As an engineering researcher focusing on thermal-management components for electrified mobility, I recognize that the performance requirements for heating elements in electric vehicles are fundamentally different from those in stationary appliances. My study was motivated by the fact that modern electric vehicles require cabin heaters and battery warmers that operate over extremely wide environment-temperature windows, usually from -40 °C to over 100 °C, while the available DC-bus voltage spans a similarly broad range. The positive temperature coefficient of resistance thermistor, commonly abbreviated as PTCR, is one of the most attractive heating sources because of its self-regulating and self-limiting electrical behaviour. In electric vehicles, PTCR ceramic heaters deliver rapid warm-up, structural simplicity, and excellent intrinsic safety.
During the development of my research plan, I reviewed the existing literature with respect to BaTiO₃-based PTCR ceramics and recognized that a critical technical bottleneck remains. Although PTCR elements have been mass-produced for decades, their resistance-temperature curve is not ideal. Below the Curie temperature, materials exhibit a tail of negative temperature coefficient behaviour: the resistance decreases as temperature rises, giving a pronounced resistance minimum before the positive jump occurs. For electric vehicles, this NTC tail is problematic. In cold regions, when a heater is switched on, the initial impedance of the PTCR element can be unnecessarily high, delaying the warming process. Once current flows and Joule heating raises the temperature, the resistance drops into the NTC region, creating a large inrush current. The impact current can stress the electric-vehicle power bus and, in severe cases, cause destructive breakdown of the ceramic. Thus, the design target of my thesis was to establish a ceramic doping and processing strategy that simultaneously suppresses the sub-Curie NTC effect, preserves a strong PTCR anomaly, raises the voltage endurance, and keeps the room-temperature resistance within a suitable range for electric-vehicle heaters.
1. Choice of Material System and Preparation Method
The classic BaTiO₃ PTCR material has a Curie point near 120 °C, which is insufficient for many electric-vehicle heating scenarios that call for higher surface temperatures. Therefore I selected a lead-containing solid solution with the general formula (Ba,Pb)TiO₃, abbreviated BPT, so as to shift the Curie temperature above the operating temperature of the heater. I adopted the two-step solid-state reaction route. In the first step, I synthesized BaTiO₃ and PbTiO₃ separately. The second step mixed the two pre-formed phases with all dopants and sintering aids. The individual synthesis of PbTiO₃ considerably reduced lead volatilization during subsequent sintering, which is important for process stability and environmental protection. My raw materials included BaCO₃, Pb₃O₄, TiO₂, CaCO₃, Nb₂O₅, Mn(NO₃)₂, Al₂O₃, SiO₂, Li₂CO₃ and a PVA binder. Powders were weighed on an analytical balance, milled with zirconia media in deionized water, dried, calcined, remilled for a long duration, granulated through a 40-mesh sieve, and pressed into rectangular green bodies with a size of 24 mm × 15 mm and a thickness of about 2.4 mm. Sintering was performed in a bell-type furnace in air. The critical high-temperature segment followed a staircase profile, with a soaking step at maximum temperature followed by controlled cooling. The resulting ceramic bodies were ground, coated with aluminium paste by screen printing, and baked to form the final ohmic electrodes.
The nominal base composition used in the systematic study can be written as
$$
(\mathrm{Ba}_{0.77-x}\mathrm{Pb}_{0.20}\mathrm{Ca}_{x})\mathrm{Ti}_{1.01}\mathrm{O}_3
\qquad (0.03 \leq x \leq 0.10).
$$
The master batch always contained a fixed semiconducting donor, Nb₂O₅, in addition to Mn as an acceptor, together with the liquid-phase promoters Al₂O₃ and SiO₂. The final combined additive system I explored is referred to below as the CAST system. The essential idea was to utilize the simultaneous occupation of A-site and B-site by calcium with the known effects of aluminium and silicon on the grain-boundary chemistry.
2. Electrical Characterization Techniques
To evaluate my samples, I employed a complete set of electrical and structural measurements. The resistance-temperature curve, denoted R-T, was measured with a programmable ZWX-BC system. From the R-T curve I extracted the room-temperature resistance R₂₅, the Curie temperature TC, the maximum resistance Rmax, the minimum resistance Rmin, and the commonly used figures of merit.
The resistance temperature coefficient α over an interval from T₁ to T₂ is defined by
$$
\alpha(T_1,T_2)=\frac{\ln R_{T_2}-\ln R_{T_1}}{T_2-T_1}
$$
and the intrinsic logarithmic derivative at a given temperature is
$$
\alpha=\frac{1}{R_T}\,\frac{\mathrm{d}R_T}{\mathrm{d}T}.
$$
The jump ratio is
$$
\beta=\frac{R_{\max}}{R_{\min}},
$$
and the NTC tail below the Curie point is quantitatively expressed by
$$
\gamma=\frac{R_{25}}{R_{\min}}.
$$
When γ is greater than unity, an NTC tail exists; when γ is smaller than or equal to unity, the resistance below TC does not fall below the room-temperature value, indicating that the unwanted NTC effect is eliminated. The voltage-current behaviour was tested with a lamp transformer combined with a calibrated digital multimeter, and the alternating-current impedance spectra were recorded from 100 Hz to 10 MHz to separate grain-boundary and grain contributions.
3. Influence of CaCO₃ on BPT Ceramics
3.1 R-T Characteristics
I prepared two groups of Ca-doped samples: the first group was sintered at a maximum temperature of 1262 °C, the second at 1240 °C. All other processing conditions were kept identical. As the calcium content increased from 3 mol% to 10 mol%, the Curie point moved slightly toward lower temperatures and the resistance transition progressively broadened. This displacement is consistent with calcium entering the B-site. Calcium substitution in the titanium sublattice interrupts the Ti-O network that controls ferroelectricity, and therefore gradually reduces the ferroelectric transition temperature.
A remarkable observation was the gradual disappearance of the negative resistance tail as x approached 8 mol%. At the 1262 °C sintering temperature, the sample with 8 mol% calcium exhibited γ = 0.9225, which is less than unity, so the ceramic showed no NTC effect at all below its Curie temperature. At the same doping level, the jump ratio remained high, β = 4.19, and the room-temperature resistance reached 4991 Ω. The 1240 °C series followed a similar trend, with the minimum γ equal to 1.062 at 8 mol% calcium.
| Ca content (mol%) | Sintering 1262 °C: R₂₅ (Ω) | Sintering 1262 °C: γ = R₂₅/Rmin | Sintering 1262 °C: β | Breakdown voltage (V) |
|---|---|---|---|---|
| 3 | 17.0 | 2.35 | 4.66 | 527.8 |
| 5 | 58.8 | 2.21 | 3.53 | 633.4 |
| 6 | 967 | 3.65 | 3.43 | 727.2 |
| 7 | 875 | 1.35 | 2.78 | 774.5 |
| 8 | 4991 | 0.92 | 4.19 | 892.7 |
| 9 | 935 | 3.63 | 3.48 | 844.5 |
| 10 | 38.9 | 2.25 | 1.79 | 521.5 |
| Ca content (mol%) | Sintering 1240 °C: R₂₅ (Ω) | R₂₅/Rmin | β | Breakdown voltage (V) |
|---|---|---|---|---|
| 3 | 46.0 | 2.41 | 3.46 | 423.0 |
| 5 | 758 | 2.09 | 3.24 | 456.1 |
| 6 | 155 | 2.49 | 3.12 | 474.1 |
| 7 | 10180 | 1.27 | 3.03 | 533.7 |
| 8 | 3287 | 1.06 | 3.00 | 665.3 |
| 10 | 1.83×10⁶ | 2.12 | 1.71 | >1100 |
The sample sintered at 1262 °C with 8 mol% Ca was selected for detailed current-time measurement. I calculated the expected resistance at low temperatures according to the Arrhenius-type NTC equation commonly used for thermistors:
$$
R_T = R_{T_{\mathrm{ref}}}
\exp\left[B\left(\frac{1}{T}-\frac{1}{T_{\mathrm{ref}}}\right)\right].
After fitting from R₂₅ and Rmin, I found that the extrapolated resistance at -40 °C remained very close to the room-temperature value. Therefore, during an electric-vehicle cold start the ceramic begins with a stable impedance and does not create a large current burst. The extraordinary stability of resistance with respect to temperature, from -40 °C up to the Curie region, provides a decisive advantage for electric-vehicle heating systems operating in cold winters.
3.2 Inrush Current and Starting Time
I compared several elements with different NTC strengths by recording their inrush behavior at 220 V, 260 V, and 300 V. I defined the starting time as the interval needed for the current to fall from the initial reading to half of the maximum value. Sample 3, with a strong NTC tail, displayed a large inrush and long starting time. In contrast, Sample 1 and Sample 2, both with nearly negligible NTC tails, showed small current overshoots and rapid stabilization. The measured current ratio was as low as 1.0 for Sample 2, meaning that no significant inrush current existed at all. This behavior is shown in the summary below.
| Applied voltage (V) | t_start, Sample 1 (ms) | t_start, Sample 2 (ms) | t_start, Sample 3 (ms) | I_max/I₀, Sample 1 | I_max/I₀, Sample 2 | I_max/I₀, Sample 3 |
|---|---|---|---|---|---|---|
| 220 | 61 | 54 | 192 | 1.08 | 1.00 | 2.23 |
| 260 | 52 | 48 | 188 | 1.09 | 1.00 | 2.25 |
| 300 | 39 | 34 | 185 | 1.05 | 1.00 | 2.26 |
According to the industrial standard used for electric-vehicle PTC heaters, the maximum inrush current of a self-controlled heater shall not exceed 2.5 times the steady-state current. The CAST-based samples with γ close to unity easily met this requirement without an external pre-charge controller. Since the current remained virtually constant during startup, the ceramic elements could be classified simply by their nominal resistance, avoiding the elaborate sorting procedure that conventional PTCR materials require because of their scattered inrush ratios. This is practically significant for electric-vehicle heating modules because it enables wider production tolerance and simplified quality control.
3.3 Microstructure and Defect Chemistry of Ca-Doped Samples
Scanning-electron-microscopy observations showed that calcium acts as an effective grain refiner when introduced in concentrations up to roughly 7 mol%. The average grain size decreased, and the grain-size distribution became more homogeneous. When the Ca content reached 8 mol%, the ceramic displayed a dense and uniform microstructure. Beyond this value, local liquid phases containing excess calcium caused abnormal grain growth and non-uniform contrast in the backscattered-electron image.
X-ray diffraction confirmed the BPT perovskite as the only main crystalline phase for all samples. However, the substitution of calcium into the titanium site changed the lattice dimensions: the perovskite peak shifted to higher diffraction angles when Ca replaced Ba but shifted back when Ca started to replace Ti as well. For samples with calcium at or above 5 mol%, weak diffraction peaks corresponding to Ca₂SiO₄ appeared. The secondary phase was strongest at 8 mol% Ca, implying that excess calcium reacted with SiO₂ along the grain boundaries and contributed to the improved voltage endurance.
I analysed the related defect equilibria. For low Ca concentrations, calcium merely fills the barium site:
$$
\mathrm{CaO \xrightarrow{BaTiO_3} Ca_{Ba}^{\times} + O_O^{\times}}.
$$
As the Ca concentration increases, calcium begins to enter the titanium site with the formation of oxygen vacancies:
$$
\mathrm{CaO \xrightarrow{BaTiO_3} Ca_{Ti}^{\prime\prime} + V_O^{\bullet\bullet} + O_O^{\times}}.
$$
The doubly positive oxygen vacancies attract molecular oxygen from the sintering atmosphere and create hole-trap states:
$$
V_O^{\bullet\bullet} + \frac{1}{2}O_2(g) \rightleftharpoons O_O^{\times} + 2h^{\bullet}.
$$
Under the simultaneous donor doping with Nb, the niobium pentoxide occupies titanium sites and releases electrons:
$$
\mathrm{Nb_2O_5 + 2TiO_2 \xrightarrow{BaTiO_3} 2Nb_{Ti}^{\bullet} + 6O_O^{\times} + 2e^{\prime} + \frac{1}{2}O_2(g)}.
$$
The electrons and holes recombine to maintain charge neutrality:
$$
e^{\prime} + h^{\bullet} \rightarrow \mathrm{nil}.
$$
With nearly equal Ca and Nb occupation at the B-site, the acceptor compensation from calcium-derived oxygen vacancies cancels the donor electrons of niobium, raising the grain-boundary potential barrier. This explains why the maximum room-temperature resistance and the strongest PTC effect appear around 8 mol% Ca.
Beyond the simple compensation effect, my X-ray and microstructure findings support a specific mechanism for the removal of the NTC tail. A previously reported explanation was that the low-temperature NTC conduction in BaTiO₃ originates from electron hopping between Ti³⁺ and Ti⁴⁺ ionic states in the crystal boundary region, a process promoted by the TiO₂ sublattice. When Ca²⁺ substitutes for Ti⁴⁺, the density of available titanium neighbours decreases, the hopping paths are interrupted, and the thermally activated conductivity responsible for the negative resistance slope is suppressed. Consequently, R₂₅/Rmin falls below unity when the calcium content is sufficient. This mechanism is remarkably well-suited to the operational needs of electric vehicles, which require resistance stability over a wide ambient temperature range before the PTC jump begins.
4. Effect of Al₂O₃ in the CAST System
With the optimized Ca content fixed at 8 mol%, I varied the Al₂O₃ addition from 2.0 mol% to 3.8 mol%. The R-T measurements showed that the temperature coefficient jumped significantly with increasing Al content. The aluminium-doped ceramics had maximum resistances exceeding 10⁶ Ω, which is valuable for safe operation under high voltage in electric-vehicle systems.
The room-temperature resistance first increased when Al rose from 2.0 mol% to 2.6 mol% and then decreased for higher additions. This non-monotonic change can be explained by the amphoteric behaviour of aluminium ions in the perovskite lattice. At low Al concentration, Al³⁺ probably substitutes for Ti⁴⁺ and behaves as an acceptor:
$$
\mathrm{Al_2O_3 \xrightarrow{BaTiO_3} 2Al_{Ti}^{\prime} + 3O_O^{\times} + V_O^{\bullet\bullet}}.
$$
The oxygen vacancy then oxidizes during cooling to generate holes, thereby raising the potential barrier. At higher Al₂O₃ amounts, however, a fraction of the Al³⁺ may enter the Ba site as an effective donor, producing self-compensation:
$$
\mathrm{Al_2O_3 \xrightarrow{BaTiO_3} Al_{Ba}^{\bullet} + Al_{Ti}^{\prime} + 3O_O^{\times}}.
$$
As a result, for very high Al concentrations the room-temperature resistance decreases again, whereas the temperature coefficient and voltage endurance remain high.
| Al₂O₃ content (mol%) | R₂₅ (Ω) | R₂₅/Rmin | I_min (mA) | Breakdown voltage (V) |
|---|---|---|---|---|
| 2.0 | 9313 | 1.01 | 18.5 | 804 |
| 2.6 | 4452 | 1.21 | 28.8 | 827 |
| 3.2 | 17471 | 1.38 | 68.6 | 841 |
| 3.8 | 12143 | 1.50 | 18.3 | 876 |
The microstructural images revealed that Al₂O₃ promoted densification by generating a liquid phase that filled the pores between grains. At concentrations higher than 3.2 mol%, X-ray diffraction disclosed a new secondary phase, Ba₃Al₁₀TiO₂₀, in addition to the Ca₂SiO₄ already present. The combination of a well-distributed glassy phase and fine-grained microstructure is beneficial to voltage endurance because the secondary phases flatten and redistribute the electric field along the grain boundaries. I also observed that all Al-containing samples retained γ below 1.5, so the NTC suppression achieved by Ca remained intact when aluminium was introduced.
5. Influence of SiO₂ Addition
Silicon dioxide is normally added to PTCR ceramics in conjunction with TiO₂ and Al₂O₃ because it forms a low-melting AST liquid that assists mass transport and semiconducting grain growth. I studied SiO₂ additions from 1.4 mol% to 3.8 mol% while maintaining the 8 mol% Ca formulation and similar firing conditions.
The sample containing only 1.4 mol% SiO₂ was not fully semiconducted; it appeared yellow and had a very high resistance. At 2.0 mol% SiO₂ the material became blue-grey and semiconducting, with a room-temperature resistance of about 1200–2100 Ω and a very weak NTC tail. The optimum voltage endurance appeared for approximately 2.6 mol% SiO₂, where the breakdown voltage reached 946 V. The resistance then rose strongly with further SiO₂ because excess silicon remained at the grain boundaries, precipitated a Ca₂SiO₄ layer, and increased the intergranular barrier. The R₂₅/Rmin ratio also deteriorated for SiO₂ concentrations larger than 3.2 mol%, showing that an excessive glassy phase can re-introduce the low-temperature resistance drop.
| SiO₂ addition (mol%) | Semiconducting behaviour | R₂₅ (Ω) | R₂₅/Rmin | Breakdown voltage (V) |
|---|---|---|---|---|
| 1.4 | No | too high | — | — |
| 2.0 | Yes | 1200–2141 | ~1.0 | 770–802 |
| 2.6 | Yes | 3032–3065 | ~1.0 | 933–946 |
| 3.2 | Yes | 10041 | >1 | 528 |
| 3.8 | Yes | 53216 | >1 | 513 |
The impedance spectra of the Si-varied samples showed only one dominant semicircle in the complex plane, confirming that the ceramic resistance is governed by the grain-boundary response. I extracted the grain and grain-boundary contributions from the intercepts on the real axis. All changes of R₂₅ with SiO₂ content were mirrored by the grain-boundary resistance, whereas the grain resistance remained almost constant. Therefore, silicon mainly modifies the potential barriers at the boundaries: a moderate amount of liquid phase improves grain contact and homogeneity, while excess SiO₂ forms a high-resistivity film that blocks conduction.
6. Optimization of the Sintering Profile
6.1 Maximum Sintering Temperature
Sintering is the most decisive processing step for (Ba,Pb)TiO₃-based PTCR ceramics. I prepared samples with Ca doping between 6 mol% and 8 mol% and changed only the maximum sintering temperature from 1200 °C to 1320 °C. For the 8 mol% Ca formulation, the lowest firing temperature produced poorly semiconducted ceramics with R₂₅ equal to several megaohms. The breakdown voltage at 1262 °C was almost 942 V, whereas an increase to 1280 or 1300 °C caused exaggerated grain growth, reduction of the grain-boundary volume, and a sudden decline of voltage endurance.
| Maximum temperature (°C) | Ca content 8 mol%: R₂₅ (Ω) | Breakdown voltage (V) |
|---|---|---|
| 1200 | 2.55×10⁶ | >1100 |
| 1240 | 3287 | 425 |
| 1250 | 7007 | 739 |
| 1262 | 4991 | 942 |
| 1280 | 1210 | 633 |
| 1300 | 661 | 373 |
| 1320 | 173 | 373 |
When plotted against the sintering temperature, both the NTC parameter γ and the PTCR jump ratio β displayed an “M” shape. I interpreted this by considering calcium segregation and dissolution. At low sintering temperatures, calcium segregates strongly at the grain boundaries and forms acceptor states, which raises the barrier height. As the temperature increases, calcium dissolves into the grains, boundary barrier height decreases, and the NTC tail becomes less pronounced. At intermediate high temperature, excessive calcium begins to precipitate as a secondary titanate layer, creating an additional interfacial barrier that changes the resistivity ratio again. Above 1300 °C, abnormal grain growth destroys the fine-grained network and both PTC and voltage properties collapse. Therefore, the maximum sintering temperature must be controlled rather carefully around 1262 °C in order to combine high breakdown voltage, low R₂₅/Rmin, and sufficient PTCR jump ratio for electric-vehicle applications.
6.2 Heating Rate
I also varied the heating rate in the high-temperature section from 7.07 °C/min to 14.13 °C/min. The colour of the finished ceramics changed gradually from dark blue-grey to yellow, indicating incomplete semiconducting conversion when the heating rate was too high. The room-temperature resistance monotonically increased with the heating rate, while the NTC tail became weak. On the other hand, the voltage endurance decreased continuously with increasing heating rate. A moderate heating rate of 7.07 °C/min yielded a well-developed microstructure and the best breakdown value of 942 V.
| Heating rate (°C/min) | R₂₅ (Ω) | R₂₅/Rmin | β | Breakdown voltage (V) |
|---|---|---|---|---|
| 7.07 | 4991 | 0.93 | 4.19 | 942 |
| 8.48 | 56500 | 1.25 | 2.16 | 845 |
| 10.60 | 126900 | 1.08 | 2.32 | 739 |
| 14.13 | 299800 | 1.00 | 2.41 | 686 |
6.3 Electrode Firing Temperature
Surface metallization is not usually discussed in depth, but my experiments demonstrated that the electrode bake-out temperature strongly influences the apparent PTC characteristics. I screen-printed aluminium electrodes and fired identical ceramic bodies at 650 °C, 680 °C, and 700 °C. The room-temperature resistance was nearly unchanged, yet the maximum resistance Rmax, the jump ratio, and the breakdown voltage all degraded at higher firing temperatures. The most likely reason is that aluminium starts to melt just above 660 °C and can partially infiltrate the porous ceramic, reducing the effective electrode thickness and creating local ohmic inhomogeneities. The sample fired at 650 °C showed the strongest PTC jump and the highest breakdown voltage, above 800 V. Thus, the optimal electrode baking temperature is 650 °C.
7. Integrated Ceramic Performance
Finally, I selected compositions combining the best features of the CAST system and refined sintering conditions. Multiple batches were produced with room-temperature resistance between 1000 Ω and 5000 Ω, γ below 1.5, breakdown voltage greater than 800 V, and a voltage–power curve in which the maximum and minimum power did not exceed the allowed ratio around the rated working point of 350 V. Table 7 shows a representative selection of these optimized elements.
| Sample designation | R₂₅ (Ω) | β | R₂₅/Rmin | Heating rate (°C/min) | Breakdown voltage (V) |
|---|---|---|---|---|---|
| C8Si2.6Al2-T1262 | 1763 | 4.19 | 1.03 | 7.07 | 942 |
| C8Si2.0Al2-T1262 | 3004 | 3.81 | 1.43 | 7.07 | 866 |
| C8Si2.6Al2.6-T1262 | 2471 | 3.66 | 1.25 | 8.48 | 845 |
| C8Si3.2Al3.2-T1262 | 3151 | 3.72 | 1.39 | 7.07 | 897 |
| C8Si2.6Al2.6-T1280 | 1421 | 3.84 | 1.39 | 8.48 | 950 |
| C8Si3.2Al2.6-T1280 | 1345 | 3.71 | 1.35 | 8.48 | 896 |
The power-voltage curves of these elements were sufficiently flat in the range of 220–400 V. Since electric-vehicle bus voltages may fluctuate when batteries are charged at high power, a ceramic heater that tolerates more than 800 V provides a large safety margin. The absence of a pronounced NTC tail, together with the high breakdown strength, makes the CAST-compounded BPT ceramic directly applicable to current electric-vehicle heating systems without an additional controller.
8. Conclusion
In this thesis I prepared lead-containing barium titanate PTCR ceramics with the two-step solid-state reaction method and systematically investigated the influence of CaCO₃, Al₂O₃, SiO₂, and sintering variables on the electrical performance. The main conclusions are as follows:
- Calcium was shown to be the decisive component for suppressing the sub-Curie NTC effect. When Ca content was raised to 8 mol%, the R₂₅/Rmin parameter dropped below unity at a sintering temperature of 1262 °C, meaning that the resistance below the Curie point never fell below the room-temperature resistance. The same composition preserved a PTCR jump ratio of 4.19 and breakdown voltage above 890 V.
- The combined effects of calcium, aluminium, and silicon form a synergistic CAST doping system. Calcium reduces the NTC tail by partial occupation of the titanium site and interruption of the Ti³⁺–Ti⁴⁺ hopping paths; aluminium raises the resistance temperature coefficient and strengthens the PTC jump by acceptor compensation; silica improves liquid-phase sintering, homogenizes the microstructure, and markedly increases voltage endurance through the formation of grain-boundary secondary phases.
- The maximum sintering temperature and heating rate must be carefully tuned. A firing temperature near 1262 °C and a heating rate around 7 °C/min produced the most favourable combination of semiconducting behaviour, high β, low γ, and high breakdown voltage.
- The best electrode firing temperature for aluminium contacts was 650 °C. Higher electrode baking temperatures caused partial melting of the aluminium film and reduced the apparent PTCR jump and voltage endurance.
Based on this work, a new generation of high-reliability PTCR ceramics is now available for the electric-vehicle thermal-management industry. The elements operate over a wide temperature range from -40 °C to above 120 °C and over a wide voltage range without an auxiliary controller. They start rapidly, draw almost no inrush current, and withstand voltages far above the rated bus voltage of an electric vehicle. Further work should focus on the long-term stability of the electrode contact under thermal cycling, the lifetime behaviour under high-voltage d.c. bias, and the elimination of lead from the material for environmental reasons. Nevertheless, the CAST-based BPT ceramic already represents a substantial improvement over traditional PTCR materials and offers an economically attractive solution for the next generation of electric-vehicle heating modules.