserts and mounted in an injection molding machine to perform a cycle of 500 high-density polyethylene (HDPE) sample plates.

The molded samples were analyzed using optical microscopy. Figure 8 displays 150 times magnified images of HDPE objects from the run number 1, 100, 300 and 500. Whatever the type of insert surface, the first injected object presents more heterogeneous surface than the objects injected in cycles 100, 300 or 500. After the first set of injections, the polymeric objects molded with both diamond coating are very alike. The molded pieces by AC3 and AC4, present some surface scratches, but maintain the optical brightness (observed by the naked eye). The samples molded with the diamond coated inserts presented a slightly more tarnished surface than the samples molded with bare steel or with the CrN coating.

This may be due to the slight increased roughness that the diamond coated inserts presented, compared to the non-diamond coated samples, and also due to the crystalline nature of the diamond coatings. It is also worth mentioning that the injected samples 100 by the bare steel insert was very clean, but as the number of injections are increased, the surface of the plastic sample became heterogeneous.

Microinjection Molding of Enhanced Thermoplastics 233

Since it was found out that the samples molded with the diamond coated inserts presented a slightly more tarnished surface than the samples molded with bare steel or with CrN coating and it is expected that if the average roughness and crystal size of the diamond film is reduced, this problem may be overcome, a new group of samples were prepared. Diamond average roughness and crystal size reduction may be accomplished by the deposition of sub-microcrystalline or nanocrystalline diamond. In order to obtain homogeneous coatings with an average crystal size of about 1 μm or less, different deposition condition where used. The deposition conditions can be seen in Neto et al. (2008c). Three different films were produced. All the as-grown films exhibited sub-micron diamond crystallite size, mainly displaying (111) crystal orientation. The average diamond crystallite sizes of the deposited films were 0.61, 0.71 and 0.83 µm, for sample FD1, FD2 and FD3, respectively. The three coated samples and a bare steel plate (sample F1) was also used

Fig. 9. Optical microscopy images of the HDPE molded surfaces, by the different inserts, in

in the adapted mould, to serve as a reference sample.

run numbers 1, 50 and 80

Fig. 7. SEM images of inserts AC1, AC2, AC3 and AC4 before injection molding

Fig. 8. Optical microscopy images of the HDPE molded surfaces

Fig. 7. SEM images of inserts AC1, AC2, AC3 and AC4 before injection molding

Fig. 8. Optical microscopy images of the HDPE molded surfaces

Since it was found out that the samples molded with the diamond coated inserts presented a slightly more tarnished surface than the samples molded with bare steel or with CrN coating and it is expected that if the average roughness and crystal size of the diamond film is reduced, this problem may be overcome, a new group of samples were prepared. Diamond average roughness and crystal size reduction may be accomplished by the deposition of sub-microcrystalline or nanocrystalline diamond. In order to obtain homogeneous coatings with an average crystal size of about 1 μm or less, different deposition condition where used. The deposition conditions can be seen in Neto et al. (2008c). Three different films were produced. All the as-grown films exhibited sub-micron diamond crystallite size, mainly displaying (111) crystal orientation. The average diamond crystallite sizes of the deposited films were 0.61, 0.71 and 0.83 µm, for sample FD1, FD2 and FD3, respectively. The three coated samples and a bare steel plate (sample F1) was also used in the adapted mould, to serve as a reference sample.

Fig. 9. Optical microscopy images of the HDPE molded surfaces, by the different inserts, in run numbers 1, 50 and 80

Microinjection Molding of Enhanced Thermoplastics 235

the polymeric produced parts and also the degradation of the coated inserts. All coated samples presented good stability at least till 500 runs (the maximum that a single diamond coated insert was subjected to, under laboratory conditions). The HDPE thermoplastic molded objects presented good quality and reproduced well the molding surface. Microcrystalline diamond coated inserts produced slightly tarnished plastic parts. The latter, seems to be considerably dimmed with the use of sub-microcrystalline or nanocrystalline films. The diamond coated featured inserts presented a good performance

and a reduced degradation trend comparatively to the non-coated surfaces.

Fig. 10. Nanocrystalline diamond deposited on Si and SiC structured substrates

Injection molding enables the large scale production of polymeric components with accuracy. This technology has been progressively used for the production of microcomponents in quantity and quality at low cost, which supports the development of microelectro-mechanical systems. Nevertheless, the dimensional reduction of the components requires a higher control of the dimensional accuracy of these devices. It is also known that in the molding blocks of this type of objects, the wear is amplified due to the fact that the surface roughness is dimensionally very close to the dimensions being controlled. Reinforcing materials such as glass or metallic fibers, or carbon nanotubes in the polymeric matrix enhances the wear capability of the material being injected, compromising the service life time of the tools. Reinforced thermoplastics, the so called nano-composites, where developed in order to test the tools and the process, using carbon nanotubes as enhancers. A gateway to reduce the deterioration of the molding impressions, increase their durability and reduce the need of corrective intervention on the tool may be by the use of appropriate surface engineering processes. The use of nanocrystalline diamond or other allotropic carbon coatings is considered a potential surface engineering coating for this type of application, since it detains high hardness and high thermal conductivity, being both properties very interesting to apply to the thermoplastic injection molding process. The

**5. Conclusion** 

Figure 9 displays micrographs of the polymeric molded surface by the different inserts, in run number 1, 50 and 80. Apart from the first set of samples, that present marks of the demolding spray used in the beginning of the processing work, the injected parts are identical. All samples presented a good finishing surface, not showing the tarnished surface that the microcrystalline diamond coated insert originated. From the images, it is evident that the molded objects with the diamond-coated inserts, present more homogeneous surfaces than the ones molded by the insert without coating. It should be noted that insert F1 has the same or better surface finish than the samples that were used to deposit diamond, because their surface was not diamond abraded as the ones pre-treated for diamond coating. The initial steel samples were not surface polished to achieve a mirror surface, in order to obtain optical smooth surfaces and, hence, good quality plastic components/parts. Therefore, the results indicate that polishing time may be saved when using diamondcoated surfaces.

For the evaluation of the degradation of the molding surface, the injected parts by the different systems were analyzed. These parts had simple features which were measured throughout the injection process. Samples molded by diamond coated inserts showed a degradation trend of 0.0001 mm/injection, but the samples molded by samples with only the CrN film or without film presented a degradation trend of 0.0004 mm/injection. This may be due to the polymer aggregation to the cavity.

#### **4.4 Performance of diamond coatings on steel substrates**

Silicon (Si) and silicon carbide (SiC) samples were prepared to be diamond coated and then tested to be used as molding blocks. The diamond growth was performed using conditions to achieve both microcrystalline and nanocrystalline diamond morphology. Microcrystalline conditions can be seen in Neto et al. (2008c). Nanocrystalline was achieved by adding Argon gas to the reactor chamber. The full deposition conditions for the latter can be seen in Neto et al. (2011).

Figure 10 presents the SEM images of the as coated diamond films on Si and SiC with the nanocrystalline conditions. Si (a) is a low magnification image showing the rip and the nearby surface of the featured Si substrate. Si (b) is a high magnification image from the top surface and Si (c) from inside a feature. SiC (a) is a 600 times magnified image of one of the SiC samples, and SiC (b) and SiC (c) are also from the top surface from inside one of the rips, respectively. All films are almost fully coalescent and are composed of nanocrystalline diamond particles with a size inferior to 100 nm. The roughness assessment by means of a profilometer lead to an average surface roughness of 0.12 and 0.19 µm for the deposited diamond films on Si and SiC, respectively.

From the coated samples, 100 of HDPE parts were molded. It seems to replicate the molding surface in a proper way, displaying a fine shining surface.

The Si samples coated with microcrystalline diamond presented similar results as the ones presented by the steel coated substrates.

Different coated systems were tested to reproduce high-density polyethylene (HDPE) components, namely microcrystalline, sub-microcrystalline and nanocrystalline diamond films. Each coated systems were tested for a number of injection cycles in order to evaluate

Figure 9 displays micrographs of the polymeric molded surface by the different inserts, in run number 1, 50 and 80. Apart from the first set of samples, that present marks of the demolding spray used in the beginning of the processing work, the injected parts are identical. All samples presented a good finishing surface, not showing the tarnished surface that the microcrystalline diamond coated insert originated. From the images, it is evident that the molded objects with the diamond-coated inserts, present more homogeneous surfaces than the ones molded by the insert without coating. It should be noted that insert F1 has the same or better surface finish than the samples that were used to deposit diamond, because their surface was not diamond abraded as the ones pre-treated for diamond coating. The initial steel samples were not surface polished to achieve a mirror surface, in order to obtain optical smooth surfaces and, hence, good quality plastic components/parts. Therefore, the results indicate that polishing time may be saved when using diamond-

For the evaluation of the degradation of the molding surface, the injected parts by the different systems were analyzed. These parts had simple features which were measured throughout the injection process. Samples molded by diamond coated inserts showed a degradation trend of 0.0001 mm/injection, but the samples molded by samples with only the CrN film or without film presented a degradation trend of 0.0004 mm/injection. This

Silicon (Si) and silicon carbide (SiC) samples were prepared to be diamond coated and then tested to be used as molding blocks. The diamond growth was performed using conditions to achieve both microcrystalline and nanocrystalline diamond morphology. Microcrystalline conditions can be seen in Neto et al. (2008c). Nanocrystalline was achieved by adding Argon gas to the reactor chamber. The full deposition conditions for the latter can be seen in Neto

Figure 10 presents the SEM images of the as coated diamond films on Si and SiC with the nanocrystalline conditions. Si (a) is a low magnification image showing the rip and the nearby surface of the featured Si substrate. Si (b) is a high magnification image from the top surface and Si (c) from inside a feature. SiC (a) is a 600 times magnified image of one of the SiC samples, and SiC (b) and SiC (c) are also from the top surface from inside one of the rips, respectively. All films are almost fully coalescent and are composed of nanocrystalline diamond particles with a size inferior to 100 nm. The roughness assessment by means of a profilometer lead to an average surface roughness of 0.12 and 0.19 µm for the deposited

From the coated samples, 100 of HDPE parts were molded. It seems to replicate the molding

The Si samples coated with microcrystalline diamond presented similar results as the ones

Different coated systems were tested to reproduce high-density polyethylene (HDPE) components, namely microcrystalline, sub-microcrystalline and nanocrystalline diamond films. Each coated systems were tested for a number of injection cycles in order to evaluate

coated surfaces.

et al. (2011).

may be due to the polymer aggregation to the cavity.

diamond films on Si and SiC, respectively.

presented by the steel coated substrates.

surface in a proper way, displaying a fine shining surface.

**4.4 Performance of diamond coatings on steel substrates** 

the polymeric produced parts and also the degradation of the coated inserts. All coated samples presented good stability at least till 500 runs (the maximum that a single diamond coated insert was subjected to, under laboratory conditions). The HDPE thermoplastic molded objects presented good quality and reproduced well the molding surface. Microcrystalline diamond coated inserts produced slightly tarnished plastic parts. The latter, seems to be considerably dimmed with the use of sub-microcrystalline or nanocrystalline films. The diamond coated featured inserts presented a good performance and a reduced degradation trend comparatively to the non-coated surfaces.

Fig. 10. Nanocrystalline diamond deposited on Si and SiC structured substrates
