Keywords: Polyether ether ketone, thermal analysis, TGA, evolved gas analysis, DSC, MDSC, FTIR, mass spectrometry, advanced materials, aerospace, electronics, biomedical
TA496
Abstract
Thermal analysis including thermogravimetric analysis (TGA), modulated thermogravimetric analysis (MTGA), evolved gas analysis (EGA), differential scanning calorimetry (DSC), and Modulated DSC™ Analysis (or MDSC™ Analysis) was performed on a screw made from polyether ether ketone (PEEK). Decomposition temperature, filler amount, quantification of pyrolytic carbon, active energy, decomposition products, glass transition, and melting and crystallization transitions were obtained. Additionally, MDSC Analysis was utilized to isolate and quantify enthalpic recovery (due to enthalpic relaxation). Enthalpic recovery is detected as an endotherm occurring at the glass transition. Due to the kinetic nature of enthalpic relaxation, enthalpic recovery can be isolated in the non-reversing heat flow signal in the MDSC Experiment. Nine (9) different decomposition products were identified using evolved gas analysis with phenol as the main decomposition product.
Polyether ether ketone (PEEK) (Figure 1) is chemically known as Poly(oxy-1,4-phenyleneoxy-1,4-phenylenecarbonyl-1,4-phenylene). It is a semicrystalline thermoplastic with excellent thermal and chemical resistance, mechanical properties including modulus, fatigue resistance, and resistance to abrasion. PEEK’s high temperature resistance makes it an excellent choice for demanding applications such as aerospace, defense, electronics, automotive, and any application where heat, chemical resistance, and electrical insulation are needed. It is also used in medical applications such as prostheses, dental implants, orthopedic applications, and other medical devices. As with any plastic material, there are issues with PEEK such as cost, processing complications, and susceptibility to certain chemicals, especially halogenated compounds.
Thermal analysis techniques such as TGA and DSC can be used to understand properties such as thermal transitions and decomposition temperatures of a material. Applying modulation to these techniques offers further insight, such as activation energy and enthalpic recovery. Coupling a TGA system with EGA further enhances the understanding of the material being examined through elucidating decomposition products and potential outgassing. In this work, PEEK is investigated using TGA, modulated TGA, TGA-EGA with FTIR and GC/MS, DSC and Modulated DSC Techniques.
Commercially available PEEK socket head screws were used in this work. Shown in Figure 2, the screws were 2-56 thread and a half inch long.
TGA experiments were run on a TA Instruments™ Discovery™ 5500 using nitrogen purge at 30 mL/min. Samples of nominally 10 mg were loaded into 100 μL platinum pans and heated at 10 °C/min from ambient to 1000 °C. For MTGA, the samples were heated at 2 °C/min with an amplitude of ± 5 °C and a period of 200 s.
Evolved gas analysis was performed by coupling the Discovery 5500 with FTIR and gas chromatography mass spectrometry (GC/MS). A Nicolet™ iS50 FTIR Spectrometer, 7890B GC from Agilent paired with MS, and HP-5MS 5% phenyl methyl siloxane 30 m × 250 μm × 0.25 μm column were used. The hyphenated TGA-FTIRGC/MS system was coupled together using the TL9000 Interface from RedShift.
A TA Instruments Discovery 2500 was used for DSC and MDSC Experiments. The sample mass was 3 mg and TZero™ Aluminum Pans were used. Heating and cooling rates were 10 °C / min. MDSC Experiments were conducted with an amplitude of ± 1 °C with a period of 60 s and a heating rate of 1.5 °C/min. All DSC experiments were run in a nitrogen purge of 50 mL/min.
TGA Results
The first approach in thermal analysis is a TGA experiment to determine relative stability, mass loss events, filler content, and decomposition temperature. The decomposition temperature was chosen as the temperature where 2% mass loss occurs in N2. TGA data and mass loss summary are shown in Figure 3 and summarized in Table 1.
Table 1. Summary of mass loss events for PEEK sample in N2
| Event | Mass (%) | Temperature at Event (°C) |
|---|---|---|
| Decomposition Onset* | 0.32 | 551.25 |
| Decomposition Temperature | 2.00 | 557.2 |
| 1st Mass Loss | 35.6 | 577.6 (max mass loss rate) |
| 2nd Mass Loss | 4.31 | 726.4 (max mass loss rate) |
| Residue | 60.5 | 1000 |
* Decomposition onset temperature is taken as a tangency line drawn from the base and initial ascending part of the derivative of mass loss with respect to temperature.
The aromatic structure of PEEK and large amount of residue remaining in the TGA experiment run in N2 necessitates another TGA experiment with a switch to air to determine the fraction of pyrolytic carbon in the sample as well as any filler content. TGA data and mass event summary are shown in Figure 4 and Table 2.
Pyrolytic carbon in the PEEK sample is 38.6% with filler content of 21.94%.
Table 2. Summary of mass events for PEEK sample with gas switch to air at 600 °C
| Event | Mass (%) | Temperature at Event (°C) |
|---|---|---|
| Decomposition Onset* | 1.36 | 552.1 |
| Decomposition Temperature | 2.00 | 556.0 |
| 1st Mass Loss | 30.90 | 577.9 (T at max mass loss rate) |
| 2nd Mass Loss | 47.81 | 635.5 (T at max mass loss rate) |
| Residue | 21.94 | 1000 |
* Decomposition onset temperature is taken as a tangency line drawn from the base and initial ascending part of the derivative of mass loss with respect to temperature.
Modulated TGA (MTGA)
The MTGA experiment allows calculation of the activation energy in a single experiment. The activation energy and pre-exponential factor can be plotted as a signal on the y-axis as a function of temperature, time, mass percent or the more familiar extent of conversion (α) which can be defined as a user variable. The MTGA experiment is a variation of the factor jump method devised by Flynn [1] and described by Blaine [2, 3]. For this sample, the activation energies at the maximum rate of decomposition were determined and shown in Figure 5 and summarized in Table 3.
Table 3. MTGA of PEEK sample summary
| Event | Activation Energy (kJ/mol) | T at maximum rate of conversion (°C) |
|---|---|---|
| 1st Mass Loss | 141.1 | 520.81 |
| 2nd Mass Loss | 272.6 | 671.47 |
Figure 6 shows the TGA mass loss with derivative and derivative of modulated mass loss. Evaluating the number of oscillations of the full width and half-height is useful for determining the validity of the MTGA experiment. This number should be at least five [2].
Another view of the MTGA experiment is shown in Figure 7: plotting the activation energy and rate of mass loss as function of mass fraction converted. Activation energies at the first and second mass loss differ significantly. The first mass loss maximum rate shows an activation energy of 141 kJ/mol at 520.8 °C at a conversion of 0.14, the second mass loss maximum rate shows an activation energy of 306.0 kJ/mol at 678.6 °C at a conversion of 0.34.
Evolved Gas Analysis
Evolved Gas Analysis (EGA) is a powerful addition to the TGA experiment. Chemical analysis of the evolved gases by FTIR, MS, or GC/MS, or all of them combined, allows identification of the decomposition products which is useful for process safety concerns as well as identification of the original sample in many cases. We have demonstrated the utility of EGA in identifying difficult to analyze samples [4]. For the PEEK sample, we utilize TGA/FTIR/GC/MS in the configuration shown in Figure 8.
First Mass Loss – TGA/FTIR
As shown in Figure 9, the TGA data is overlaid with the mass loss, derivative, and Gram-Schmidt reconstruction from the FTIR data. Often the Gram-Schmidt reconstruction will resemble the derivative of the rate of mass loss, which makes identifying spectra corresponding to mass loss events simple. Two mass loss events occurring at 577 °C and 725 °C will be considered.
Figure 9. TGA / EGA analysis in N2
The FTIR spectrum at 577 °C is shown in Figure 10. Obvious components include CO2, CO, water, aromatic hydrocarbons, and free hydroxyls, which is likely a phenolic species based on the structure of the polymer.
Subtraction of water vapor cleans the spectrum up significantly (Figure 11). The reference water vapor spectrum was generated using the first mass loss of calcium oxalate. From the structure in Figure 1, one can infer that one of the major decomposition products is likely to be phenol. A comparison of the unknown spectrum and a NIST web book reference of phenol is shown in Figure 12. Spectral features assignable to phenol confirm it as a major component. Of course, this does not account for all the spectral absorbances indicating that the by-product is a complex mixture. The two possible lower frequency carbonyls at 1682 and 1656 cm-1 are likely indicative of aryl ketones and quinones. There are also other C-O stretching bands. A close-up of the fingerprint region is shown in Figure 13.
Second Mass Loss – TGA/FTIR
The second mass loss occurs at approximately 726 °C. Most of the spectral features in the second mass loss (Figure 14) are comparable to the first mass loss but also with evolution of methane (Figure 15).
Patel and coworkers published extensive work on the decomposition products and proposed decomposition mechanisms of PEEK [5]. The decomposition products identified by Patel et. al. are shown in Table 4. The infrared spectra at both mass losses are consistent with many of the components identified in Patel’s work.
Table 4. Decomposition Products of PEEK identified by Patel and coworkers
| 4-Phenoxyphenol | 4-Hydroxybenzophenone |
| 1,4-Diphenoxybenzene | p-Benzoquinone |
| CO + CO2 | Benzophenone |
| Diphenyl Ether | Biphenyl |
| Phenol | Naphthalene |
| Benzene | Fluorene |
| Dibenzofuran | 4-Hydroxybenzophenone |
| Hydroquinone | 1,4-Diphenoxybenzene |
| 4-Dibenzofuranol | 4-Phenylphenol |
TGA/GC/MS
Each of the mass losses obtained in the TGA experiment and shown in Figure 9 were injected as separate run onto the GC column, and mass spectra were obtained. The chromatogram of the first mass loss event at 577 °C is shown in Figure 16, with corresponding mass loss components listed in Table 5. Figure 17 shows the chromatogram of the second mass loss event at 726 °C and Table 6 summarizes the second mass loss components that show good agreement with those identified by Patel, et. al.
Table 5. Summary of 1st mass loss components at 577 °C
| TIC Peak | RT (min) | Species |
|---|---|---|
| 1 | 3.261 | Benzene |
| 2 | 6.991 | 671.47 |
| 3 | 8.550 | Phenol |
| 4 | 20.024 | Diphenyl ether |
| 5 | 22.872 | Dibenzofuran |
| 6 | 27.620 | 4-phenoxyphenol |
| 7 | 30.366 | 2-Dibenzofuranol |
| 8 | 34.184 | 4-Hydroxybenzophenone |
| 9 | 41.394 | 1,4-diphenoxy benzene |
Table 6. Summary of 2nd mass loss components 726 °C
| TIC Peak | RT (min) | Species |
|---|---|---|
| 1 | 8.447 | Phenol |
| 2 | 20.028 | Diphenyl Ether |
| 3 | 22.863 | Dibenzofuran |
| 4 | 27.615 | 4-phenoxyphenol |
One of the advantages of the TGA/FTIR/GC/MS configuration is the ability to monitor the evolution of gases in real time on the continuous collecting FTIR and inject analytes of interest to the GC/MS at any time. In this experiment, injections were made at the temperature at maximum decomposition rate.
DSC
The PEEK screw as received was cut and fit to the Tzero Aluminum Pan. Running the part as received is designed to assess any process history imparted to the sample during the manufacturing process. The first heat, shown in Figure 18, shows a prominent glass transition at 144 °C, cold crystallization at 171 °C, and a melting transition at 343.2 °C with a heat of fusion of -50.7 J/g. Integration forward of the glass transition subtracts the heat associated with cold crystallization from the heat of fusion of the melt transition. It also accounts for any melting / recrystallization that can occur during a crystal perfection process.
Cooling at 10 °C/min shows a crystallization temperature of 292 °C, heat of crystallization of 40.2 J/g, and glass transition at 145 °C (Figure 19).
The second heat (Figure 20) shows a glass transition at 150.3 °C, melting temperature at 340.9 °C, and heat of fusion of -42.4 J/g.
A summary of the DSC data is shown in Table 7. Corrected heat of fusion and crystallization is based on the sample containing 21.9% filler determined in the second TGA experiment. Heat of crystallization in first heat is estimated from a running integral of the cold crystallization peak.
Table 7. Summary of PEEK DSC data
| 1st Heat | Cool | 2nd Heat | |
|---|---|---|---|
| TG °C | 144.4 | 145.3 | 150.3 |
| TM °C | 343.2 | – | 340.9 |
| TC °C | 171 | 291.5 | – |
| ΔHF (J/g) | -50.7 | – | -42.41 |
| ΔHC (J/g) | 2.8 (est) | 40.24 | – |
| ΔHF (J/g) (Corrected) | -64.95 | – | -54.33 |
| ΔHC (J/g) (Corrected) | 3.6 (est) | 51.55 | – |
The relatively high heat of fusion observed in the first heat of the DSC experiment is likely due to crystallinity or other metastable structure imparted during the screw manufacturing process. The integration limits were taken forward of the glass transition to subtract the cold crystallization to the baseline beyond the melt. A comparison of the first and second heats is shown in Figure 21.
Modulated DSC (MDSC) Experiments
The observation of cold crystallization and obvious glass transition in the first heat of the DSC experiment is indicative of metastable amorphous phase imparted during processing. Often PEEK is used in high temperature applications that may allow enthalpic relaxation of that amorphous phase and may be worth considering. Indication of enthalpic relaxation is observed as an endotherm at the glass transition in the DSC and is referred to as enthalpic recovery. Enthalpic relaxation is a kinetic process dependent on time and temperature and as such can be isolated from the heat flow by means of a Modulated DSC Experiment. MDSC Techniques superimpose a sinusoidal heating rate on the linear heating rate and separate the heat flow signal into a heat capacity component (reversing heat flow) and a kinetic component (non-reversing heat flow) [6] [7]. The total heat flow is the sum of the reversing and nonreversing heat flow signals.
The results of the MDSC Experiment are shown in Figure 22 and summarized in Table 8. MDSC Techniques separate the reversing and non-reversing heat flow signals, allowing calculation of enthalpic recovery. The energy associated with enthalpic recovery and cold crystallization can be quantified with a simple integration. A comparison of the first and second heats of the non-reversing heat flow is shown in Figure 23. As the glass transition temperature is approached, the rate of enthalpic relaxation will increase – resulting in reduction in free volume, densification, increased modulus, and decreased heat capacity resulting in a structure with more tendency to brittle fracture rather than ductile deformation. The effects of processing, additives, impact modifiers, etc. on the kinetics of enthalpic relaxation can be directly measured by MDSC. The heat of crystallization of the cold crystallization transition can be isolated and quantified. The heat capacity change, which is proportional to the amount of amorphous phase, can be more accurately determined by isolating it in the reversing heat flow signal.
Table 8. Summary of MDSC Analysis of PEEK at glass transition
| Transition | Value |
|---|---|
| Enthalpic recovery – non reversing heat flow (J/g) | -0.283 |
| TC – non reversing heat flow (°C ) | 162.9 |
| ΔHC – non reversing heat flow (J/g) | 6.09 (observed), 7.80 (corrected) |
| TG Total heat flow (°C) | 142.3 |
| Heat capacity change- total heat flow (J/g °C) | 0.202 |
| TG Reversing heat flow (°C) | 141.9 |
| Heat capacity change – Reversing heat flow (J/g °C) | 0.055 |
Leonard Thomas has published more details in the analysis of the glass transition and quantification of enthalpic recovery in one of a suite of papers on the subject of MDSC Techniques, available on the TA Instruments website (www.tainstruments.com) [8].
Basic thermal analysis of materials has evolved over the years as instrumentation technology has advanced. Modulated thermal analysis and evolved gas analysis are needed to augment the basic understanding of thermal properties of materials.
For this work, a screw made from polyether ether ketone (PEEK) was analyzed. In addition to TGA providing decomposition temperature and filler content, modulated TGA provides activation energy at the two observed mass loss events. The effects of additives, fillers, copolymers, and impact modifiers on the relative stability of plastics can be investigated in a single experiment to obtain activation energy (E), and the pre-exponential frequency factor (A). Kinetic information can be used in mathematical models to estimate polymer lifetime as a function of temperature or time. Evolved gas analysis (EGA) combines TGA with FTIR, mass spectrometry (MS), and GC/MS. In this work, decomposition products at the two mass losses were evaluated using TGA/FTIR/GC/MS and show excellent agreement with previous investigations [5].
DSC analysis of the screw show processing effects in the first heat as well as cold crystallization. The prominent glass transition and cold crystallization due to the amorphous phase led to an MDSC Analysis to determine the presence of enthalpic relaxation. The kinetic nature of enthalpic relaxation can be isolated and measured as enthalpic recovery in the non-reversing heat flow signal.
References
- J. Flynn and B. Dickens, “Steady State Parameter-Jump Methods and Relaxation Methods in Thermogravimetry,” Thermochimica Acta, vol. 15, pp. 1-16, 1976.
- R. Blaine and B. Hahn, “Obtaining Kinetic Parameters by Modulated Thermogravimetry,” Journal of Thermal Analysis, vol. 54, pp. 695-704, 1998.
- R. Blaine, “A Faster Approach to Obtaining Kinetic Parameters,” American Laboratory, 1998.
- Browne J.A., “Idenification of a Cross-Linked Rubber by Evolved Gas Analysis and Thermal Analysis Techniques,” TA Application Note TA416, pp. 1-5.
- P. Patel, R. Hull, R. McCabe, D. Flath, J. Grasmeder and M. Percy, “Mechanism of Thermal Decomposition of Poly(Ether Ether Ketone) (PEEK) From a Review of Decomposition Studies,” Polymer Degradation and Stability, vol. 95, pp. 709-718, 2010.
- M. Reading, A. Lacey and D. Price, Modulated Temperature Differential Scanning Calorimetry; Theory and Practice of Modulated Temperature Scanning Calorimetry, D. H. Mike Reading, Ed., 2006, pp. 1-80.
- L. C. Thomas, “Modulated DSC Paper #2: Modulated DSC Basics; Calculation and Calibration of MDSC Signals,” TA Instruments, New Castle.
- Leonard Thomas, “Modulated DSC Paper #5 – Measurement of Glass Transitions and Enthalpic Recovery,” TA Instruments, New Castle.
Acknowledgement
For more information or to request a product quote, please visit www.tainstruments.com to locate your local sales office information.
This application note was written by James Browne, Senior Applications Scientist – Waters TA Instruments.
TA Instruments, Discovery, Modulated DSC, MDSC, and TZero are trademarks of Waters Technologies Corporation. Nicolet is a trademark of Thermo Fisher Scientific Inc.
Click here to download the printable version of this application note.