You've probably stared at a DSC curve and wondered why your hard segment melting peak sits where it does. Still, maybe you're tweaking a formulation and the numbers aren't matching the literature. Or you're just trying to explain to your boss why switching from BDO to hexanediol shifted everything by fifteen degrees.
Here's the thing — most papers give you a single number. "MDI-HDO hard segments melt at 180°C.That said, " But that number is a lie. Or at best, a half-truth.
What Is MDI-Hexanediol Polyurethane Hard Segment Melting Temperature
Let's start with what we're actually measuring. They form crystalline or pseudo-crystalline domains through hydrogen bonding and aromatic stacking. When you synthesize a polyurethane from MDI and 1,6-hexanediol (HDO), the hard segments — those MDI-HDO-MDI sequences — want to pack together. The melting temperature (Tm) is where those ordered domains fall apart.
But here's what nobody puts in the abstract: that Tm isn't a fixed property of the chemistry. It's a property of your specific morphology*.
MDI is 4,4'-methylenebis(phenyl isocyanate) — two phenyl rings linked by a methylene bridge, each bearing an isocyanate group. Symmetric. HDO is a linear C6 diol. Regular. React them in the right ratio and you get a hard segment repeat unit with a molecular weight around 382 g/mol. In theory, it crystallizes beautifully.
In practice? So your Tm could be anywhere from 160°C to 210°C. I've seen both in the same lab, same batch of MDI, different prep methods.
The difference between Tm and hard segment dissociation
This trips people up constantly. But then there's the order-disorder transition* where hydrogen-bonded domains lose registry but don't fully dissolve. Even so, there's the true* melting point — crystals becoming amorphous melt. And then there's the soft segment phase mixing* temperature where everything blurs together.
DSC shows an endotherm. If you're publishing, you need to specify which technique, which heating rate, which thermal history. Because of that, they're rarely the same temperature. Which means dMA shows a tan delta peak. If you're formulating, you need to know which one correlates with your property of interest.
This part deserves a bit more attention than it usually gets.
Why It Matters / Why People Care
You care because hard segment Tm sets your processing window. It sets your upper use temperature. It controls creep, compression set, and whether your part fails at 80°C or 120°C.
Processing window reality check
Extrusion? You need to process above* the hard segment Tm but below* degradation. That's tight. Injection molding? If your Tm is 205°C, you have a 15-degree window. MDI-HDO systems degrade around 220–230°C. Screw speed, residence time, moisture — everything matters more.
I've watched a production run go from perfect parts to brittle scrap because the barrel temperature drifted 8°C. The operator didn't know the hard segment Tm was that close to the degradation onset. Nobody had told them.
Mechanical properties above and below Tm
Below Tm: hard domains act as physical crosslinks. High modulus, good creep resistance, shape memory. Above Tm: those crosslinks vanish. You get viscous flow. The transition isn't always sharp — broad peaks mean a gradual property drop over 20–30°C.
For thermoplastic polyurethanes (TPUs), this is the switching temperature. Still, for thermosets with low hard segment content, you might never see a distinct Tm — just a broad shoulder. Know which system you have.
How It Works — The Structural Factors That Move Tm
This is where the real formulation work lives. But every variable below shifts your melting temperature. Some you control. Some you don't.
Hard segment length and distribution
Short hard segments (low DP) melt lower. There's a critical length — roughly 3–4 repeat units — below which you don't get stable crystals at all. You get hydrogen-bonded clusters with no long-range order. In practice, no sharp Tm. Just a glass transition that shifts with hard segment content.
But here's the kicker: distribution* matters more than average length. A prepolymer route with controlled chain extension gives you a sharp one. Now, same chemistry. A Poisson distribution from one-shot polymerization gives you a broad Tm. Different morphology.
Soft segment molecular weight and chemistry
This is the variable everyone forgets. Lower Tm. Your soft segment dilutes* the hard segments. PTMEG 2000 vs PTMEG 650 — the longer soft segment spaces hard domains further apart. Even so, less hard-hard contact. Sometimes 20–30°C lower.
Want to learn more? We recommend how do you measure the density of a liquid and jobs you can get with a chemistry degree for further reading.
And the chemistry: polyether vs polyester. But PPG's methyl groups create steric hindrance at the interface. So polyethers (PTMEG, PPG) generally don't. Think about it: polyester soft segments (PCL, PBA) can co-crystallize or interfere with hard segment packing. PTMEG's linearity lets hard segments pack tighter.
I once switched a formulation from PTMEG 1000 to PPG 1000 — same Mn, same hard segment content. That said, the methyl groups disrupted the hard domain interface. Worth adding: tm dropped 18°C. Took me three weeks to prove it wasn't moisture.
Hard segment content (hard segment weight fraction)
More hard segment = more hard-hard contact = higher Tm. No continuous network. Below ~25% hard segment, domains are isolated. Tm is depressed and broad. And there's a percolation threshold. But it's not linear. Above ~40%, you get connected hard domains and a sharper, higher Tm.
The relationship flattens out around 50–60%. After that, you're limited by crystal perfection, not domain connectivity.
Chain extender ratio and stoichiometry
Off-ratio formulations kill Tm. Excess diol gives you hydroxyl end-caps on hard segments — they disrupt packing. Excess MDI gives you isocyanate end-caps that can trimerize to isocyanurates (great for thermal stability, terrible for sharp melting).
Target 1.That said, 00 ± 0. 02 NCO:OH. Analyze your residuals. If you're seeing a broad, low Tm, check your titration data before blaming the raw materials.
Thermal history — the variable you can't ignore
This is the biggest source of "my data doesn't match the literature."
Anneal a sample at 120°C for 2 hours? In practice, tm drops 15°C and broadens. Tm goes up 10°C. Consider this: age at room temperature for six months? On the flip side, quench from the melt? Physical aging increases domain order — Tm creeps up.
Always report your thermal history. Always condition your samples the same way before testing. And for the love of polymer science, don't compare a compression-molded plaque to an injection-molded bar without acknowledging the difference.
Common Mistakes / What Most People Get Wrong
Treating literature Tm as a specification
"Paper X says 185°C. My material melts at 172°C. Something's wrong.
No. Because of that, your material is different. Different Mn soft segment. Because of that, different chain extender ratio. Different thermal history. Different DSC heating rate. Literature values are reference points*, not acceptance criteria.
Confusing hard segment Tm with soft segment Tm
PTMEG 2000 melts around 45–55°C. PCL melts around 55–60°C. If you see
a melting endotherm around 45–55°C in your DSC trace, it’s likely the soft segment—not the hard segment. Always check both Tm values and their relative intensities. A low soft segment Tm can mask a high hard segment Tm if the soft phase melts too quickly. Hard segments in TPUs typically melt above 150°C. And don’t forget: the glass transition temperature (Tg) of the soft segment affects the overall material behavior below Tm, but it doesn’t influence the melting behavior directly.
Another pitfall: assuming all hard segments are equal. Polycaprolactone (PCL) hard segments melt around 60°C, while aromatic-based hard segments (e.Consider this: g. Think about it: , from MDI or TDI) can exceed 200°C. The same TPU with different hard segment chemistries will exhibit wildly different Tm values. If you’re comparing formulations, ensure the hard segment type is identical.
Lastly, don’t overlook the role of processing temperature. On the flip side, conversely, underheating can leave residual amorphous regions that depress Tm. Plus, heating the melt above Tm during processing can cause partial crystallization, leading to a higher measured Tm in the final part. Always match your processing temperature to the material’s thermal requirements.
So, to summarize, melting temperature in TPUs is a nuanced interplay of chemistry, formulation, processing, and measurement. There’s no single factor that dictates Tm—it’s the cumulative effect of all these variables. That said, whether you’re optimizing a formulation or troubleshooting a failure, start by mapping out every parameter: soft segment type, hard segment content, chain extender stoichiometry, thermal history, and testing protocol. On the flip side, cross-reference your data with literature values, but treat them as guides, not gospel. And remember: a 10°C shift in Tm can mean the difference between a material that performs or fails. Master the variables, and you’ll master the melt.
