Capacitor-grade polypropylene film is selected on the basis of two measurable quantities: the dielectric loss tangent at the intended operating frequency and the short-term breakdown field under standardized electrodes. These quantities are not independent; both are governed by the isotacticity distribution of the base resin, the concentration of catalyst residues and ash, the quench rate of the cast sheet, and the sequential biaxial orientation ratios imposed on the film line. Commercially qualified capacitor-grade resins are typically specified with isotactic index above 96%, ash content below 30 mg/kg, and chlorine residues below 5 mg/kg; these limits are used by polymer suppliers to reduce ionic conduction and high-field dielectric loss. The dielectric loss tangent of well-processed biaxially oriented polypropylene film at 1 kHz and 23 °C is normally reported between 0.0002 and 0.0005 when tested according to IEC 62631-2-1 or ASTM D150-18. The short-term DC breakdown field of films in the 3–8 µm thickness range is commonly reported as 180 kV/mm to 300 kV/mm when measured with cylindrical edge-defined electrodes at a ramp of 500 V/s under ASTM D149-20 or IEC 60243-1. These values are, however, thickness-, area-, and humidity-dependent; a single-point comparison between film grades is insufficient for capacitor design.

On production-scale sequential tenter lines, the film is extruded through a flat die, quenched on a chill roll, reheated, stretched in the machine direction, and then stretched again in the transverse direction. Melt pump pressure oscillations and die-lip temperature nonuniformities generate thickness bands that are detected by beta gauges; a thickness variation of ±0.2 µm across the web translates into local electric field enhancement because the applied voltage divides across a thinner region. The dielectric loss map across the web is also affected by thermal history: edges of the cast sheet quench faster than the center, producing lower crystallinity and different β-phase populations if the extrusion temperature is not tightly controlled. Production campaigns where edge trim is not removed cleanly from the quench system and re-enters the melt are avoided in capacitor-grade BOPP because gel contamination creates point defects with reduced breakdown field. These process interactions explain why capacitor-grade film cannot be selected solely from resin data; the final film morphology must be qualified by destructive breakdown tests and dissipation factor scans across the web. The film must also meet thickness and defect requirements of IEC 60674-3-4, which is the relevant specification for biaxially oriented polypropylene film intended for capacitor use.

What Minimum Breakdown Field Is Required for DC-Link Capacitors in 800 V Traction Inverters?

For an 800 V DC bus in an automotive traction inverter, the DC-link capacitor is exposed to a superimposed AC ripple at the switching frequency of the inverter, typically 4 kHz to 30 kHz depending on silicon carbide or IGBT module technology. The film in this application is usually 2.5–5 µm thick, and the capacitor is designed with two or more metallized film layers in series across the bus. The short-term breakdown field of the individual film is not the sole acceptance criterion; the long-term design electric stress is commonly limited to 120–200 V/µm at maximum rated voltage and maximum hot-spot temperature. This derating is required because the breakdown strength of polypropylene decreases as temperature rises and because the metallized electrode system is designed to self-heal repeatedly. A dielectric loss tangent of 0.0002 at 1 kHz can still produce appreciable heat because the reactive power density in a DC-link capacitor is large. The dielectric heating per unit volume is proportional to 2πf ε₀ ε_r E_rms² tanδ, where ε_r for polypropylene is 2.2–2.3. In a high-ripple bus, the equivalent series resistance from the end-spray, electrode segmentation, and film loss must be low enough to keep the hot spot below 105 °C, which is the common continuous temperature limit in manufacturer datasheets. Qualification of DC-link film is therefore performed with ripple current tests, thermal stability tests, and voltage endurance tests under IEC 61071 or automotive OEM specifications aligned to AEC-Q200. Published data for specific film thickness and segmentation configurations is limited, but the design rule of selecting a film with a short-term breakdown field at least 2–3 times the maximum working stress is widely applied in power electronic capacitor design. Production-scale failure analysis of DC-link capacitors shows that hot-spot failures are more often associated with localized electrode oxidation near the end-spray contact than with bulk dielectric failure. The film selection must therefore include a surface specification, typically a low extractable level and no slip aid, because end-spray adhesion to the exposed film edge is sensitive to surface contamination. Batch-to-batch variation in surface roughness of 0.02–0.08 µm Ra is sufficient to alter the end-spray contact resistance and the resulting local temperature under 100 A rms ripple currents.

Where the application is a continuous AC motor-run or power-factor correction capacitor, the breakdown field of the film is not the only selection pressure: the capacitor may be connected across a sinusoidal line voltage for 50 Hz or 60 Hz for more than 10,000 h per year, and the metallized electrode is exposed to atmospheric oxygen and humidity through the package. In these designs, film thickness is usually 5–10 µm and the working electric stress is held near 60–100 V/µm, substantially below the short-term breakdown field, because the requirement is not immediate withstand strength but long-term clearing stability. The film specified for these capacitors must have a dielectric loss tangent below 0.0005 at 50 Hz and 23 °C and a volume resistivity above 1×10^14 Ω·m when measured under IEC 60093; otherwise self-discharge and ionic conduction produce local thermal gradients that accelerate oxidation of the aluminum or zinc-aluminum electrode. Film surface quality is equally important. Capacitor grades avoid migratory slip additives because erucamide or oleamide blooms can move to the film surface during storage and alter the metallization adhesion, increasing the measured loss after temperature cycling. The metallized electrode is typically segmented with fuse structures, and the film must have uniform thickness and low gel content so that clearing events do not cascade into a thermal runaway. Tests according to IEC 60384-17 for AC and pulse capacitors, or IEC 60384-14 for electromagnetic interference suppression capacitors, use a combination of voltage endurance, impulse, and active flammability; the film is therefore selected not only by dielectric loss and breakdown field but also by its ability to survive a defined number of self-healing events without excessive capacitance loss. Production experience with AC capacitor banks in humid environments shows that exposed end-spray zinc layers are prone to contact oxidation if the film surface contains more than trace levels of polar low-molecular-weight species. For this reason, capacitor-grade film suppliers specify low extractables and avoid corona treatment after orientation; corona treatment raises surface energy for metallization but can introduce polar groups that increase dielectric loss at elevated humidity. The selection balance in motor-run applications therefore shifts toward low surface polarity and high surface smoothness rather than maximum dielectric strength.

When Surface Roughness and Dielectric Loss Interact in High-Frequency Resonant Capacitors

In resonant inverters and induction heating converters operating between 100 kHz and 1 MHz, the intrinsic dielectric loss of polypropylene is so low that the measured dissipation factor of the finished capacitor is usually dominated by the electrode system. A capacitor-grade BOPP film with a tan δ of 0.0003 at 1 kHz may exhibit an apparent loss tangent of 0.001–0.005 at 500 kHz after metallization, not because the polymer has changed, but because the metallized electrode has sheet resistance and the contact between electrode segments and end-spray creates a series resistance. The effective loss tangent of a capacitor can be expressed in the form tanδ_eff = tanδ_dielectric + R_s C ω, where R_s is the equivalent series resistance associated with the metallization and contacts. Because the dielectric term is small, film selection for high-frequency capacitors is driven by the surface roughness, surface energy, and surface cleanliness that control the morphology and sheet resistance of the vacuum-deposited electrode. The deposited aluminum or zinc-aluminum layer has a sheet resistance typically between 2 Ω/sq and 8 Ω/sq for AC capacitors. A film with an average surface roughness Ra below 0.04 µm and a low coefficient of friction is preferred because the deposited metal layer is more uniform and the current distribution across the electrode is more homogeneous. However, an excessively smooth film can develop winding slippage and air entrapment, which increases partial discharge and reduces capacitance stability. In practice, the roughness specification is a compromise: for high-frequency resonant capacitors, the film is often specified with Ra between 0.02 µm and 0.05 µm, measured by optical profilometry according to ISO 25178 or equivalent. The dielectric loss measurement under IEC 62631-2-1 at 1 kHz is therefore not sufficient; the film must be characterized at the intended switching frequency after a standard metallization process to separate polymer loss from electrode loss. Without this separation, a capacitor manufacturer may reject a film that has excellent dielectric loss but poorly controlled surface roughness, or accept a film with low roughness that develops high end-spray contact resistance after thermal cycling. Current crowding at segment edges and at the contact areas between overlapping metallized layers becomes severe above 300 kHz; therefore the selection process includes a high-frequency ESR measurement on a wound capacitor sample, not merely a film dissipation factor test. The film’s low dielectric loss remains an enabler, but breakdown field is not the main failure boundary in this frequency range; instead, the thermal limit is reached by ohmic heating in the electrodes. This is a key distinction from DC-link selection: a high-frequency capacitor may operate at a dielectric stress below 50 V/µm and still fail thermally because the electrode loss generates a hot spot at the center of the winding. Production experience with induction heating capacitors shows that the axial temperature gradient in a non-segmented winding can exceed 20 K between the core and the end-spray, with the core being hotter; the film must therefore have stable shrinkage characteristics at 105 °C and low moisture absorption so that dimensional changes do not increase electrode resistance during service.

Thermal Runaway Boundaries in Metallized Self-Healing Film

Self-healing is the mechanism by which a metallized polypropylene capacitor clears a localized breakdown without failing short-circuit. The process evaporates or oxidizes the thin metallization around the defect, but the clearing event generates a localized plasma and a small amount of gas. If the available energy per clearing is too high, the gas pressure and thermal pulse can damage surrounding film and create new weak points, leading to a cascade. The energy available for a clearing event is related to the local capacitance, the square of the voltage, and the segmentation geometry. This is why high-voltage metallized polypropylene capacitors are divided into many electrode segments connected by fuse links; segmentation limits the energy to a fraction of the total stored energy. Film selection enters because the dielectric loss tangent determines continuous heating, while the breakdown field and thickness profile determine the frequency and spatial distribution of self-healing events. If a film has a broad distribution of breakdown strengths, clearing events will cluster in thin or contaminated regions; repeated clearing at one location raises the local temperature and reduces the film’s ability to survive further stress. A film with a more uniform thickness and a higher Weibull modulus exhibits more distributed clearing and lower probability of thermal runaway. Published Weibull modulus values for high-quality capacitor-grade PP film are often reported in the range 8–15, but the exact value depends on electrode area, voltage ramp, and defect density. The continuous hot-spot temperature must remain below 105 °C; above this limit, polypropylene softens, the metallization adhesion decreases, and self-healing events become larger. Thermal stability tests specified in IEC 61071 and steady-state life tests under IEC 60384-16 are used to verify that the selected film can sustain the rated voltage at maximum temperature without progressive capacitance loss. The dielectric loss of the film is a critical input because the heat generation rate scales with tanδ, and even a modest increase from 0.0002 to 0.0004 doubles the dielectric loss contribution at the same voltage and frequency. In a densely packed capacitor bank with limited airflow, that increase can shift the thermal equilibrium from stable to unstable. Production-scale measurements on segmented metallized PP film show that capacitance loss during a 1,000 h endurance test at 1.2 times rated voltage and 85 °C is not uniform across the winding; the central layers degrade first if the film has high internal haze or if the metallized electrode is too thick. The selected film must therefore be evaluated not only as an unmetalized film but also after vacuum metallization. The metallization process itself, which operates at pressures below 1×10^-4 mbar, can modify the film surface by electron or ion bombardment; an unsuitable film may show increased surface roughness or reduced contact angle, changing the sheet resistance. This is one reason why a film qualification program includes measurements of surface energy, roughness, and dielectric strength before and after metallization. Published data for the exact relationship between metallization process parameters and dielectric loss of capacitor-grade PP film is limited, so qualification is typically performed on a pilot winding rather than on a flat film coupon.

Breakdown Field Selection in High-Voltage Energy Storage and Pulse Capacitors

High-voltage energy storage capacitors for pulsed lasers, defibrillators, and electromagnetic forming require a different selection logic because the film operates at high electric stress for short discharge times and must withstand rapid polarity reversals. Thicknesses in these capacitors are often 5–12 µm, and the working stress can be 120–180 V/µm, which is closer to the short-term breakdown field than in AC or DC-link designs. The key film properties are the short-term DC breakdown field, the thickness distribution, and the density of gel particles or other dielectric defects. Because the discharge time is short, the dielectric loss contribution to average heating is smaller than in continuous ripple applications, but the fast charge-discharge cycles create mechanical stress in the film and electrode layers. The breakdown field must be measured under IEC 60243-1 with a ramp rate consistent with the fast charge application, and the data must be analyzed statistically rather than as a single mean value. A capacitor with a mean breakdown field of 220 kV/mm but a low Weibull modulus may be less reliable than one with a mean of 200 kV/mm and a tighter distribution. Partial discharge inception voltage at the electrode edge and at the free margin is another selection criterion; a film with poor surface smoothness may trap air at the electrode edge and generate partial discharges at high voltage. The dielectric loss of the film at 1 kHz is often a secondary consideration, but the loss at higher harmonics of a fast discharge can contribute to local heating at the electrode edges. Specifications for high-voltage pulse capacitors may reference IEC 61881 for railway applications or IEC 60384-16 for DC capacitors; the film is often tested with a minimum breakdown voltage per unit thickness, a maximum dissipation factor, and a minimum insulation resistance after temperature cycling. Because high-voltage designs use thicker films and higher stress, the selection process must pay particular attention to film winding defects such as telescoping, wrinkles, and gauge bands. A thickness variation of ±0.1 µm across a 10 µm film creates less relative field enhancement than the same variation across a 3 µm film, but the stored energy is higher and a clearing event can be more damaging. Production experience with pulse capacitor windings shows that failures often initiate at hard wrinkles or at particle inclusions, not at the average thickness location. The film is therefore inspected with on-line optical detection systems and a sample of film from each roll is destructively tested for dielectric strength at multiple points across the web.

Processing Windows that Determine Film Dielectric Strength

The dielectric strength of the final BOPP film is determined less by the resin’s intrinsic properties than by the thermal and mechanical sequence that converts the cast sheet into a highly oriented film. In sequential stretching, the cast sheet is first reheated and stretched in the machine direction, typically at 135–150 °C, and then in the transverse direction, typically at 155–175 °C. These temperature windows are narrow; deviations of more than ±5 °C across the web produce uneven orientation and localized variations in thickness, crystallinity, and dielectric loss. If the machine-direction orientation temperature is too low, the film surface may develop microvoids from excessive stress, reducing breakdown strength. If the temperature is too high, the orientation is incomplete, and the film lacks the dense, microfibrillar structure that suppresses charge injection at high fields. The orientation ratios are also important: machine-direction ratios of 4:1 to 5:1 and transverse ratios of 8:1 to 10:1 are used in many capacitor-grade films, but exact values vary by line configuration and resin. Production lines use a melt pump between the extruder and the flat die to dampen pressure fluctuations; the extruder is typically a single-screw machine with a screw L/D ratio of 25:1 or greater and a filter mesh pack downstream of the screw to remove gels. Melt temperature at the die is usually controlled in a band near 230–260 °C, depending on the resin melt flow rate. If the melt temperature is too high, oxidative degradation increases carbonyl groups and dielectric loss; if too low, the cast sheet may contain unmelted resin particles that become breakdown weak points. The chill roll temperature also influences the crystalline morphology. A quench temperature around 20–40 °C produces a small spherulite or mesomorphic structure that is easier to orient later; slower cooling at elevated temperatures can grow large spherulites that leave boundary regions with lower breakdown strength after orientation. These process parameters are not selected independently; a resin with an MFI of 2.0–3.5 g/10 min under ISO 1133-1 at 230 °C and 2.16 kg may require a lower extrusion temperature than a high-MFI resin, but the orientation temperatures must remain within the narrow window that produces uniform film. The resulting film is then wound into large mill rolls and later slit to the required capacitor width. During slitting, edge damage and debris must be minimized because the film edge becomes the dielectric margin in the capacitor winding. A single hard blade or dull knife can generate enough edge debris to reduce the partial discharge inception voltage of the finished capacitor. For this reason, capacitor-grade film is slit with clean, sharp knife blades or laser slitting, and the slit edges are inspected before metallization.

Resin selection and process selection are coupled: high isotacticity and low ash are necessary but not sufficient. A resin with isotacticity above 96% and ash below 30 mg/kg can still produce poor capacitor film if the cast sheet is quenched nonuniformly or if the transverse orientation oven has a temperature spread greater than ±3 °C. On production-scale tenter lines, edge ducts and center ducts are controlled separately, and the measured film thickness profile often shows thinner edges if the transverse stretch is not balanced with the clip speed. The breakdown field map across the web therefore reveals process health better than a single sample from the film roll. A film roll that passes a 200 kV/mm average breakdown test but contains a region with 120 kV/mm breakdown at a known edge band is unsuitable for high-stress capacitors, even if the dissipation factor is uniformly below 0.0003. This is why capacitor manufacturers require destructive testing of multiple samples across the web and along the roll length, and why film suppliers must provide lot-specific data rather than a general datasheet.