Did you know that two differential scanning calorimetry plots can display the same melting event as peaks pointing in opposite directions? If you need to translate a DSC thermogram into clear English, accessible narration, a technical report, or another language, your first task is not to name the peak. It is to establish what the axes, sign convention and experimental programme actually say.
This guide explains how to translate DSC thermograms without reversing heat flow, misidentifying a melting onset, or reporting a peak area as though it were a temperature. Differential scanning calorimetry, commonly abbreviated DSC, tracks heat-flow behaviour as a material is heated, cooled or held at a specified temperature. The curve is evidence of thermal events, but it is not a complete molecular explanation by itself.
For an accurate translation, keep three questions separate: What was measured? What pattern was observed? What interpretation is justified? This distinction helps students reading scientific diagrams, researchers preparing multilingual documents, accessibility editors describing graphs and technical teams comparing laboratories. A good translation makes the thermal event understandable while preserving its units, limitations and conditions.
Begin with the simplest reliable reading
A DSC thermogram is a plot produced by an instrument that compares the thermal response of a sample against a reference under a controlled temperature programme. The instrument records a heat-flow-related signal, while software displays it against temperature or time. The signal may rise or fall around an event such as melting, crystallisation, a glass transition, evaporation or a chemical reaction. The shape alone is not enough to label a process with confidence: readers also need the instrument convention, scan direction, material and experimental history.
When translating a figure, begin with a neutral description. Say that the trace shows a downward feature near a stated temperature, not that it automatically shows melting. Then establish whether the instrument labels downward signals as endothermic and whether the trace is from heating or cooling. Only after those checks should the final wording identify an endothermic or exothermic event. The result is less dramatic than guessing, but far more useful.
Why the word translate belongs in a thermal-analysis laboratory
Translation here involves more than replacing English words with words from another language. A DSC image contains information expressed by line direction, coordinates, tick marks, instrument labels, colour legends and annotations. Someone reading aloud must convert a spatial graphic into a sequence of sentences. Someone preparing a bilingual report must transfer the technical meaning of words such as onset, heat flow, enthalpy and transition without altering what the experiment demonstrated.
The same information may travel through a journal figure, equipment screenshot, lecture slide, data sheet and quality-control report. Each format loses something unless the translator deliberately restores it. A screenshot may omit the axis caption; a slide may crop a sign-convention arrow; a published graph may normalise the signal by sample mass. Ask which features were present in the source and which were inferred later. Keeping that distinction visible is a core professional skill.
Identify the horizontal axis before interpreting a peak
Many DSC figures use temperature in degrees Celsius on the horizontal axis, but time in minutes or seconds is also possible. A heating scan typically progresses toward higher temperature; a cooling scan progresses toward lower temperature. Software may still arrange the plot from low to high temperature even when the experimental direction was reversed. This makes the direction of the plotted axis different from the direction in which data were acquired.
Write down the axis title, unit, displayed range and scan direction as four separate facts. If the legend indicates cooling at a specified rate, include that fact in a figure description instead of assuming the left-to-right movement of your eyes represents the experiment. For a time plot, do not assign every time point a particular temperature unless the temperature programme is supplied. Hold segments, ramps and cooling phases can create different relationships between time and temperature.
Read the vertical axis and all of its units
A DSC vertical axis may be labelled heat flow, heat-flow rate, specific heat-flow rate, differential heat flow, or a manufacturer’s calibrated signal. A trace shown in milliwatts reports power-like quantities; a trace in watts per gram has been normalised to a specimen mass. The interpretation of numerical peak height changes when the unit changes. A milliwatt peak should not be translated as milliwatts per milligram, and a per-gram value must not be described as total power for the specimen.
Look for offsets as well. A figure comparing multiple samples may move traces vertically for readability. Those offsets are not measured changes in enthalpy unless the graph explicitly says so. A label such as endo up can matter as much as the numeric ticks. When neither label nor documentation establishes the sign convention, describe the peak direction visually and state that its thermal sign remains to be confirmed.
Endothermic up and endothermic down are both possible
The most avoidable translation error is assuming that an upward peak means a substance absorbed heat. DSC plotting conventions vary. Some software displays endothermic events upwards; other displays show them downwards. Endothermic describes a heat-absorbing event under the relevant conditions, whereas exothermic describes a heat-releasing event. Neither word inherently means a graphic must rise on a page.
Suppose a plot shows a strong downward peak near a known melting interval. If its annotation states endothermic down, a reader may reasonably describe a downward endothermic feature. If the annotation says endothermic up, that same downward feature should not be called melting solely because the temperature seems plausible. It might represent another process or a differently processed signal. Check the legend, instrument method or source notes before completing the translation.
Distinguish heat flow from accumulated energy
Heat flow concerns the rate of energy transfer; enthalpy associated with an event comes from integrating an appropriate peak relative to a baseline, with calibration and normalisation considered. These are related but different quantities. Describing a larger instantaneous signal as a larger total enthalpy is not automatically valid. A narrow tall peak and a broad short peak can have different integrated areas and different total energies.
A careful translation separates peak height, peak width and peak area. If a table says delta H equals a value in joules per gram, explain that the reported enthalpy change is normalised by mass. If a curve only shows milliwatts against degrees Celsius, avoid asserting a precise joules-per-gram value without the integration result and heating rate needed for conversion. The technical meaning depends on what the software actually calculated.
Preserve the complete temperature programme
A DSC method is not just a start and end temperature. It may include a first heating ramp, a cooling segment, an isothermal hold and a second heating scan. Those stages can reveal different effects of processing history. A first heat may show events caused by earlier storage or manufacturing; a second heat may reflect the sample after the instrument imposed a controlled thermal history.
In translation, name each scan explicitly: first heating, controlled cooling or second heating. Do not combine results from separate scans into one anonymous thermogram. If a material crystallises during cooling but melts during subsequent heating, readers must know which event belongs to which scan. A description should also preserve the heating or cooling rate where available, because rate changes can shift or reshape features and complicate comparisons between reports.
Find the baseline before measuring an event
The baseline is the signal used as a reference outside the event under consideration. Depending on the method and the sample, it may slope or change around a transition. Software may fit a straight line, construct a sigmoidal baseline or follow a defined integration rule. A shaded peak region can look precise while still depending on the analyst’s baseline choice.
A translator should never erase that analytical decision. If a paper states that enthalpy was calculated using a linear baseline between two temperatures, preserve both the method and the range when they matter. If the figure only shows a shaded region, say that the shaded area represents the software’s integrated event region. Do not imply that a baseline is a physical boundary inside the sample. It is an analytical reference established to estimate a thermal contribution.
Understand what an onset temperature claims
An onset temperature commonly marks the beginning of an identified transition as defined by a particular analytical convention. One common reported value is an extrapolated onset, derived using tangents or an intersection with an estimated baseline. It is not necessarily the first pixel in a curve that deviates from straightness. Different onset definitions can produce different numbers for the same recorded event.
If the label reads extrapolated onset, translate that complete term. If a figure reports onset of melting at a given temperature, do not casually rewrite it as the peak temperature. In formal reports, note the onset method or software convention where needed. A reader using the measurement for process control may make a different decision depending on whether the value refers to extrapolated onset, maximum heat-flow position or completion of an event.
Separate peak temperature from onset and endset
The peak temperature is often associated with a local maximum or minimum in the measured heat-flow signal, depending on the display sign. It is not identical to the onset, which is an earlier convention-defined point, nor to the endset or end of integration. For a broad event, those differences may be substantial, and they communicate information about the temperature interval over which the process occurs.
Imagine a labelled example with an onset at 112 °C and a peak at 119 °C. A poor translation says the material melts at 119 °C and ignores the onset. A better one says that the reported melting-related endotherm begins at an extrapolated onset near 112 °C and reaches its heat-flow extremum around 119 °C. If the material is a mixture or polymer, avoid implying it has the sharp equilibrium melting point of a pure ideal substance.
Glass transitions are often steps, not classical peaks
A glass transition in a polymer frequently appears as a step or shift in the DSC baseline because the heat capacity changes across the transition region. It should not be automatically treated as a melting peak. Some traces also exhibit enthalpy-relaxation overshoots around the glass transition, which can resemble a separate peak and confuse inexperienced readers.
When translating a report of glass-transition temperature, preserve the chosen criterion: onset, midpoint, half-step or another stated method. The symbol Tg is useful but does not replace a clear definition. Write that a baseline shift is consistent with a glass transition under the recorded conditions rather than announcing that every observed step definitively establishes one. Sample thermal history, heating rate and moisture can influence the feature and how the analyst reports its position.
Melting has a direction and an interpretation
Melting requires energy input under ordinary DSC heating conditions and is therefore generally associated with an endothermic event. But a correct phrase such as melting endotherm still requires the sign convention to be read accurately. A polymer may display multiple melting features due to crystal populations, reorganisation or sample history, and a mixed material may show overlapping transitions.
Do not translate several peaks into a claim that there are necessarily several chemically distinct ingredients. Describe what appears, then qualify the explanation. For example, two endothermic peaks during heating can be reported as two endothermic features within the observed temperature interval. Further analysis may support a particular structural interpretation. Respecting this boundary between observation and hypothesis helps readers trust the report.
Crystallisation usually releases heat
Crystallisation during cooling is often accompanied by heat release and an exothermic signal. A cold-crystallisation event can also occur during heating when a previously insufficiently crystallised material develops additional order before later melting. If a graphic shows exothermic activity during heating, calling it cooling crystallisation will mislead the reader about the experimental sequence.
Use the terms cooling crystallisation and cold crystallisation where they apply; they do not refer to the same sequence. Make the scan stage visible in each caption. A readable description might state that an exothermic peak appears during the cooling run and that a separate exothermic feature appears during reheating. That wording preserves both events without pretending a diagram alone proves how the molecular structure changed.
Curing and chemical reactions require additional caution
Reactive resins and adhesives may release heat as curing progresses. DSC can record exothermic behaviour that helps evaluate the reaction under specific test conditions. But the area of a cure-related peak is not automatically a direct measure of product safety, complete reaction, final mechanical strength or readiness for service. These are different properties that can need additional measurements.
If a technical document calls a peak a residual cure exotherm, preserve the modifier residual: it signals that a specimen already underwent some processing before the measurement. If a second heat no longer shows the same feature, describe that observation rather than concluding that all reactions are impossible thereafter. Calibration, sensitivity and the measured temperature window set limits on what absence of a peak means.
Evaporation and mass loss may mimic other thermal events
An endothermic-looking feature may arise from evaporation of water or solvent rather than melting of the main material. DSC alone does not always identify the gas or liquid released. A concurrent change in mass measured by thermogravimetry can provide supporting evidence, but the two instruments do not measure the same physical quantity. Heat flow is not mass loss.
Therefore, a translation should not turn an unassigned broad endotherm into a definitive chemical name. If the original says consistent with removal of absorbed moisture, keep that degree of confidence. If the source reports a mass-loss event in a separate TGA analysis, connect the observations with a phrase such as occurring over a similar temperature interval, rather than claiming one graph proves the other.
Overlapping events are not one neat story
A sample can melt, crystallise, evaporate or react across nearby or overlapping temperature ranges. Their heat-flow contributions may partially cancel or alter the visible shape. A shoulder on a peak is a feature worthy of description but not a direct count of microscopic processes. Software deconvolution may assign components, but the result depends on its assumptions.
When figures contain shoulders or mixed peaks, state that the event appears broad, asymmetric or partially overlapping. Preserve any uncertainty attached to separation procedures. Avoid translating a mathematically fitted component as an independently observed physical transition. The fit is a model of the data, not a second experiment. If the study proposes identities for the components, attribute them to the analysis instead of stating them as obvious facts on the raw trace.
Heating rate changes what can be compared
Increasing the heating rate can shift some measured transition positions, change peak shape and make temperature gradients more important. Slower scans may improve resolution of certain overlapping events but can also allow more time for reactions or reorganisation. There is no universally correct rate for every specimen and question. The experimental goal drives the method.
In a translated comparison, always carry over units such as degrees Celsius per minute or kelvin per minute. A rate of 10 °C/min is not the same as a hold at 10 °C, and a typographic slash is meaningful. If two published curves use different heating rates, warn that peak temperatures should not be compared as though rate were controlled. Explain the limitation without declaring that either dataset is defective.
Sample mass and pan configuration matter
DSC responses depend on how the sample and reference are prepared. Crucibles or pans may be open, sealed, pierced or designed for particular pressures, depending on instrument capabilities and safety requirements. Specimen mass, contact with the pan, distribution of material and atmosphere can affect the trace. A small sample in one setup may yield a cleaner transition than a poorly contacting specimen in another.
For translation, include relevant sample mass and pan type when interpreting differences among results. If an experimental description says sealed pan, do not silently replace it with open pan. The distinction can alter volatile retention and observed events. Likewise, do not mistake an instrument’s automatic baseline correction for evidence that preparation differences have disappeared. The metadata are part of the meaning of the graph.
Atmosphere and gas flow belong in the caption
DSC experiments are often run under nitrogen, argon, air or another specified atmosphere. The gas can influence oxidative reactions and affect what is observed. A material stable under an inert purge may behave differently in oxygen-containing conditions. Descriptions such as tested in air and tested under nitrogen should not be interchanged simply because both plots look similar.
Record the purge gas, nominal flow rate and any gas switching when the source provides them. If this information is missing, do not invent it. A technical translator can say that the atmosphere was not specified in the supplied figure and refer readers to the full method. This is especially important when a curve is used to justify a material choice in manufacturing or research, where environmental assumptions matter.
Calibration turns a line into a measurement
A DSC instrument produces a signal that must be related to temperature and heat flow through appropriate calibration and verification practices. Reference materials, calibration procedures and instrument condition influence measurement quality. A visually smooth plot should not be treated as proof of reliable temperature or enthalpy values. The apparatus and method determine how confidence is established.
When localising a laboratory document, distinguish calibration performed, calibration interval, verification check and instrument qualification. Those are related but not interchangeable claims. If an article reports values without an uncertainty estimate, the translator should not add invented precision. Reproduce the significant figures and accuracy qualifications provided. Readers will appreciate a transparent statement of what was measured rather than a polished sentence that implies more certainty than the experiment earned.
Make numerical expressions robust across languages
Numbers can be damaged during localisation even if every technical noun is correct. Decimal marks, thousands separators, negative signs, nonbreaking spaces and unit abbreviations vary across conventions. In a DSC report, 0.8 J/g must not become 8 J/g because a decimal separator vanished. The temperature interval 40–80 °C is a range, not a subtraction problem. Preserve the mathematical relationship when reformatting.
For quality assurance, compare the translated values directly with the original figure: the onset, peak, enthalpy, heating rate, sample mass and atmosphere. Work with structured values where possible instead of copying numbers from screenshots. Avoid reversing labels in right-to-left writing systems: a technical abbreviation such as DSC or Tg may require bidirectional text handling so that adjacent numbers and units remain in the intended order.
Build a bilingual glossary before translating the explanation
A compact glossary can include heat flow, heat-flow rate, differential scanning calorimetry, endothermic, exothermic, melting, crystallisation, cold crystallisation, glass transition, onset temperature, peak temperature, enthalpy change, baseline, heating rate and cooling run. Definitions should be tied to how the relevant report uses each term rather than to a vague dictionary synonym.
Do not force one everyday-language word to stand for several distinct scientific concepts. For example, transition is broader than melting, and onset is not the same as origin or source. Where a term lacks an exact common-language equivalent, a short explanation in the target language may be clearer than an unnatural borrowed noun. Maintain consistent abbreviations in equations, figure labels, running text and tables so readers can map each expression to the same quantity.
A worked example with a melting-related feature
Consider an illustrative DSC heating scan of a semicrystalline polymer. The figure’s horizontal axis reads temperature, °C; the vertical axis reads heat flow, W/g; the note says endothermic down. The curve shows a small baseline step, a later exothermic feature and finally a downward endothermic peak. An accompanying table reports an extrapolated melting onset, a peak temperature and an integrated enthalpy per gram.
A reliable translation first reports the heating scan and sign convention. It describes the baseline step as glass-transition-related only if supported by the source analysis, distinguishes a reheating crystallisation feature from melting, and records melting onset, peak temperature and enthalpy as separate values. It does not call the lowest part of the melting peak the onset, nor describe the area under the peak as an additional temperature. This method transfers the relationships, not merely the picture.
A worked example with a cooling curve
Suppose a second figure shows cooling from a high temperature after the specimen was melted. A pronounced upward feature appears, while the caption indicates exothermic up. The original report labels the feature crystallisation. A translator who assumes every DSC plot is a heating curve might describe this as melting, especially if the sample name is familiar. That error changes the physical event rather than merely its wording.
The better account begins, During controlled cooling, an exothermic crystallisation feature is observed. It adds the reported onset or peak temperature with the appropriate definition. If the cooling rate differs from the heating rate, note the difference. When comparing cooling crystallisation with subsequent melting, resist implying that the two event temperatures must match; thermal history, kinetics and experimental programme may influence the observed positions.
Write short accessible descriptions before long interpretations
A person using a screen reader needs the graph’s purpose and message before hearing every number. A useful short description may say: DSC thermogram comparing first and second heating traces, with a baseline step and two labelled thermal events. Then a longer description can specify axes, units, legend, sign convention, key onsets and peak positions. That order helps readers form a mental model before hearing detailed measurements.
Avoid descriptions like interesting wavy line or dramatic dip, which omit the meaning of the plotted quantities. But also avoid constructing a full molecular mechanism from an unlabeled graphic. If several traces overlap, narrate the legend and each trace’s relationship to the others. When colours are used, give an identifier other than colour alone, such as sample A or second heating scan, to make the chart interpretable without relying on vision.
Translate captions and tables as a matched pair
A caption may say endothermic down while a results table uses a symbol for enthalpy of fusion. Translating the caption alone can leave the reader with two incompatible vocabularies. Cross-check all occurrences of the temperature value and event name across the figure, caption, results paragraph and table. Include subscripts or explanatory words when they are needed to tell melting enthalpy from crystallisation enthalpy.
Captions should identify the material, scan, axes and major inference at an appropriate level. They should not substitute for the experimental method. A table provides exact reported values and labels; the prose explains what can be concluded from them. If values conflict, flag the source inconsistency rather than silently choosing one. The translator’s role is fidelity and clarity, not unacknowledged data correction.
Distinguish direct observation from interpretation
An observed peak, a software-integrated enthalpy and a proposed material explanation occupy different evidence levels. You can see a peak. The software calculates an area under specified assumptions. The research team may interpret the area and transition sequence as consistent with crystallinity, thermal history or reaction progress. Language such as proves, establishes, indicates and is consistent with carries different strength.
Match the original strength of claim unless the original is demonstrably ambiguous and a clarification is sought. Do not upgrade suggests into confirms. Conversely, do not remove confidence where a well-controlled experimental method supports a clear conclusion. When discussing proprietary samples or incomplete protocols, highlight what cannot be inferred without overloading the text with disclaimers. A strong technical translation is precise and readable at the same time.
Review three representative errors
First error: The downward peak at 125 °C is exothermic. Without a displayed sign convention that sentence can be wrong. Repair it by stating the peak direction and identifying the thermal sign only after confirming the annotation. Second error: The sample’s melting point is 125 °C. The number might be peak temperature, extrapolated onset or a broad transition midpoint. Specify which number the source reports.
Third error: The peak contains 20 joules of heat. If the reported enthalpy is 20 J/g, the statement removes normalisation and changes the measurement. A revised translation states that the reported integrated enthalpy is 20 J/g under the defined scan and baseline method. These small adjustments matter because another reader may use the translated information to compare lots, choose materials or teach the underlying science.
A practical verification sequence for a finished translation
Read the original axis labels against the final sentence. Check that the scan direction and sign convention have not disappeared. Verify that every reported temperature is identified as onset, peak, midpoint or endset where the source distinguishes them. Confirm the units for heat flow and enthalpy, then check whether the values are normalised by mass. Review sample history, pan type, purge conditions and heating rate if they affect the comparison.
Next, compare the translated figure caption with the body text and results table. Finally, ask a reader who has not seen the original graph to describe what the experiment measured and what its limits were. If the reader cannot distinguish an observation from a hypothesis, the translation needs another pass. If the reader can reconstruct the main plot logic without inventing a peak identity, the work is doing its job.
When DSC should be supplemented by other evidence
A thermogram tells a thermal story, but it is not a universal chemical identifier. Thermogravimetric analysis can reveal mass changes across a temperature programme. Infrared or Raman spectroscopy may help investigate chemical bonds and structure. X-ray diffraction may provide evidence about crystalline structure. Microscopy can show morphology. The right combination depends on the research question, and the instruments should not be described as interchangeable.
For example, a broad endothermic feature near the same temperature as a TGA mass loss may motivate a hypothesis involving volatilisation. It does not automatically identify the volatile species. A glass-transition-related shift may support a polymer interpretation, but an additional experiment can be needed to understand material composition. Make these distinctions visible in teaching notes and multilingual technical documents rather than portraying one curve as an answer to every question.
Learn the vocabulary as connected concepts
Scientific vocabulary becomes easier when connected terms are studied together. Onset belongs beside peak and endset; heat-flow rate belongs beside power and enthalpy; melting belongs beside crystallisation and glass transition; heating rate belongs beside thermal history. Learning only isolated dictionary entries often produces translations that sound plausible but are scientifically contradictory.
An effective practice session uses one graph and asks for three outputs: a two-sentence plain-language explanation, a detailed technical caption and a glossary of crucial labels. Compare these outputs for consistency. Revisit the graph after a delay and attempt the description without seeing the original translation. That combination of explanation, retrieval and comparison builds durable understanding that transfers to other scientific plots.
Frequently asked questions about translating DSC thermograms
Does an upward DSC peak mean exothermic? Not necessarily. Read the instrument’s sign convention. Can onset temperature be called peak temperature? No; the values follow different definitions. Can DSC determine an exact percentage crystallinity? A calculation may be attempted when appropriate reference enthalpies, sample composition and model assumptions are available, but an unlabeled curve does not provide that percentage by itself. Is a broad endothermic feature always melting? No; overlapping processes or volatilisation may contribute.
Should the translator change temperature units? Only with careful conversion and clear identification of the new unit, preferably preserving the original where the report requires it. Is a second heating run interchangeable with the first? No; the specimen may have a different thermal history. Can an AI-generated caption be trusted without review? It should be checked against the source axes, sign convention, numerical values and supporting method. The standard is faithful meaning, not fluent-looking text.
Further reading and learning routes
Readers wanting to understand the experimental basis can consult the NIST guide to heat-flux DSC measurements and the ASTM description of enthalpy measurements by DSC. The eduKateSG article on differential scanning calorimetry and phase-transition evidence adds scientific context.
For neighbouring translation problems, read the eduKateSG guides to FTIR spectra and UV–visible absorption spectra. Explore the science learning hub, vocabulary learning through morphology and the guide to translation verification and bilingual review.
The reliable takeaway
A DSC trace becomes easier to explain when every sentence preserves the measurement’s structure. Identify the axes and sign convention. Separate scan direction from graph direction. Distinguish onset, peak, heat flow and integrated enthalpy. Treat captions, tables and method notes as parts of the same evidence. Above all, let the graph tell readers what was observed before proposing why it happened.
That is what accurate translation accomplishes: it changes the language and presentation while keeping the scientific relationship intact. A reader should leave with a clearer mental picture of the thermogram, not with a different experiment.