๐ก In This Article
- DSC analysis: How Tg can be identified from changes in heat capacity
- How to read Tg from a DSC curve: Why Tg is usually identified from a baseline shift rather than simply looking for a peak
- Why comparing DSC curves can be more informative than comparing Tg values alone
- Using DSC for quantitative evaluation of degree of cure
In the previous article, “Why Are There Different Analytical Techniques for Measuring Tg?“, I covered how Tg isn’t simply a single number but a reference point tied to changes in several material properties. DSC (Differential Scanning Calorimetry) is usually one of the first techniques a developer reaches for when evaluating Tg, because it gives a direct way to watch how a material’s heat flow and heat capacity change with temperature.
What Does a DSC Actually Measure?
DSC measures the difference in heat flow between a sample and a reference while temperature is controlled, typically during a programmed heating or cooling scan. When a material is heated without a significant thermal transition, its heat-flow behavior changes relatively smoothly. But when a physical or chemical transition occurs inside the material, the amount of heat required or released changes, and DSC captures that as a curve plotted against temperature.
That means a single DSC scan can carry information about several different thermal phenomena at once โ glass transition, crystallization, melting, and curing reactions. In that sense, DSC isn’t just telling us a temperature; it’s showing how the material responds to thermal energy as its temperature changes.
There are different DSC configurations, including heat-flux DSC and power-compensation DSC. For polymer and epoxy characterization, heat-flux DSC is the more common choice, though the right instrument setup and test conditions still depend on the material and the purpose of the measurement. Heating rate matters too โ 10ยฐC/min is a common choice in polymer characterization, but the appropriate rate should follow the applicable standard and test purpose. A faster heating rate can push the apparent glass-transition region toward a higher temperature, since the polymer has less time to respond to the changing thermal conditions, which is why heating rate and thermal history should always be part of the conversation when comparing DSC data.
Why Can DSC Detect Tg?
When an amorphous polymer moves from a glassy state toward a more rubbery one, the mobility of its polymer segments increases, and that change in molecular mobility comes with a change in heat capacity โ which is the principle DSC-based Tg measurement rests on. DSC picks up the change in heat flow that accompanies that shift in heat capacity.
Importantly, Tg usually doesn’t show up as a large, sharp peak the way melting or crystallization does. It tends to appear as a step-like change or shift in the baseline instead, and that’s one of the first things a developer needs to understand when reading a DSC curve.
A simplified view of a DSC curve
The exact appearance and direction of the curve depend on the instrument convention and whether the plot is presented as exo-up or exo-down. So whether the curve moves up or down isn’t the important part โ what matters is that the baseline changes around the glass-transition region.
How Is Tg Determined from a DSC Curve?
The Tg value a DSC report gives depends on the standard and data-analysis method being used. Common approaches identify characteristic points such as onset temperature, midpoint temperature, or other method-specific criteria. Midpoint is often used as the representative Tg value, but there’s no single universal definition that applies across every DSC measurement โ which is why two reports can occasionally give slightly different Tg numbers for the same material even when both measurements are technically valid.
The transition itself also spans a finite temperature range, so when developing an epoxy formulation, it’s often worth looking past the reported Tg number to the shape and width of the transition. A broad transition, for instance, might point to compositional or structural heterogeneity, overlapping transitions, differences in network structure, or other formulation- and history-related effects โ it shouldn’t automatically be read as evidence of just one specific cause. That’s another reason reading the whole curve tells you more than recording a single temperature.
In the Lab, We Looked at the Curve Before the Number
In actual development work, it’s not unusual to find two epoxy samples with similar reported Tg values but noticeably different DSC curves โ the baseline shift might differ in size, a residual curing exotherm might show up, a small unexpected peak might appear, the transition might be broad in one and sharp in the other, or a new thermal event might emerge after aging.
That pushes a developer toward a much more practical question than “what is the Tg?” โ it’s “why does this curve look different?” A numerical value tells you what was measured; the shape of the curve gives clues about what may have happened inside the material. That becomes especially useful when comparing samples before and after storage or aging.
Physical Aging Can Also Appear in the DSC Curve
Epoxy samples stored for a long time, particularly below their Tg, can sometimes show an endothermic feature near or just above the glass-transition region. This is often associated with physical aging or enthalpy relaxation โ after sitting below Tg for an extended period, the polymer network can gradually settle toward a lower-energy, more relaxed glassy state, and that stored relaxation shows up as an endothermic overshoot or peak during the next DSC heating scan.
This distinction matters because it shouldn’t automatically be read as chemical degradation or thermal decomposition. A small endothermic feature near Tg may simply reflect the material’s thermal history and physical relaxation, while a different thermal event at a higher temperature could indicate something else entirely. The location, shape, reproducibility, and relationship to the sample’s thermal history all need to be weighed together when interpreting an unexpected peak.
Compare the New Curve with the Previous One
A DSC curve becomes far more useful once it’s compared against another curve. In formulation development, I rarely looked at a newly obtained DSC curve in isolation โ I’d compare it against previous results after changing curing conditions, modifying the formulation ratio, adding or removing an additive, changing curing time, or finishing an aging test.
One of the simplest and most powerful methods is simply overlaying the curves and asking: how much did the baseline change? Did Tg shift? Did the residual exotherm shrink? Did a new peak appear? Did the transition get broader or sharper? Did the overall thermal behavior change after aging?
These changes can say a lot about what happened inside the material. A shift in Tg alongside a reduced residual cure exotherm, for instance, can suggest additional curing occurred during a thermal treatment, while a new thermal event appearing after aging might call for a completely different investigation. The direction and pattern of change can matter more than the absolute Tg value itself.
DSC Can Also Tell Us How Completely an Epoxy Has Cured
In epoxy development, DSC often gets used even more for evaluating degree of cure than for measuring Tg. When an epoxy sample isn’t fully cured, additional curing can occur during the DSC heating scan itself, and since curing is exothermic, that extra reaction shows up as a residual exothermic peak. A well-cured sample should show little or no residual curing exotherm, depending on the formulation and measurement conditions โ which makes DSC useful not just for Tg but for tracking the remaining cure reaction too.
More importantly, the residual heat of reaction can be quantified. If ฮHresidual is the residual heat of reaction measured from the partially cured sample and ฮHtotal is the total heat of reaction for the same formulation in an appropriate uncured reference state, degree of cure can be estimated as:
ฮฑ = ( 1 – ฮHresidual / ฮHtotal ) X 100 %
When developing a new epoxy formulation, this can be used to work out the curing time and temperature needed to reach a target degree of cure โ 95% or higher, say โ provided the reference and measurement procedures are properly controlled. The absolute value should always be read in the context of the formulation, reference sample, baseline treatment, and test method.
A Development Note: Comparing the 1st and 2nd Heating Scans
One common lab technique is comparing the 1st heating scan against the 2nd. The first heating reflects the sample’s current condition, including previous thermal history, residual moisture, physical aging, and incomplete curing. The sample is then heated through the relevant range and cooled under controlled conditions, and the second heating gives a measurement after much of the original thermal history has been erased โ often producing a more reproducible picture of the material’s thermal transition under the defined test conditions.
That said, I wouldn’t call the second-heating Tg the material’s “true Tg.” For reactive systems like epoxy, additional curing can happen during the first heating, and the thermal history built up during that first scan can itself affect the second. So the difference between the 1st and 2nd heating curves is valuable information in its own right โ if Tg shifts significantly upward in the second scan while a residual exotherm disappears, that tells you something real about the state of cure and thermal history of the original sample. The curve is telling us more than a single Tg number, again.
Tg Is a Number, but the DSC Curve Tells the Story
DSC is certainly useful for getting a Tg value, but from a formulation developer’s perspective, its real value runs much deeper. A DSC curve can show how the glass transition occurs, whether residual curing remains, whether physical aging has taken place, whether a new thermal event has appeared, and how the material has changed compared with an earlier formulation or condition.
That’s why I learned to look at the curve before the number. A number is useful for spec sheets and comparisons, but when something changes during formulation development, aging, or process optimization, the curve is usually what tells you why.
Coming Next: Reading a TMA Curve
DSC tells us about changes in heat flow and heat capacity. Another important technique, TMA (Thermomechanical Analysis), approaches Tg from a completely different angle by measuring dimensional change โ and the resulting curve can look very different from a DSC curve, even though both are probing the same glass-transition region.
In the next article, Reading a TMA Curve: Why Do We Measure TMA?, I’ll look at how TMA detects changes in dimensional behavior and how developers can read TMA data alongside DSC results.
