Polycarbonate granules have no melting point. It is an amorphous polymer, which means it does not melt — it softens progressively as it passes through its glass transition at about 145–150 °C, and there is no latent heat, no sharp transition, and no thermodynamic melting temperature.
The number that search engines and old textbooks quote, around 220–230 °C, is not a melting point; it is roughly where an unloaded PC sample becomes soft enough to deform visibly, which is a different measurement with a different meaning. This matters practically because people run PC at 240 °C on the assumption that they are 20 °C above its “melting point”, and then wonder why thin walls do not fill and why the parts are brittle. The correct processing window is 280–310 °C, and the reason is viscosity, not melting. On the drying side, the trap runs the other way: drying too hot is as damaging as not drying enough, and the correct setpoint is 120 °C with desiccant air, not 140 °C for speed.
1. Polycarbonate Is Amorphous, So “Melting Point” Is a Category Error
Bisphenol-A polycarbonate is an amorphous thermoplastic. Its chains are entangled but not ordered into lamellae, so there is no crystal lattice to melt. Heating it produces a glass transition — the point at which the chains gain enough thermal energy to move cooperatively — and then progressive softening, and eventually thermal degradation.
| Transition | Temperature | What Happens | Measurement |
| Glass transition (Tg) | 145–150 °C | Chain segments gain mobility; modulus drops by orders of magnitude | DSC, ISO 11357-2 |
| Softening (Vicat) | 150–160 °C | Defined penetration under a specified load | ISO 306 |
| Heat deflection (HDT) | 125–135 °C at 1.8 MPa | Deflection under a specified bending load | ISO 75-2 |
| Processing window | 280–310 °C | Viscosity low enough to fill and pack | Practical |
| Thermal degradation onset | ~340–350 °C | Chain scission, yellowing, gas evolution | TGA |
Notice the gap. Tg is 150 °C and the processing window starts at 280 °C. That gap is not a melting range — it is the range over which viscosity falls from “solid” to “injection-mouldable”, and it spans 130 °C. In a semi-crystalline polymer like PA66 or PBT the equivalent gap is small, because melting happens at a specific temperature and viscosity drops abruptly. In PC it is gradual, which is why PC is processed so much hotter than its Tg and why the “melting point” question has no clean answer.
What the 220–230 °C number is: it is the temperature at which an unloaded or lightly loaded PC specimen deforms noticeably — close to the Vicat softening range under some conditions, and often confused with a “softening point” from an older test method. It is a real observation and a useless processing number. At 230 °C, PC has a viscosity in the range of 10⁴–10⁵ Pa·s, which is far too stiff to fill a thin wall at any reasonable injection pressure.
2. Why the Processing Window Is 280–310 °C
The window is set by two boundaries: fillability at the bottom and degradation at the top.
| Melt Temperature | Viscosity | Fill Capability | Degradation Risk |
| 240 °C | Very high | Thick sections only, short flow | Low |
| 260 °C | High | Moderate walls, short-to-medium flow | Low |
| 280 °C | Moderate | General purpose, thin walls feasible | Low |
| 290–300 °C | Low | Thin walls, long flow, optical parts | Moderate |
| 310 °C | Low | Excellent flow | Residence time critical |
| 320 °C+ | Very low | — | Rapid yellowing, molecular weight loss |
The practical rule: run as hot as the part requires and no hotter. A thick-walled, short-flow part at 280 °C will have better colour and better impact than the same part at 310 °C, because thermal history is the main enemy of PC properties. A thin-walled, long-flow part needs 300 °C, and the correct response is to accept that and control residence time rather than to run cold and increase injection pressure.
Running cold and increasing pressure does not work. Viscosity at 250 °C is roughly five to ten times that at 300 °C. You cannot buy that back with pressure — the machine’s pressure limit arrives long before the cavity fills, and the extra pressure generates shear heating that raises the local melt temperature anyway, so you get the worst of both: high pressure and hot spots.
3. What Happens When You Run PC Too Cold
This is a real and common failure, and its symptoms are misread as material problems.
| Symptom | Mechanism | What People Blame |
| Short shots on thin walls | Viscosity too high to fill before freeze | “The grade is too low flow” |
| High residual stress, cracking | Orientation frozen in under high pressure | “The material is brittle” |
| Weld lines that fail | Poor polymer diffusion at the interface | “Bad material” |
| Flow marks and jetting | High viscosity, unstable flow front | “Wrong grade” |
| Dimensional variation | Incomplete packing, variable cushion | “Inconsistent material” |
| Delamination at the surface | High shear stress at the gate | “Contamination” |
Every one of these is a temperature problem wearing a material problem’s clothes. The giveaway is that they cluster: if a part has short shots and high stress and weak weld lines, the common cause is almost always insufficient melt temperature or insufficient injection speed, not a material defect.
4. Drying: Why PC Must Be Dried
Polycarbonate hydrolyses. Water attacks the carbonate linkage and cleaves the chain, producing a phenol end group and a carboxylic acid end group. Two things make this worse than a simple chain scission:
It is autocatalytic. The carboxylic acid end group it produces catalyses further hydrolysis. Once started, the reaction accelerates, which is why a wet lot does not degrade linearly with moisture content — it degrades slowly, then rapidly.
The impact loss precedes the visible symptoms. Notched impact falls before tensile strength, before elongation, and before any visible defect. A part moulded from wet PC can pass a dimensional check and a tensile test and still fail a drop test, because the molecular weight has dropped below the threshold for rubber-particle cavitation.
| Moisture Content | Molecular Weight | Notched Impact | Visible Symptoms |
| ≤ 0.02% | Full | Full | None |
| 0.02–0.05% | Slight reduction | 10–20% loss | None |
| 0.05–0.10% | Moderate reduction | 30–50% loss | Slight splay, yellowing |
| > 0.10% | Severe reduction | 60–80% loss | Splay, silver streaks, brittleness |
The specification is ≤ 0.02% moisture, measured by Karl Fischer per ISO 15512, on the pellet as it enters the feed throat — not as it leaves the dryer.
5. The Drying Window: 120 °C, Not 140 °C
The instinct when a part has splay is to raise the drying temperature. That is the wrong move for PC, and it is why so many processors fight a problem they are making worse.
| Drying Temperature | Time to 0.02% | Risk |
| 100 °C | 6–8 h | Slow, but safe |
| 110 °C | 4–6 h | Safe |
| 120 °C | 3–4 h | The standard recommendation |
| 130 °C | 2–3 h | Oxidation begins, yellowing |
| 140 °C | 1.5–2 h | Significant oxidation, colour shift, branching |
| 160 °C | — | Pellet agglomeration, severe degradation |
Above about 130 °C, polycarbonate begins to oxidise in air. The oxidation produces yellowing and, at longer times, branching and crosslinking that raise viscosity and create gels. You are trading a moisture problem for an oxidation problem, and the oxidation problem is harder to detect because it does not produce splay — it produces a colour shift and occasional gel particles, which get blamed on the supplier.
The correct drying specification:
| Parameter | Value |
| Temperature | 120 °C |
| Time | 3–4 h (4 h for thick-walled or optical parts) |
| Air dew point | −20 °C or lower |
| Airflow | Per dryer capacity, typically 1.5–2.5 m³/h per kg/h throughput |
| Dryer type | Desiccant (not hot air) |
| Target moisture | ≤ 0.02% |
| Over-dry limit | Do not exceed 6 h at 120 °C |
Hot-air dryers do not work for PC. A hot-air oven at 120 °C with ambient humidity of 50% RH has a dew point around 10 °C — it will dry the surface and then re-equilibrate the pellet to whatever the ambient humidity supports, which is typically 0.05–0.15%. A desiccant dryer at −20 °C dew point is the only way to reach 0.02%.
6. Moisture Regain: The Hopper Problem
Dried PC does not stay dry. In an open hopper at 40% RH and 25 °C, PC regains moisture steadily, and the rate is fast enough to matter within a shift.
| Exposure | Approximate Moisture Regain | Consequence |
| 10 min in open hopper | +0.01% | Marginal |
| 30 min in open hopper | +0.03% | Over the limit |
| 2 h in open hopper | +0.08% | Splay and embrittlement |
| Overnight in an open bin | +0.15–0.25% | Unusable without re-drying |
The practical controls:
- Keep the dryer on the machine and the hopper sealed. A dried material in a heated hopper with dry air blanketing holds its moisture; the same material in an open hopper does not.
- Do not dry a whole shift’s material at the start. Dry to demand, or hold in a sealed dry hopper.
- Watch the weekend. Material left in an unsealed hopper over a shutdown needs re-drying, not a top-up.
- Regrind dries differently. Regrind has higher surface area and often higher initial moisture; it needs longer, and it should be dried separately or in a known blend ratio.
7. Measuring Moisture Properly
Three methods, three levels of usefulness.
| Method | Principle | Accuracy | Time | Use |
| Karl Fischer titration | Coulometric or volumetric water determination | ±0.005% | 15–30 min | The reference method, ISO 15512 |
| Halogen moisture analyser | Weight loss on heating | ±0.02% | 10–20 min | Routine incoming and in-process |
| Loss on drying (oven) | Weight loss at 120 °C | ±0.05% | 2 h | Crude, volatile loss confounds |
Karl Fischer is the specification method and the only one that measures water specifically. A halogen analyser measures total volatile loss, which includes any residual monomer or additive volatility, so it reads high on some grades. That is acceptable for routine control as long as you have correlated it against Karl Fischer for your specific grade — which takes an afternoon and is worth doing once.
The measurement point matters as much as the method: sample at the feed throat, not at the dryer outlet. Material that measures 0.01% at the dryer can be 0.04% by the time it reaches the screw if the conveying path is open to ambient air.
8. Drying Equipment
| Type | Dew Point Achievable | Suitability for PC | Notes |
| Hot air oven | Ambient dew point | Not suitable | Cannot reach 0.02% |
| Desiccant wheel dryer | −20 to −40 °C | Suitable | Standard choice |
| Compressed air dryer | −20 to −40 °C | Suitable | Needs clean, dry supply |
| Vacuum dryer | Very low | Suitable, fast | Higher capital cost, fast cycles |
| Infrared dryer | n/a | Not for bulk | Surface drying only |
| Inline crystalliser + dryer | n/a | Not needed | PC is amorphous, no crystallisation step |
Two equipment notes:
PC does not need a crystalliser. Unlike PET, which must be crystallised before drying to prevent pellet agglomeration, PC is amorphous and can go straight into a desiccant dryer. If a supplier tells you to crystallise PC, they are confusing it with PET — a confusion we address in our comparison of PBT and PET, where the crystallisation difference is the whole story.
Dryer sizing is usually wrong in the direction of too small. A dryer rated for 50 kg/h at 120 °C/−20 °C dew point will not deliver 50 kg/h at 0.02% if the incoming material is at 0.2% and the residence time is short. Size for the worst case, or accept longer residence.
9. Processing Parameters
| Parameter | Recommendation | Notes |
| Drying temperature | 120 °C | Not 140 °C |
| Drying time | 3–4 h | 4 h for optical or thick wall |
| Moisture target | ≤ 0.02% | Karl Fischer, ISO 15512 |
| Melt temperature | 280–310 °C | As hot as needed, no hotter |
| Feed zone | 260–280 °C | Gradual rise |
| Compression zone | 290–305 °C | |
| Metering zone | 290–310 °C | |
| Nozzle | 290–310 °C | Slightly below metering to avoid drool |
| Mould temperature | 80–110 °C | 100–120 °C for optical, low stress |
| Injection pressure | As required | High pressure indicates too-cold melt |
| Holding pressure | 40–70% of injection | |
| Back pressure | 0.5–1.5 MPa | Low; high back pressure adds shear heat |
| Screw speed | 40–80 rpm | Slow, to limit shear heating |
| Residence time | < 6 min | Critical for colour and impact |
| Cushion | 4–8 mm | Stable cushion indicates good packing |
| Ejection temperature | Below 130 °C | Above this, ejection stress is high |
Residence time is the parameter that is most often violated and least often measured. A 300 °C melt held for 15 minutes loses measurable molecular weight and yellows visibly. The calculation is simple: shot weight divided by (machine capacity × cycles per hour) gives the number of shots’ worth of material in the barrel, and that number times the cycle time is the residence time. If it exceeds 6 minutes, either the barrel is too large for the mould or the cycle is too slow — and both are fixable.
10. Verifying the Grade You Received
Before chasing a processing problem, confirm the material is what the label says. Five checks cover it, and they are all fast.
| Check | Standard | What It Confirms |
| MFR | ISO 1133, 300 °C/1.2 kg | Molecular weight band, regrind content |
| Density | ISO 1183 | Grade identity, filler presence |
| Glass transition | ISO 11357-2, DSC at 20 K/min | That it is PC and not a blend |
| Moisture | ISO 15512, Karl Fischer | Drying state at receipt |
| Notched Charpy | ISO 179-1eA | Whether the lot has already been degraded |
| Tensile properties | ISO 527-2 | Baseline for comparison against the TDS |
| Shrinkage | ISO 294-4 | Whether the lot matches the mould it was bought for |
A lot that passes all seven is a good lot. A lot that fails MFR and notched impact together has almost certainly been through an extra thermal pass, which is a supplier conversation rather than a process fix.
11. Troubleshooting
| Defect | Likely Cause | Check First |
| Silver streaks / splay | Moisture | Karl Fischer at feed throat |
| Brittle parts, low impact | Moisture or over-temperature | Moisture, then residence time |
| Yellow parts | Over-temperature, long residence, over-drying | Melt temp, residence, drying temp |
| Short shots | Melt too cold, insufficient speed | Raise melt temperature before pressure |
| Flow marks / jetting | Melt too cold, gate too small | Melt temperature, gate size |
| Weld line failure | Melt/mould too cold, gate position | Mould temperature, gate location |
| Black specks | Degraded material, worn barrel | Purge, inspect screw |
| Gels / unmelted particles | Low melt temperature, short residence | Melt temperature |
| Warpage | Uneven cooling, high packing | Mould temperature uniformity |
| Stress cracking in service | High moulded-in stress | Mould temperature, packing, annealing |
The pattern across the table: five of the ten defects are caused or worsened by running too cold, and three by running too hot or too wet. The window is real and it is narrower than most people assume, which is why the parameter table matters more for PC than for a forgiving commodity resin.
12. Frequently Asked Questions
What is the melting point of polycarbonate?
It does not have one. PC is amorphous, so it has a glass transition at about 145–150 °C rather than a melting point. The 220–230 °C figure often quoted is a softening observation, not a melting point, and using it as a processing reference leads to running the material far too cold. The practical processing window is 280–310 °C.
Why is the processing temperature so far above the glass transition?
Because the window is set by viscosity, not by a phase change. At 230 °C, PC viscosity is in the 10⁴–10⁵ Pa·s range, which cannot fill a thin wall at achievable injection pressures. At 290–300 °C it falls to a workable range. The 130 °C gap between Tg and processing temperature is normal for an amorphous engineering resin.
What moisture content should PC be dried to?
≤ 0.02%, measured by Karl Fischer per ISO 15512, sampled at the feed throat. Above 0.05% you lose 30–50% of notched impact with no visible defect, which is why wet PC fails drop tests while passing dimensional checks.
Can I dry PC at 140 °C to speed it up?
No. Above about 130 °C, PC oxidises in air, producing yellowing and eventually branching and gels. You trade a moisture problem for an oxidation problem that is harder to diagnose. Dry at 120 °C for 3–4 h with desiccant air at −20 °C dew point.
Do I need a crystalliser for PC?
No. PC is amorphous, so there is no crystallisation step before drying — that requirement belongs to PET. PC goes straight into a desiccant dryer.
Why do my parts have silver streaks even though the dryer says the material is dry?
Because the measurement point and the use point are different. Material at 0.01% at the dryer outlet can be 0.04% at the feed throat if the conveying path is open to ambient air. Sample at the feed throat, and check hopper seals and the conveying air supply.
How do I calculate residence time?
Shot weight divided by machine shot capacity gives the fraction of a shot in the barrel; multiply by the cycle time and invert appropriately, or more simply: (barrel capacity in shots) × (cycle time) = residence time. If it exceeds 6 minutes at 300 °C, use a smaller barrel or a faster cycle.
13. What We Can Send You
We supply PC across the melt flow range, from optical grades at 4 cm³/10 min to high-flow grades at 25 cm³/10 min, and we can tell you the drying and processing window for the specific grade rather than a family average. For a processing programme we can send:
- A processing guide for the specific grade — barrel profile, mould temperature, residence limit, drying specification
- DSC data giving Tg per ISO 11357-2, so you can verify the grade rather than assume it
- Initial moisture and drying curve data, so you can set your own dryer rather than guess
- Rheology at three temperatures, so you can predict fill at your wall thickness
- A paired trial against your incumbent, moulded at the same conditions
Send us the part, the machine, the wall thickness and the current defects, and we will tell you whether the problem is the material or the process — and in our experience it is the process more often than the purchase order suggests.
Related Reading
- Polycarbonate (PC): Properties, Grades and Applications
- Why Does Polycarbonate Yellow? Causes and Prevention
- PC Annealing and Residual Stress Relief
- PC/ABS Processing Guide: Drying, Melt Temperature and Mold Design
- PBT vs PET: Why Crystallization Rate Changes Everything
- PC for Medical Devices: Sterilization Compatibility