PSLE-SCI-REALITY-0569
Wait, what? One peak owns 92% of the chart, so is 92% of the sample that substance?
A laboratory report shows a chromatogram: a line rises into several peaks, then falls back toward the baseline. The largest peak is labelled “92.0% area”. A smaller peak is labelled 5.1%, and two tiny peaks together make up the rest.
A learner reads the report and says, “Easy. The sample is 92% pure because 92% of the peak area belongs to the main substance.”
That may be the conclusion a particular validated method allows under particular conditions. But the number 92% peak area does not, by itself, prove that 92% of the sample’s mass, molecules or material is the named substance. A chromatogram is a measurement representation. Its peak areas are produced by a detector responding to compounds that have been separated by a method. To turn detector response into a composition claim, we must know what the detector responds to, whether different substances produce equal response for equal amount, whether any substances were missed or overlapped, and how the laboratory defined the calculation.
This Reality Lab owns one narrow learner job: how to stop a percentage of detector signal from silently becoming a percentage of sample purity.
Quick answer
No. “Peak area = 92%” means, at minimum, that under the stated chromatographic method and integration rules, the selected peak contributed about 92% of the integrated detector-response area included in that calculation. Whether that can be interpreted as 92% purity depends on the method and its assumptions.
A scientifically careful learner asks six questions before turning area share into composition share:
- What exactly did the detector measure?
- Do equal amounts of different compounds give equal detector response?
- Were all relevant compounds separated into distinct peaks?
- Could any relevant material be invisible or weakly visible to this detector?
- How were peak boundaries and baselines integrated?
- Does the method explicitly validate area percentage as a purity or composition estimate for this sample and purpose?
The exact learner job this page owns
This page is not a chromatography course. It does not own stationary phases, mobile phases, retention mechanisms, column chemistry, HPLC design, gas chromatography, detector engineering or advanced analytical chemistry. Those mechanisms remain with the existing eduKate scientific concept owners.
It also does not re-own the broad question “What does 99.9% pure mean?” Reality Lab Vol.216 already owns the general job of asking pure by what measure, on what fraction basis, and what is in the remainder?
The job here is more specific: a learner sees a chromatogram peak-area percentage and must decide what that percentage actually represents before using it as evidence about sample composition.
Why this is PSLE Science reasoning, even though the instrument is advanced
You do not need to operate a chromatograph to practise the reasoning. The 2023 Primary Science syllabus emphasises interpreting information, healthy scepticism, evaluating evidence and understanding that science is communicated through different forms and media. The 2026 PSLE Science assessment objectives include interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning.
The transferable habit is simple: name the measured quantity before you interpret the number. A percentage can be a percentage of mass, volume, people, time, pixels, detector counts, peak area, probability, coverage or something else. The percent sign does not tell you the denominator. The method does.
The Signal Ledger: keep four quantities separate
Instead of memorising a slogan, build a small signal ledger. Keep these four quantities in separate boxes until evidence connects them.
| Ledger box | Question | Example |
|---|---|---|
| Amount in sample | How much material is physically present? | milligrams, moles, concentration, mass fraction |
| Separated fraction | What reached the detector at this time? | a chromatographic band or peak |
| Detector response | How strongly did the detector respond? | signal intensity through time |
| Integrated peak area | How much response was accumulated under the chosen peak? | area units or area percentage |
Sometimes these boxes are closely related. A validated quantitative method can establish a dependable relationship between amount and response. But the relationship is evidence, not magic. You should not jump from the last box to the first just because both numbers can be written as percentages.
Rebuild the real-world evidence object
Consider this original composite laboratory report. It is not copied from any commercial report or exam question.
| Peak | Retention time | Integrated area | Area % |
|---|---|---|---|
| A | 2.1 min | 3,000 | 2.0% |
| Main | 5.8 min | 138,000 | 92.0% |
| B | 7.2 min | 6,000 | 4.0% |
| C | 9.5 min | 3,000 | 2.0% |
The four integrated areas total 150,000 response-area units, so the main peak contributes 138,000 ÷ 150,000 = 0.92, or 92% of the included area. That arithmetic is sound.
But the scientific question comes next: 92% of what? The answer supported directly by the table is “92% of the integrated detector-response area included in this calculation.” The table alone does not say “92% of sample mass.”
Observed, calculated, claimed, inferred
| Layer | What belongs here? | Common overreach |
|---|---|---|
| Observed by instrument | Detector signal varies with time | “The detector directly saw purity.” |
| Calculated | Peak boundaries are integrated; area percentages are computed | “The percentage must be a mass percentage.” |
| Method-defined claim | A validated method may define how area relates to purity or impurity | “Every chromatography method uses the same rule.” |
| Inference | Composition or purity conclusion within method limits | “The number is universally true for the whole sample.” |
The detector is not a neutral camera
Imagine two coloured lamps. One sensor is twice as sensitive to blue light as to red light. If both lamps send the same physical amount of light to the sensor, the blue lamp may produce a larger electrical response. A larger response does not automatically mean a larger physical amount unless the sensor response has been characterised.
Chromatographic detectors have response behaviour too. The National Institute of Standards and Technology notes, in guidance on chromatographic purity measurements, that relative peak-area approaches can be affected by differences in detector response factors among impurities and the main component. NIST specifically warns that assuming equal response can be incorrect for many detector situations, and it identifies co-eluting peaks, response factors, bias and precision as sources that matter to the measurement.
For a Primary 5/6 learner, the practical translation is: the instrument’s response rule matters.
Response factor: why equal amount can produce unequal signal
Suppose Substance X and Substance Y are each present at the same amount. Under a particular detector and method, X produces 100 response units while Y produces only 50. If you compare raw peak areas without correction, X looks twice as abundant even though the physical amounts are equal.
Now reverse the situation. An impurity produces a very strong detector response. A small amount could create a surprisingly large peak. A second impurity gives a weak response. A larger amount could create a modest peak. Therefore peak area and material amount are not automatically interchangeable quantities.
Quantitative methods can account for these relationships using calibration, response factors or other validated calculations. The Reality Lab habit is not “never trust peak area.” It is “find the method that earns the translation.”
Separation check: two substances can hide under one apparent peak
A chromatogram is useful because components can be separated before detection. But separation is not automatically perfect. Two compounds can emerge close together. If their signals overlap, one apparent peak can contain contributions from more than one substance. This is called co-elution.
Imagine a smooth mountain-shaped peak labelled “Main”. Underneath, two substances may contribute to the shape. The chart still shows one broad feature. If the method has not demonstrated adequate separation, treating the whole area as one pure component can overstate what the representation proves.
A strong report therefore gives evidence about method suitability: resolution, standards, known retention behaviour, spectral checks or other method-specific validation. A Primary learner does not need to calculate those advanced quantities. The transferable question is: How do we know this peak belongs only to the thing we say it belongs to?
Visibility check: not every substance must produce a counted peak
Area percentage uses the peaks that were detected, integrated and included under the method. But a detector may not respond equally to every possible component. Some material may be outside the method’s target range, poorly detected, not volatile enough for a particular technique, not absorbing at the selected wavelength, lost in sample preparation, or excluded by an integration threshold.
This creates an important denominator question. If the report says “92% of detected integrated area,” the denominator is the integrated area included by that procedure. It is not automatically “all matter that exists in the vial.”
Integration check: where does a peak begin and end?
Peak area is not a coloured region painted by nature. Software estimates a baseline and integrates signal above it over a chosen interval. Different defensible integration settings can sometimes change the area assigned to neighbouring or irregular peaks. Good laboratory methods specify how integration is handled and review unusual cases rather than treating software output as unquestionable truth.
For a learner, this is another representation lesson. A printed percentage may look final, but it can depend on a chain of prior choices: sampling, preparation, separation, detection, baseline selection, peak identification and calculation.
Baseline check: a flat-looking line is still part of the measurement
Look at any stylised chromatogram and the baseline may appear to be a perfectly flat zero line. Real instruments can show noise, drift and background response. If a baseline changes, the estimated area under a small peak can change too. This matters especially for tiny impurity peaks near detection limits.
The PSLE transfer is powerful: before interpreting an area, understand where the baseline came from. The same habit applies to graphs that start at a non-zero axis, before/after images with different brightness, and instruments that subtract a blank reading.
Calibration check: when can signal support amount?
A calibration experiment measures known amounts and observes detector responses. If the method shows a dependable relationship over a stated range, an unknown response can be translated into an estimated amount. This is much stronger evidence than merely assuming “big peak means proportional amount.”
But calibration has boundaries. A relationship verified from 1 to 100 units should not automatically be stretched to 10,000. Earlier Reality Lab volumes have explored why measurements above a calibration range should not be treated as precise exact concentrations. The same principle applies here: translation rules have domains.
Method-definition check: sometimes area-normalisation really is the specified procedure
It would be equally unscientific to swing to the opposite extreme and say, “Peak area can never tell us anything about purity.” Analytical methods are designed precisely to turn instrument evidence into useful estimates. Some methods define relative area calculations for a specific material, detector, range and impurity profile. When validation supports the assumptions, the area result can be informative and fit for its intended purpose.
The scientific habit is conditional trust: trust the result to the extent that the method, calibration, separation, detector response and validation support it.
The denominator audit
Whenever a report gives “92%”, finish this sentence before interpreting it:
92% of ____________________.
Possible completions include:
- the total integrated peak area;
- the corrected detector response after applying response factors;
- the estimated mass fraction under a validated method;
- the peaks above a reporting threshold;
- the compounds detectable under the selected conditions.
Those are different denominators. Scientific reading begins when the blank is filled correctly.
Worked case 1: equal areas, unequal response factors
An original composite report shows two peaks with equal areas: 5,000 and 5,000. A student says the two substances must be present in equal amounts.
Additional method information says Substance B produces twice the detector response per unit amount as Substance A. The equal areas no longer support equal amounts. If all other conditions are valid, B could require only about half as much material to create the same response.
Lesson: equal signal does not guarantee equal amount when response differs.
Worked case 2: one tall peak, one hidden neighbour
A chromatogram shows a single broad main peak with 96% of total area. A second analytical check indicates that two compounds overlap at nearly the same retention time.
The 96% area remains a real property of the integrated signal region, but its meaning changes. It can no longer automatically be assigned entirely to one compound. The evidence-transfer job is to update the claim when new evidence reveals that the representation merged contributions.
Worked case 3: a detector that misses one class of material
A product is analysed by a method that strongly detects the target compound and several known organic impurities. Water in the sample gives little or no relevant response under that detector setting. The main peak is 99% of integrated area.
The chart alone cannot prove the entire sample is 99% target by total mass because material outside the detector’s useful response could contribute to the sample without contributing proportionally to the chromatogram area. A separate water measurement or a validated comprehensive method could change the conclusion.
Worked case 4: corrected response factors
Another method has measured relative response factors for three known impurities. The laboratory applies those corrections, demonstrates adequate separation, uses standards across the working range and reports an uncertainty suitable for the purpose.
Now the link from detector response to composition is much better supported. A cautious learner should not reject the composition result merely because it began as detector signal. The method has supplied the missing bridge.
Worked case 5: software changes the integration boundary
Version 1 of a report assigns 91.8% area to the main peak. After a reviewed integration rule separates a shoulder peak, the main area becomes 90.9% and the shoulder becomes 0.9%.
The sample did not physically change after the computer reprocessed the data. The representation and calculation changed. The learner should ask which integration procedure is method-defined and why the revision is more defensible.
Worked case 6: same sample, different detector
The same prepared sample is measured with two legitimate detectors. Detector 1 gives the main peak 94% of integrated area; Detector 2 gives it 89%. Is one necessarily wrong?
Not necessarily. Different detectors can have different sensitivities to different compounds. To compare the percentages, you need the method definitions, detector response behaviour and calibration evidence. “Same sample” does not guarantee “same raw signal percentage” across different measurement systems.
What evidence strengthens the claim “this area percentage estimates purity”?
- The method explicitly defines the area-normalisation calculation for this material and purpose.
- Known relevant impurities are separated sufficiently from the main component.
- Detector response factors are demonstrated to be similar or are measured and corrected where needed.
- Calibration covers the relevant working range.
- Standards or reference materials confirm peak identity and method behaviour.
- Integration rules are specified and reviewed.
- Important components are not systematically invisible to the measurement.
- Replicate measurements show acceptable precision.
- Independent or orthogonal evidence agrees with the composition estimate.
What evidence weakens the claim?
- The report gives only an unlabeled “area %” with no method definition.
- Known compounds have very different detector responses.
- A peak is unresolved or visibly overlaps another.
- The detector is known to miss a relevant component.
- The main peak is outside the validated detector or calibration range.
- Integration settings are arbitrary or changed without justification.
- A large unidentified peak is excluded from the denominator.
- The conclusion changes sharply when reasonable processing choices change.
- No evidence links the area percentage to the claimed composition basis.
How far can the conclusion travel?
Suppose a validated method supports 92.0% purity by a defined chromatographic procedure for one prepared sample. How far may we generalise?
Not automatically to every bottle from the factory, every point inside a non-uniform batch, every future production run, every type of impurity or every possible analytical method. The result belongs first to the sample, preparation, method, instrument conditions and reporting rules that generated it. Broader claims need broader evidence: representative sampling, process controls, repeated batches and suitable validation.
Tempting reasoning that fails
| Tempting thought | Why it fails |
|---|---|
| “92% area means 92% mass.” | The denominator is detector-response area unless the method establishes the translation. |
| “The tallest peak is the substance present in the greatest amount.” | Detector responses can differ among compounds. |
| “One peak means one compound.” | Co-elution can combine signals. |
| “No peak means none of the substance exists.” | The detector or method may not detect it adequately. |
| “Software gave the percentage, so the value is objective.” | Integration and baseline rules are part of the method. |
| “Because area % is imperfect, chromatography is useless.” | Validated quantitative methods can establish reliable relationships and uncertainties. |
The three-denominator test
When you see a percentage in a scientific report, ask three denominator questions:
- Mathematical denominator: What numbers were actually divided?
- Measurement denominator: What could the instrument and method detect and include?
- Claim denominator: What population, material or system is the writer talking about?
A misleading conclusion often appears when these three denominators are treated as if they were automatically the same.
PSLE-style transfer case: coloured beads and a biased sensor
A box contains red and blue beads. Instead of counting the beads, a sensor gives red beads 2 signal units each and blue beads 1 signal unit each. The total signal is 80% red and 20% blue.
Can you say 80% of the beads are red? No. The sensor gives unequal response per bead. You need the response rule. If there are 40 red beads and 20 blue beads, the signal is 80 red units and 20 blue units: 80% of signal but only 40 out of 60 beads, or about 67% of the beads.
This simple model captures the same evidence-transfer problem without teaching advanced chromatography.
Transfer case: camera brightness
A camera records one surface as twice the pixel value of another. Does that mean the surface reflects exactly twice as much light? Not automatically. Exposure, sensor response, processing and calibration matter. Again, measured signal and physical quantity must be connected by a method.
Transfer case: microphone amplitude
A sound recording has a waveform with a taller peak. That does not automatically mean “twice as loud” to a person. The representation has a measurement scale and a response process. The scientific move is the same: identify what the signal encodes.
Model and measurement limits
Every analytical measurement simplifies a physical sample into selected evidence. A chromatogram is especially instructive because it compresses a complicated chain into a clean plot: sample preparation, separation, detector response, timing, baseline, integration, identification and calculation. The clean plot can create an illusion that the number fell directly out of the sample. It did not. It was produced by a method.
That does not make the result untrustworthy. It tells us where trust should be placed: in a well-characterised method, appropriate controls, standards, calibration, separation, suitable detector response, documented calculations and a conclusion that stays inside those limits.
Practice: read the signal before the story
- A report says “Main peak area = 97%.” What is the safest direct interpretation?
- Why can two equal amounts give different peak areas?
- Why can one apparent peak contain more than one compound?
- If a detector does not respond to water, what problem could occur when interpreting area % as total mass %?
- What question should you ask about the baseline before trusting a very small peak?
- What evidence would make an area-based purity claim stronger?
- Why is “the method is imperfect” not enough reason to reject every chromatographic result?
- A sample gives 95% area today and 95% area tomorrow. Does that prove every future production batch will be 95% pure?
- Write the missing denominator in the statement “92% of ________.”
- Name one non-chromatography situation in which sensor response should not automatically be treated as physical amount.
Explained answers
1. The main peak contributes 97% of the integrated detector-response area included under that method and calculation. Any purity interpretation needs method support.
2. Different compounds can produce different detector response per unit amount.
3. The separation may be incomplete, so two compounds can co-elute and contribute to one signal region.
4. Water could contribute physical mass while contributing little or none of the counted detector area, so area % would not automatically equal total mass %.
5. Ask how the baseline was defined, how much noise or drift exists, and whether the tiny peak is reliably above the method’s detection and integration rules.
6. Stronger evidence includes adequate separation, appropriate response factors or calibration, standards, method validation, specified integration rules and corroborating measurements.
7. Scientific methods can quantify and control known limitations. A validated result may be reliable for its intended purpose even though no measurement is magically assumption-free.
8. No. Repeated results support repeatability for those samples and conditions. Future batches require representative process evidence.
9. In the raw area-normalisation example: “integrated detector-response area included in the calculation.”
10. Examples include camera pixels, microphone signals, fluorescence intensity, satellite digital numbers and many other instrument outputs.
Delayed independent return
Tomorrow: invent a detector that responds twice as strongly to green tokens as to yellow tokens. Create a signal percentage that differs from the token-count percentage.
In three days: take any scientific percentage you encounter and write its denominator in words. If you cannot, mark the claim as incomplete until you can.
In one week: revisit the original 92% chromatogram case without looking at this article. Explain why the arithmetic can be correct while the purity inference can still require more evidence.
Route to existing canonical owners
For the mechanisms of chromatographic separation and HPLC, use How to Learn Chromatography and Chemical Separation: From Paper Spots to HPLC and Analytical Chemistry. For the broader meaning of a purity percentage and the importance of its fraction basis, use PSLE Science Reality Lab Vol No.216 | “99.9% Pure” — Pure by What Measure, and What Is in the Other 0.1%?. For keeping conclusions at the level actually supported by evidence, use How to Keep a PSLE Science Claim at the Right Evidence Level. For separating what was observed from what was inferred, use How to Tell Observation, Inference, Prediction and Explanation Apart in PSLE Science.
Parent and tutor teaching guide: use a biased counter, not a chromatography lecture
Do not begin by teaching instrument hardware. Begin with a deliberately biased counter. Put ten blue counters and ten red counters on a table. Tell the learner that the imaginary detector gives each blue counter one signal unit and each red counter two signal units. Ask for the percentage of objects that are red, then the percentage of signal that is red.
The object percentage is 50%. The red signal share is 20 red units out of 30 total units, about 67%. That mismatch creates the insight immediately: a signal share can differ from an amount share.
Next show a simple four-peak sketch and label only the peak areas. Ask, “What do we know directly?” Reward the answer “relative integrated signal” before asking what extra evidence would be needed to talk about purity. This teaches evidence discipline without pretending a Primary learner must master analytical chemistry.
Finally transfer the habit to a different representation: a camera image, sound waveform or fluorescence graph. The learner has succeeded when they ask “What does the sensor respond to?” without being prompted.
Authoritative sources
- Ministry of Education Singapore: 2023 Primary Science Teaching and Learning Syllabus.
- Singapore Examinations and Assessment Board: 2026 PSLE Science syllabus.
- National Institute of Standards and Technology, Special Publication 1209: Analytical Measurement Procedures for Purity Determination — discusses chromatographic purity approaches, response factors, detector range, co-eluting peaks, bias and precision.
The quiet habit to keep
A percentage is never complete until you know its denominator. A detector signal is never a composition claim until the method earns the translation.
SAMPLE → SEPARATION → DETECTOR RESPONSE → INTEGRATION → METHOD RULE → LIMITED COMPOSITION CLAIM.
That chain is the difference between reading a number and evaluating evidence.
