Properties and Medical Applications of Biocompatible Polyamide in Additive Manufacturing

The paper presents the structural and dimensional differences of rectangular shape samples manufactured by selective laser sintering (SLS) according to three process sinterization energies. In this study we determine the geometrical and structural characteristics of biocompatible PA2200 polyamide manufactured at a low energy density E1, and intermediate value E2 and a high energy density E3. Relaying on these results, four custom-made medical parts have been built using biocompatible polyamide PA2200 powder, in order to evaluate their performances.


Introduction
Combining the geometrical versatility of additive manufacturing (AM) process with biocompatible polymers opens wide possibilities in custom manufacturing of complex shape structures designed for human body. The polymeric materials can be used as bone substitutes while additive technologies allow constructing the appropriate shape and functional characteristics of a biological structure [1,2]. The custom implant and prosthesis are continuously developing due to additive manufacturing technologies which allows constructing mechanical parts that mimics the local anatomy of a patient starting directly form computer tomographic images [3,4]. Beside these the ready to use parts, the AM models successfully serves as disposable tools used in surgery or visualization elements for surgeons preoperative planning [5][6][7][8][9].
Independent on the type of AM technology, the dimensional accuracy [10,11], the mechanical properties [12,13] and the material used [14,15] represents challenging research issues. The inconsistencies in shape, size and mechanical properties are associated with a series of process parameters that are very specific to each technology [16][17][18][19].
The characterization of the products from all engineering perspectives is critical, especially because there is no standardization related with these technologies [20][21][22]. Many authors focus their work on process dependent properties of AM products. The process parameters are in large number and they play a great role in geometrical and mechanical properties particularly on selective laser sintering (SLS) [23][24][25]. Some of these are controllable: laser power, temperatures, position and orientation of the samples, layer thickness, material, scanning strategies; while other are difficult to control and estimate during the process: humidity, powder degradation, temperatures gradients in powder, bulk density, contractions at cooling, electrostatic charging and compaction rate.
The geometric assessment and structure microscopy were used by some authors in order to characterize the parts resulted from AM process. Some authors work on geometrical characterization of samples obtained at different orientations and/or energy density by measuring the size and shape and inspecting the surface and internal cross sections by stereomicroscopy [26,27]. The inconsistency in shape size and properties of SLS polymers caused by the uneven temperature distribution in the powder volume were studied [28]. Other authors put efforts in identification of mechanical properties of poly -amide according to some process variables, in both tensile and bending tests [29][30][31]. Correlations between process parameters and real properties of the parts obtained by mechanical testing (tensile, compression, bending) and also geometric shape and tolerances were also studied [32][33][34]. The consistencies of the results were point out using ANOVA statistics underlining the possibility of adjusting the laser exposure as long as the energy density delivered in the powder layer is optimal [35].
The aim of this study was to determine the geometrical and structural characteristics of biocompatible PA2200 polyamide manufactured by SLS additive process under three different manufacturing conditions: a low energy density E1, and intermediate value E2 and a high energy density E3. Excepting the energy density, the other AM parameters remains constant in order to observe the influence of a single variable. The novelty of the study consists in identification of a manufacturing parameter for PA2200 biocompatible polyamide manufactured by SLS, that will lead to better dimensional stability and internal structure. Based on these findings, four custom-made medical parts were built in order to evaluate their performances.

Materials
The material used in manufacturing samples and medical applications was polyamide PA2200 in powder form. This is a multipurpose material developed by Electro Optical Systems (EOS) [36] and which exhibit high strength and stiffness, good chemical and long-term stability, high detail resolution, good biocompatibility according to EN ISO 10993-1 [37] and food contact approval in compliance with the EU Plastics Directive 2002/72/EC. In the powder form it possesses the following characteristics: average grain size 56 µm according to ISO 13320-11; bulk density of 0.45 g/cm³ according to EN ISO 60; melting point 172-180°C according to EN ISO 11357-1 [38]. The mechanical properties of PA2200 that are highly influenced by the process parameter [39][40][41]

Methods
The AM process was conducted on laser machine EOS Formiga P100 (Electro Optical Systems GmbH, Germany). The machine uses a 30 W CO2 laser for sintering the powder particles and two electrical heating sources to rise up the temperature of the powder in the softening range, in such way that the lase only give the additional energy required for particle bonding. The machine and software preparation consist in virtual positioning of the parts in the building envelope and building envelope setup. In this stage, the process parameters were set to the following values: building chamber temperature 169.5 °C; Removal chamber Temperature 152°C; Layer thickness 0.1 mm; Uniform scaling factors 2.2 %; Beam offset 0.15 mm. The scan spacing (s) was 0.25 mm constant while the laser power (P) and laser beam velocity (v) were set to 24.5, 22.8 and 21.8 W respectively 1500, 2000 and 2500 m/s. The energy densities resulted from these combinations of power, beam velocity and scan spacing were: E1 = 0.065 J/mm 2 ; E2 = 0.045 J/mm 2 ; E3 = 0,035 J/mm 2 (Eq. 1).
The sample geometry of 60 x 22 x 12 mm rectangular section beam and the medical custom parts were designed in SolidWorks 2013. Fifteen rectangular section samples were manufactured for each energy density: E1, E2 and E3.
The investigations consist of geometrical measurements, weighting and microscopic investigations. The measurements were acquired using a digital caliper having a resolution of 0.02 mm. The optical microscopy was conducted on Olympus BX51M microscope at a magnification of 100 and 200X respectively, the acquired images being analyzed using Analysis 5.0 software. Every sample was weighted using 500 g Kerrn balance having a resolution of 0.001 g. Based on mass results (M) and volume (l x b x h), the volumetric mass density (Eq. 3) was computed for each process energy.

Results and discussions
At the end of the manufacturing process, the samples were let to cool down inside the machine for 1 day. The parts extraction and coding were done in order to keep the connection between the part position in the building envelope and their dimensions, weight and structure. The location of each sample was coded as: front, middle or back, where the front positioning was considered the vicinity of the machine's door, while back is the far plane. The post-processing was conducted prior to any investigation and it consists in air blowing for removing the excess non-sintered powder.
The measurements on every direction were conducted 3 times and an average value was saved. The samples were individually weighted and relying on size and mass, the volumetric mass density was computed. In the Figure 1  The highest recorded error was for linear dimension on Z direction of the machine. This behavior is rather common in additive manufacturing since the Z direction is the direction of the layer deposition and part growing. The X and Y errors are more reduced in values due to the symmetry in the sinterization process (X-Y laser traces). From the position point of view, differences in dimensional accuracy are associated with the front-back direction of the machine. It appears that samples that grows in the middle position of the building plane are more stable from dimensional perspective. This phenomenon may be explained by the heat transfer inside the envelope. The parts that are surrounded by a large volume of powder (middle parts) are cooling much slower and for this reason their dimensions are more stable.
The volumetric mass densities of the samples according to the energy used in the sinterization process is depicted in the Figure 2. As it can be observed, the highest energy E1 lead to the highest volumetric mass density of the samples, around 0.85 g/cm 3 while the lowest energy E3 conducts to a much lower value, of about 0.65 g/cm 3 . Taking into consideration that all mechanical properties are directly connected with the structure and therefor with density, the energy providing the higher volumetric mass density is the one that leads to superior mechanical properties. The low standard deviation indicates a good structural stability in all energies that been used. The microscopic images AM polyamide structure were taken at a magnification of 100 and 200X. In the Figure 3 a-f, the structure is presented for all three energies. In images, the sinterization aspect of the structures shows a partial coalescence of the particles no matter the energy value. The bonds between the powder particles instead are in the shape of small bridges proportional in size with the energy value.  Although the porosity cannot be measured, by overlapping color filters on the images, a difference between particles and voids can be objectively measured. In Figure 4, the overlapping color masks are presented. Here, the red color was applied on highly focused particles while the blue color on particles originated in the deeper planes of the structure. A percentual result of partially coalesced particles and voids is presented in the Figure 5. The sum of percentages of particles and voids for one energy is not 100% because of transitional area between sharp focused particles and deep blurred particles (voids) that cannot be associated with any of the two colors. In other words, this can be considered the error and the limitation of this method. Anyway, the presence of bonded particles for E1 is around 20% larger than  In the next phase of the study, based on the dimensional and structural properties determined, four prototypes belonging to the medical field were built. The custom designs were chosen in such a way to cover a large variety of shapes at comparable sizes. In Figure 6 the pats manufactured from PA2200 using E1 energy are presented: a) depicts a fusion plate implant designed for cervical vertebrae fusion; b) lattice structure for bone reconstruction and/or substitution; c) intervertebral cage designed to recreate the anatomical wedge shape of the lumbar intervertebral discs; d) a prototype of short stem one component hip prosthesis.

Conclusions
The paper presents the dimensional accuracy, volumetric mass density and internal structure and of polyamide PA2200 manufactured by selective laser sintering using three energy densities E1, E2 and E3. Based on these findings, four custom made medical parts that were built in order to evaluate their performances.
The dimensional accuracy of laser sintered polyamide is dependent on process energy and parts position on the building environment. High dimensional errors are recorded for vertical direction of the samples comparing to the X-Y directions. The volumetric mass density and structure are directly dependent on energy density used in the sinterization process. Higher process energies will lead to higher volumetric mass density, which as consequence will give better mechanical properties. The laser sintering can become a useful technology in manufacturing the medical components. Due to biocompatible properties of PA2200 and versatility of geometrical shape given by AM, spare parts and disposable surgical guides can be produced.