1 Introduction

Polyether-ether-ketone (PEEK) is a kind of polyaryl-ether-ketone (PAEKs), which is a semi-crystalline thermoplastic material, firstly developed by British scientists in 1978 [1]. PEEK material has excellent high temperature resistance, chemical resistance and wear resistance than other thermoplastic materials. Therefore, PEEK is widely used as a high-performance material. In addition, PEEK material has good biocompatibility and elastic modulus similar to human bone. These characteristics show great potential in the medical field, especially in orthopedics, trauma, spine and other medical implant products [2].

The PEEK material products also exhibit a diverse range of processability. Within the medical field, conventional processing techniques such as machining, injection molding, extrusion, and compression molding enable the production and processing of PEEK products to meet specific requirements [3]. However, these conventional methods have inherent drawbacks including high mold costs, long delivery cycles, and design limitations. In recent years, the rapid advancement of additive manufacturing technology has brought forth PEEK orthopedic implants with intricate biomimetic structures and personalized designs into the public’s field of vision. At the same time, there have been recent reports on PEEK products and their derivatives printed based on three additive manufacturing processes [4,5,6]. These technologies are selective laser sintering (SLS), direct ink writing (DIW), and fused filament fabrication (FFF). The main problem with SLS technology is its relatively high equipment cost, low density of molded parts, inferior mechanical properties, and the difficulty in recycling PEEK powder after printing [7, 8]. This leads to increased material costs and constrains the further application of SLS technology. Direct ink writing technology is restricted in the field of customized medical implants due to low tensile strength and complex post-processing processes [5]. Unlike SLS and DIW, FFF 3D printing technology has the advantages of simple structure, easy operation, and low printing cost. It is currently the most popular additive manufacturing process choice for 3D printed PEEK materials used in production, research and education [9, 10]. The forming principle of FFF 3D printing is to heat and melt the filament, extrude it from the nozzle, and fall on the specified position layer upon layer to form the required three-dimensional shape.

Many existing studies focus on the impact of FFF 3D printing parameters on final product performance. In addition, the effect of the moisture absorption of PEEK filament used for FFF 3D printing on the final product has also aroused great interest. Different storage and drying conditions may affect the mechanical properties of the final product. Radoslaw et al. explored the influence of moisture compensation of acrylonitrile–butadiene–styrene (ABS) filament in the process of FFF 3D printing [11]. Filaments with moisture shown poor interlayer quality after FFF 3D printing, resulting in greater anisotropy. This was because water penetrates into the filament, causing the filament to expand and change the glass transition temperature, resulting in a change in the viscosity of the extruded material and a weak interlayer bonding force. Hamrol et al.'s study also confirmed that the tensile strength of FFF 3D printed products decreased with the increase of the moisture content of ABS filament [12].

Not only ABS materials, Mehmet et al. tested the effect of environmental relative humidity on the mechanical properties of FFF 3D printed poly(lactic-acid) (PLA) components [13]. The results showed that the bending strength and bending modulus of the sample were lower with the increase of ambient relative humidity, no matter how the filling ratio changed. This was mainly because the surface quality of 3D printed PLA components was negatively affected by relative humidity. The higher the moisture content of the sample, the more surface pores there were. Ma et al. 's research also supported the above view [14]. That is, with the increase of moisture content, the bending strength decreased, but the elongation at break increased. On the other hand, storage under dry conditions helped to improve the tensile strength of the part [15]. In addition, Zaldivar et al. also tested the effect of initial moisture content of ULTEM 9085 (polyetherimide, PEI) filament on microstructure and mechanical performances of FFF 3D printed parts [16]. When the filament moisture content exceeded 0.4%, the 3D printed sample produced a highly irregular surface with a large number of gaps. When this value was below 0.1%, the printed sample surface looked almost identical to the completely dry sample. The mechanical properties of the sample were inversely proportional to the moisture content of the filament. Although the modulus and strength were slightly higher than those of the dry sample when the moisture content was below 0.16%, it showed earlier failure.

For high-performance thermoplastic materials, Amedewovo et al. explored the characterization of moisture transport in high-performance carbon fiber-reinforced thermoplastic composites at high temperatures [17]. For carbon fiber (CF)/PEEK and CF/PEKK, the adsorption weight was proportional to time, reaching a maximum value of 0.15 wt% at approximately 10 h. Selzer et al. investigated the effect of water up-take on interlaminar fracture properties of carbon fiber-reinforced polymer composites [18]. The maximum moisture content of CF/PEEK samples was 0.3 wt%. The higher the water temperature, the faster the sample absorbed moisture. The comparison of the fracture area showed that the dry sample broke immediately at the crack. However, the samples after moisture absorption had fiber adhesion at the crack. In addition, the toughness of the sample was analyzed. With the increase of moisture content, the toughness of the CF/PEEK sample decreased by 20% compared with that of the dry sample. Similarly, the study of Michael et al. also confirmed similar views [19]. When the temperature increased from 35 to 95 ℃, the content of moisture in PEEK increased from 0.44 to 0.55 wt%. Finally, adsorption saturation was achieved at 95 ℃.

Similarly, PEEK filament is hygroscopic. Printing with undried filaments may result in different surface or internal qualities. Anouar et al. studied the influence of hot and humid environment on FFF 3D printed PEEK blends [20]. The study showed that the Young’s modulus and tensile strength of samples in hot and humid environment decreased significantly compared with dry environment. This was because after filament printing in a hot and humid environment, the interlayer adhesion was weak and there were some voids. This phenomenon led to delamination between layers and decreased the elastic modulus. Qu et al. studied the influence of FFF 3D printed PEEK material on mechanical properties at different ambient temperatures [21]. When the ambient temperature increased from 60 to 90 ℃, the tensile strength increased by 5.14% and the bending strength increased by 5.32%. The bending strength was increased by 13.86% at room temperature, and the decrease in ambient temperature can also affect the crystallinity of the sample. Wu et al.'s research also indicated the same opinion [22]. When the ambient temperature increased from 90 to 130 ℃, the warp deformation of the printed PEEK sample decreased from 1.9 to 0.65 mm. Therefore, higher ambient temperatures can effectively eliminate moisture in the filament, while reducing temperature differences and enhancing interlayer bonding. Ultimately, the performance of the sample was improved. Andrey et al. studied the hygroscopic process of 12 typical filaments (including PEEK filaments) under different humidity conditions (16–97%) [23]. The moisture absorption behavior of all test materials follows the classic Fick's law. For PEEK filament, hygroscopic behavior occurred when the humidity reaches 47%, and the most significant behavior was about 0.6 wt% when the humidity reached 97%. Under these conditions, the strength of PEEK filament was reduced by about 6.6 MPa. It could be seen that PEEK filament will adsorb a certain amount of moisture over time until it reaches a saturation value. The use of non-dry PEEK filament or PEEK filament in a humid environment after printing for a period of time will significantly affect the surface and internal quality of the final product. Such as surface pores, internal filling dissatisfaction and other bad conditions. This will mainly lead to the decrease of interlayer bonding force.

Therefore, understanding when PEEK filament reaches the moisture content that can easily cause product quality problems is crucial for users to avoid potential printing problems and control product quality. This paper will explore the relationship between the moisture absorption of PEEK filament and its standing time at room temperature. The effect of moisture content of PEEK filament on surface quality and mechanical properties of printed parts will also be verified.

2 Experiment

The material used in this study was Evonik INFINAM PEEK 9359 F polyether-ether-ketone filament (diameter of 1.75 mm). PEEK filaments were fed into the extrusion-based APIUM M220 (Apium Additive Technologies GmbH, Germany) 3D printer for sample printing. This printer has a closed chamber. During the printing process, hot air circulates upward from the build plate, and the filament is stored inside the printer. 3D printing parameters are shown in Table 1.

Table 1 PEEK specimen parameters for FFF 3D printing
Full size table

2.1 Moisture content test

The PEEK filament was placed in the vacuum drying oven (HZK-LC-210, Shanghai Yuejin Medical Equipment Co., LTD, China) to dry, set at a constant temperature of 90 ℃ and a drying time of 72 h to ensure that the filament was completely dried. Weigh the completely dry PEEK filament to m0, then place the filament in an environment of 25 ℃ and 60% RH for i h, and weigh the weight mi (i = 0, 24, 48, 72, 96…) until the weight does not change. The length of the filament for weight measurement is 20 cm, and the average of three filaments is measured at each time node. The formula for calculating the moisture absorption of filament through weight change is:

$$moisture\,content\, = \,{{\left( {m_{0} \, - \,m_{i} } \right)} \mathord{\left/ {\vphantom {{\left( {m_{0} \, - \,m_{i} } \right)} {m_{0} \, \times \,100}}} \right. \kern-0pt} {m_{0} \, \times \,100}}\%$$
(1)

2.2 3D printing specimen for testing

In order to compare the effects under different conditions, three types of samples were printed: extruded materials, tensile testing specimens, and density specimens. Tensile testing specimens and density specimens are shown in Fig. 1, respectively.

Fig. 1
figure 1

Dimension of 3D printed PEEK. a tensile testing specimen and b density specimen

Full size image

2.3 Image analysis

The biological microscope (XSP-BM-2C, Shanghai BM optical instruments manufacture co.,Ltd, China) was used to observe the morphology of extruded materials. The tensile testing specimens were fixed on the sample stage by conductive adhesive, and the morphology of the fracture surfaces of tensile testing specimens were observed by scanning electron microscope (SEM) (Prisma E, Thermo Fisher, America) with an accelerating voltage of 20 kV after ion sputtering coating with gold.

2.4 Tensile testing

The tensile specimen was tested using electronic universal testing machine (UTM5105, Shenzhen Sansi Zongheng, China) with a 100kN load cell at the testing speed of 50 mm/min. The tensile stress was given by the following equation:

$$\sigma \, = \,F/A$$
(2)

where σ is the tensile stress, F is the measured corresponding load, and A is the original cross-sectional area of the specimen. The nominal tensile strain is given by the following equation:

$$\varepsilon \, = \,\Delta L/L$$
(3)

where ΔL is the distance increment between the fixtures, and L is the initial distance between the fixtures (set at "58" mm with a tolerance of + 2 mm, determined by actual measurement). Stress-nominal strain curves are constructed based on the measured values, and tensile strength is the global stress maximum within the stress–strain.

2.5 Density testing

Density specimens were tested according to ISO 1183 using a density balance. The immersion liquid is anhydrous ethanol. PEEK density specimens manufactured by FFF process (Fig. 2) were used to observe the surface morphology, test the density, and measure the average hardness values of the top and side surfaces (Shore hardness), respectively.

Fig. 2
figure 2

FFF 3D printed PEEK density and hardness specimen

Full size image

2.6 Hardness test

The digital Shore hardness tester (LXD-D, Wenzhou Weidu Electronics Co., LTD, China) was used to perform the hardness of the density specimens.

2.7 Measurement of MFR

MFR measurements were performed on melt flow indexer (HT-3682VM-BA, HongTuo Instrument, China). Testing weight was 5 kg and testing temperature was 380 °C.

3 Results and discussion

3.1 Extruded material

In this experiment, PEEK filaments with different standing times were selected and extruded by FFF process for sampling. The microscopic observation results of extruded PEEK materials at different standing times are shown in Fig. 3. The shape and smoothness of the extruded material are significantly affected by the moisture content of the filament. Among them, at 0 h and 24 h, the shape of the extruded PEEK materials is approximately circular and the surface is very smooth. As the moisture content of PEEK filaments increases, the water will evaporate to form bubbles and expand when it is extruded by the nozzle. Therefore, a small number of internal defects can be observed in the 48-h filament 3D printed sample. However, when the moisture content of PEEK filaments reaches or is nearly saturated, the surface shape of the printed sample will be destroyed, with a large number of depressions. This is caused by the bursting of internal bubbles and the formation of voids after the increase of moisture. The higher the moisture content of PEEK samples 3D printed by fused filament fabrication (FFF) technology, the more surface pores there were, which has the same characteristics as 3D printed poly (lactic-acid) (PLA) samples [13]. Based on the above results, when the water content in PEEK filament reaches a certain range, the water will vaporize and expand internally through high-temperature nozzle extrusion, forming bubbles. In this experiment, the range is from 24 to 48 h. When the water content in PEEK filament is below this range, there is no significant change due to insufficient water to vaporize and form bubbles. When the water content in PEEK filament is higher than this range, due to the high amount of water, the number of bubbles formed inside will also increase, and some may even rupture and form depressions on the surface. Further analysis shows that an increase in moisture content can also cause a decrease in the viscosity of the melted material, thereby accelerating the generation of bubbles and voids. This proves that high moisture content can have a significant negative effect on FFF 3D printed samples.

Fig. 3
figure 3

The shape of the extruded materials (40 × magnification) after 3D printing of PEEK filament for different standing times

Full size image

3.2 Moisture content of filament

PEEK materials have a lower rate of moisture absorption from the environment than other hygroscopic materials. For example, nylon material has a moisture content of up to 5.5 wt% under certain environmental conditions [24]. In this experiment, PEEK filament was first put into a vacuum drying oven for drying. Then the completely dry filament was placed in an environment of 25 ℃, 60%RH, and the standing time of the filament was changed (0 h, 24 h, 48 h, 72 h). When the time is 0 h, it indicates that the filament has just been taken out of the drying oven. The moisture content test results of PEEK filament are shown in Fig. 4. With the increase of standing time, the moisture content of the filament increases significantly.

Fig. 4
figure 4

Moisture content of PEEK filament at different standing times (n = 3)

Full size image

Before testing, the PEEK filament was completely dried in a constant temperature drying oven until there was no change in weight. PEEK materials have a relatively slow hygroscopic rate, which is different from other hygroscopic materials. Therefore, at time T = 0 h, we default that the moisture content of PEEK filament is 0 wt%. From Fig. 4, it could be seen that the moisture content of PEEK filament is directly proportional to the standing time. After a certain standing time, the moisture content of PEEK filament tends to be saturated. When the standing time is 72 h, the moisture content of PEEK filament is about 0.19 wt%. Although subsequent measurements (96 h, 120 h) revealed slight changes in the weight of the filament, the difference in change was already close to the measurement error. Therefore, in this experiment, PEEK filament was placed in an environment of 25 ℃ and 60%RH (simulated conventional room temperature) for 72 h, and the moisture content tended to be saturated, about 0.19 wt%. This value is different from the maximum moisture content of 0.3 wt% for carbon fiber (CF)/PEEK materials at high temperatures [18]. It is worth noting that when the humidity is significantly higher than 60% RH, the moisture content of PEEK filament will also increase and stabilize to about 0.26 wt%. This indicates that the saturation value of moisture content in PEEK filament varies with changes in environmental humidity. The increase in environmental humidity may lead to an increase in the saturation value of moisture content in PEEK filament. This trend is consistent with the upper limit of moisture content of PEEK materials increasing with the increase of water temperature or humidity [19, 23]. However, this experiment simulates a conventional room temperature environment, so this data will not be further explored.

3.3 Surface morphology, density and hardness

The surface morphology of the specimen after 3D printing with different standing times is shown in Fig. 5. It can be clearly seen that the surface of the specimen at 0 h has consistent regular patterns. As the standing time increased, this consistency gradually became less apparent in the specimens at 24 and 48 h. Especially for the 72-h specimen, due to the high moisture content of the filament, the internal bubble rupture, and the external void, resulting in obvious surface defects of the printed specimen. PEEK filament and polyetherimide filament have similar poor surface quality of 3D printed parts under high moisture content conditions [16]. Therefore, the moisture content of PEEK filament will significantly affect the surface morphology of printed specimens. The main reason may be that for the 72h specimen, the surface of the filaments extruded by the nozzle is uneven due to high moisture content. In the process of layer by layer, some positions appear to be filled with dissatisfaction and some positions appear to bubble expansion. Therefore, the quality of the printed specimen surface is reduced.

Fig. 5
figure 5

Surface morphology of PEEK filament printed specimens with different standing times (A 0 h, B 24 h, C 48 h, D 72 h)

Full size image

The water tightness method was used to measure the density of PEEK specimens placed for different times, and the results are shown in Fig. 6. The specimen density of 0 h and 24 h is closer to the density of PEEK raw material (1.28 g/cm3) than that of 48 h and 72 h. The reason why the density is inconsistent with the raw material at the initial time of 0 h may be related to the compactness of 3D printing. The compactness of the specimen after FFF printing did not reach 100%, resulting in a slight difference. When the standing time is 72 h, the density of the specimen is only 94% of the raw material. The main reason is that the excessive moisture content in PEEK filament will cause defects (bubbles or voids) in the 3D printing process, and also lead to a decrease in the viscosity of the molten PEEK filament. In order to verify the changes in filament viscosity, additional tests were conducted. The result confirms our hypothesis that the viscosity of the filament has decreased (Fig. 7). Under the combined effect of the above two reasons, the density of the printed specimen decreases significantly with the increase of the standing time (moisture content).

Fig. 6
figure 6

Density of PEEK specimens printed at different standing times (n = 3)

Full size image
Fig. 7
figure 7

MFR of PEEK filaments at different standing times (n = 3)

Full size image

According to the information shown in Fig. 2, the hardness of the top and side surfaces was tested, and the test results are shown in Fig. 8. Compared with the hardness of PEEK materials produced by conventional processes, the completely dry PEEK specimens printed by FFF have a hardness of D81.25, which is slightly lower than the hardness D85 of PEEK materials produced by conventional processes. In addition, with the increase of standing time, the hardness of printed PEEK specimens decreased, reaching the lowest point of D77.5 at the time = 72 h. It is worth noting that at 24 h, the hardness of the printed PEEK specimen is slightly higher than that of the specimen at 0 h (completely dry). The specific reasons will be verified in subsequent research. For the comprehensive results of the experiment, the hardness of PEEK specimen is inversely proportional to the moisture content of the filament. The main reason is the impact of reduced density and poor surface quality.

Fig. 8
figure 8

Hardness of PEEK specimens printed at different standing times (n = 3)

Full size image

3.4 Tensile strength

The relationship between the moisture content of PEEK filament and the tensile strength of 3D printed parts was investigated by tensile test. It can be seen from the Fig. 9 that the tensile strength of the specimen with time = 0 h is 91.2 MPa. When the time = 72 h, the tensile strength decreased significantly to 86.6 MPa, which was 5% lower than that of the specimen printed with dry PEEK filament. Then, it was observed that the tensile strength of the specimen with time = 24 h was 92.0 MPa, which was slightly higher than the test result with time = 0 h. It should be noted that it is also within the standard deviation of the specimen with time = 0 h, so there is no significant difference here. In addition, the tensile strength of the specimen with time = 48h was 88.9 MPa, which began to decrease compared with the completely dry specimen. The decrease of tensile strength is mainly due to the increase of moisture content which leads to the increase of defects in the layer-by-layer deposition process. The characteristics exhibited by PEEK filament are similar to those of ABS filament [12]. These different defects result in reduced density and weak inter-layer bonding of printed parts. As a result, the overall tensile strength is reduced.

Fig. 9
figure 9

Tensile strength of PEEK specimens printed at different standing times (n = 3)

Full size image

The fracture cross-section of 3D printed tensile specimens of PEEK filament with different standing times after testing is shown in Fig. 10. It can be observed that there is only a small amount of interlayer cracking in the fracture cross-sections of the 0 h and 24 h specimens. As the moisture content of the filament further increases, the interlayer cracking of the fracture cross-section of the 48 h and 72 h specimens significantly increases. The main reason is that the increase in moisture content of PEEK filament leads to a decrease in interlayer bonding and a decrease in tensile strength. The test results of elongation at break are shown in Fig. 11. The elongation at break at 0 h is about 12.3%, while the elongation at break at 72 h reaches the lowest value about 7.91%. The 72-h result showed a significant downward trend compared to the 0 h result, with a 35.5% decrease in elongation at break. However, there is no significant difference in the standard deviation of the observed data between groups. This may be caused by a small sample size or accidental data. Therefore, it was decided to increase the sample size for a second experiment, and the results are shown in Fig. 12. As shown in Fig. 12, there was no significant difference in the results between 0 h, 24 h, and 48 h. However, there was a significant decrease in the results after 72 h compared to 0 h. This result is not consistent with other testing trends. The moisture content of PEEK filament does not significantly affect the elongation at break. In future research, the storage time interval will be reduced to confirm the trend of changes in elongation at break.

Fig. 10
figure 10

SEM images (left: 25 × magnification, right: 100 × magnification) of tensile fracture cross-sections of 3D printed PEEK specimens at different standing times. A is fracture cross-sections of specimens at 0 h; B is the fracture cross-sections of 24 h specimens; C is the fracture cross-sections of 48 h specimens; D is the fracture cross-sections of 72 h specimens

Full size image
Fig. 11
figure 11

Elongation at break of PEEK specimens printed at different standing times (n = 3)

Full size image
Fig. 12
figure 12

A second experiment of elongation at break of PEEK specimens printed at different standing times (n = 5)

Full size image

According to the above research results and analysis, the effect of moisture content on PEEK materials, especially for fused extrusion 3D printing PEEK filaments, should be concerned. The increase of moisture content of PEEK filament will significantly reduce the extrusion quality, the surface morphology, and mechanical properties of 3D printed parts. Therefore, although PEEK filament has a lower moisture content than other hygroscopic materials, it still needs to be stored in a dry environment to ensure the quality of 3D printed products as much as possible. In general, when using PEEK filament for 3D printing, the ambient temperature and humidity should be controlled. The increase of humidity will increase the upper limit of PEEK filament moisture content to a certain extent. When printing PEEK filament using the FFF process, it is recommended to use a closed chamber to avoid absorbing moisture from the air. Unused or leftover filament should be stored under vacuum drying conditions, otherwise necessary drying should be carried out before use. In addition, even when printing in a relatively enclosed chamber, it is still recommended to replace the PEEK filament no more than 24 h to improve the stability of the 3D printed PEEK part quality.

4 Conclusion

This study explored the relationship between the standing time and the moisture absorption of dried polyether-ether-ketone (PEEK) filament in a simulated room temperature environment. Through the measurement of moisture content, it was found that the maximum moisture content of PEEK wire was close to 0.2 wt% after being placed in the simulated room temperature environment for a certain time. When the humidity significantly increases, the maximum moisture content of PEEK filament increases to 0.26 wt%. In addition, through the analysis of the surface morphology, density, hardness, and tensile strength of the extruded PEEK material and 3D printed specimens by fused filament fabrication (FFF), it was found that the moisture absorption has an adverse effect on the PEEK filament. The increase in moisture content of PEEK filament significantly reduces the quality of extruded materials, resulting in poor surface morphology of 3D printed specimens. Meanwhile, weak interlayer bonding also leads to a decrease in the density and hardness of the specimen. In addition, considering the influence of moisture on crystallinity, the tensile performance also decreased. Therefore, it is necessary to establish relevant standards to control the storage and usage cycle of PEEK filament, ensuring consistency in the quality of 3D printed PEEK parts.