Evaluation of Spatial Distribution of Crystallinity Induced by Local Heating Using Low-Frequency Raman Spectroscopy on Polyether Ether Ketone (PEEK)

Abstract

Local heating was performed on a thermoplastic polymer film by contact with the tip of a soldering iron heated above the glass-transition temperature. The locally heated area was measured using microscopic Raman scattering spectroscopy, and the spatial distribution of the crystallinity was obtained from the low-frequency peak. The crystallinity distribution can be evaluated using the microscale spatial resolution. The temperature distribution around the locally heated area was calculated by applying the heat conduction equation, and good correspondence was obtained with the obtained crystallinity.

1. Introduction

Polyether ether ketone (PEEK, Figure 1) is a thermoplastic polymer with excellent mechanical, chemical, and thermal stabilities [1]. Compared with thermosetting polymers, it is expected to have shorter molding times, a lower cost, and higher productivity [2]. It is also attracting attention as a substrate for carbon fiber-reinforced plastics (CFRPs), which are expected to replace steel materials and are being considered as high-performance polymers for a wide range of applications, from automotive and aerospace to dental and orthopedic biomaterials. Its mechanical, optical, electrical, and other properties are highly dependent on its degree of crystallinity [3,4,5,6,7,8,9,10,11,12]. Therefore, evaluation of the degree of its crystallinity is of great importance [13,14,15].

Various techniques have been used to evaluate the crystallinity of polymers. Representative techniques include wide-angle X-ray diffraction (WAXRD), differential scanning calorimetry (DSC) [16], Fourier-transform infrared spectroscopy [13], and Raman spectroscopy [14,15,17,18,19,20]. Raman spectroscopy is one of the most commonly used techniques for investigating polymer materials [21]. In addition, high spatial resolution can be achieved using microscopic optics. It has also been reported that carbon fibers promote crystallization on PEEK [22]. Therefore, high spatial resolution is a useful feature for comparing the degree of crystallinity near and far from carbon fibers in composite materials.

Using the high spatial resolution of micro-Raman scattering spectroscopy, Uematsu et al. reported that in carbon fiber-reinforced polyamide (PA6) composites, various polymorphs (α-phase, intermediate β-phase, and γ-phase crystals) exist near the carbon fiber [23]. The spatial distribution of crystallinity has also been reported using micro-Raman scattering spectroscopy. Yang et al. used Raman spectroscopy to investigate the spatial distribution of crystallinity in poly(ε-caprolactone) (PCL) films and reported that the crystallinity was not uniform throughout the film [24]. Doumeng et al. measured the spatial distribution of crystallinity in wear tracks in PEEK and glass-fiber-reinforced PEEK composites and reported that crystallinity was heterogeneous within the wear tracks [15]. Segawa et al. performed 3D Raman mapping measurements on poly(ε-caprolactone) (PCL) films during marine degradation and reported that crystallinity changed mainly near the film surface [25]. In particular, in recent years, Raman scattering in the low-frequency range has attracted attention in various fields, along with terahertz-range absorption spectroscopic techniques [26,27,28,29]. In some polymer materials, the peaks obtained by low-frequency Raman scattering spectroscopy [30,31,32,33] and THz vibrational spectroscopy [26,27,28,29] have been discussed as being related to intermolecular vibrational modes. It has also been reported that the intensity of low-frequency vibrational peaks in PEEK depends on the degree of crystallinity [34,35], and the origin of these peaks has been explained using molecular dynamics simulations [36].

Laser processing is expected to be applied to polymer materials, including CFRP, as the next-generation processing method [37,38]. It is expected that the crystallinity will change locally owing to the local heating caused by laser processing. Evaluating the spatial distribution of crystallinity owing to local heating in thermoplastic polymers is an important issue for high-quality processing.

In this study, low-frequency Raman scattering spectroscopy was used to measure the spatial distribution of crystallinity caused by crystallization in locally heated amorphous PEEK samples upon heating to the glass-transition temperature.

2. Materials and Methods

2.1. Sample Preparation

Sheets of PEEK polymer (Victrex, Cleveland, UK, Grade 2000-025G, 15 mm × 15 mm, t = 25 µm) were used as samples. Local heating was performed by bringing the tip (R = 0.6 mm) of a temperature-controlled soldering iron (Hozan Corporation, Osaka, Japan, HS-51, 25 mm tip curvature) in contact with the sample surface in atmosphere. As shown in Figure 2, the sample was placed on a microbalance (Conkoo, MH-500, 0.01–500 g), and the load on the microbalance was used to confirm that the tip of soldering iron and sample were in contact. Heating was performed under a load of 2.0 g. The temperature at the tip of the soldering iron was 300 °C, and the heating time was 60 s. After 60 s, the tip of the soldering iron was lifted, and the sample was left to cool in air.

2.2. Experiments

The local heated area was observed by optical microscopy, and the Raman spectra were obtained using micro-Raman spectroscopy (LabRAM HR Evolution, Horiba, Kyoto, Japan). The micro-Raman system consisted of laser (1064 nm, 200 mW), an edge filter (Horiba, Ultra low frequency Raman module, <20 cm⁻¹) to remove the scattered excitation laser, a single spectrograph, and InGaAs detector cooled by liquid nitrogen. The focal length of the spectrograph was 800 mm, and a 300 L/mm grating was used. All measurements were performed at room temperature. The laser power was limited to approximately 200 mW on the sample to minimize the heating effects. A 50× objective lens was used to focus the laser spot to a diameter of approximately 1 µm. The samples were then moved under a microscope using a computer-controlled XY stage. The measurement time was 30 s, and the number of integration times was three. Line measurements were carried out with the center of heating as the origin, sufficiently far from the heated area.

3. Results

Optical microscopy images of the local heating area are shown in Figure 3(1). A circle with a radius of approximately 100 µm, where the tip is expected to be in contact, and a discolored region with a radius of approximately 500 µm centered on the heating point were observed. The Raman spectra of points A and B, indicated by arrows in Figure 3(1), are shown in Figure 3(2), where A and B are the typical Raman spectra of the heated and unheated regions, respectively. In spectrum A of the discolored region, three peaks at 50 cm⁻¹, 97 cm⁻¹, and 135 cm⁻¹ were observed. The peak at 97 cm⁻¹ is assigned to the Ph-O-Ph vibration, and the peaks at 50 cm⁻¹ and 135 cm⁻¹ are assigned to the Ph-CO-Ph vibration. In the spectrum B, the peaks at 50 cm⁻¹ and 97 cm⁻¹ have disappeared. In previous studies, spectra with this feature have been reported to be due to amorphous PEEK. The crystallinity of the area in which this spectrum was observed for amorphous PEEK was set as 0.

The crystallinity was calculated from the intensity ratio at 97 cm⁻¹ and 135 cm⁻¹ obtained by fitting analysis, as reported in a previous study [34]. The intensity of the Raman spectra in the low-frequency region is often corrected by the Bose–Einstein factor, as it is influenced by the measurement temperature [39]. To maintain consistency with a previous study [34] that calculated spectral intensity without the Bose–Einstein factor correction, our spectra have not been corrected. The two peaks are represented by a Gaussian function, and the baseline is a straight line. An example of the fitting analysis is presented in Figure 4. The open circle shows the experimental data, and the solid curve shows the fitting result. The model function accurately reproduced the experimental data. In addition, the fitting elements used are indicated by the dashed lines.

Figure 5 shows the crystallinity distribution obtained using Raman spectroscopy. The horizontal axis represents the distance r from the center of the heating point. Crystallinity was almost constant up to approximately 100 µm. The point at which crystallinity began to decrease was designated as r₁. r₁ is thought to be the area where the tip is in contact with the polymer. The crystallinity gradually decreases from r₁, dropping sharply at approximately 500 µm. The position at which the crystallinity became zero was designated as r₂. The area outside r₂ is thought to be the area where the temperature of the sample was below the glass-transition temperature. We demonstrated that the spatial distribution of crystallinity over a few microns caused by localized heating can be evaluated using Raman scattering spectroscopy.

From the crystallinity obtained from Raman spectroscopy, we considered the temperature distribution within the film surface owing to localized heating. Assuming that the sample is sufficiently thin, the temperature is constant in the depth direction. The heat conduction equation at a distance r from the heating point is expressed as follows.

$${\displaystyle \frac{\mathsf{\partial }}{\mathsf{\partial }r}}\left(r{\displaystyle \frac{\mathsf{\partial }T}{\mathsf{\partial }r}}\right)=0$$

(1)

The radius of the heated area where the tip of the soldering iron was in contact was r₁, the temperature of the soldering iron was T₁, the position where the crystallinity was 0 was r₂, and the glass-transition temperature was T₂. The boundary condition is:

$$T\left(r_1\right)=T_1, T\left(r_2\right)=T_2$$

(2)

T₁ and T₂ were set to 300 °C, and the glass-transition temperature was set to 143 °C. As shown in Figure 4, r₁ and r₂ are set to 100 µm and 554 µm, respectively.

Solving Equation (1) under the conditions of Equation (2) yields:

$$\begin{gathered} T\left(r<r_1\right)=T_1 \\ T\left(r>r_1\right)=T_1-{\displaystyle \frac{\ln r-\ln r_1}{\ln r_1-\ln r_2}}\left(T_2-T_1\right) \end{gathered}$$

(3)

According to Hertz’s contact theory, the contact radius between the copper and PEEK at room temperature is approximately 30 µm. The Young’s modulus of PEEK is known to increase sharply from the glass-transition temperature, and if it is estimated to be approximately 1% of the room temperature, the contact radius would be approximately 130 µm; therefore, r₁ = 100 µm is considered reasonable.

$$a={\left({\displaystyle \frac{3WR}{2E^{\prime }}}\right)}^{{\displaystyle \frac{1}{3}}}$$

(4)

$${\displaystyle \frac{1}{R}}={\displaystyle \frac{1}{R_1}}+{\displaystyle \frac{1}{R_2}}, {\displaystyle \frac{1}{E^{\prime }}}={\displaystyle \frac{1-{\nu }_1^2}{E_1}}+{\displaystyle \frac{1-{\nu }_2^2}{E_2}}$$

(5)

Equation (3) was used to convert the distance r from the heating point to the temperature during heating. The relationship between the temperature distribution during heating and crystallinity is shown in Figure 6. The filled circles indicate the results of previously reported data on the heat treatment temperature and crystallinity of samples heat-treated in a conventional electric furnace [34]; those results are in good agreement with the results of this study.

4. Conclusions

Local heating was performed on a thermoplastic polymer film by contact with the tip of a soldering iron heated above the glass-transition temperature. The locally heated area was measured using microscopic Raman scattering spectroscopy, and the spatial distribution of the crystallinity was obtained from the low-wavenumber peak. The crystallinity distribution can be evaluated using the microscale spatial resolution. The temperature distribution around the locally heated area was calculated by applying the heat conduction equation, and good correspondence was obtained with the obtained crystallinity. In applications that require a high degree of reliability in the mechanical strength of PEEK polymer, such as laser welding in the medical and aerospace fields, crystallinity with micro-level spatial resolution is essential, and it has been shown that evaluation of crystallinity using Raman scattering spectroscopy is useful.