AllMyPeptidesLaboratory reference
The science desk
Reference material for laboratory work: the arithmetic that gets redone every week, the constants worth keeping to hand, and a plain account of what the common analytical instruments actually measure — and what they cannot tell you.
Quick reference
Units, constants and conversions
The numbers worth having in one place. Since 20 May 2019 the SI has been built on constants fixed by definition rather than by artefact — these are exact, not measurements with error bars.
Everything above is standard laboratory practice, not house style. Where a value is exact by definition it is marked as such; everything else carries the uncertainty of the method behind it.
Working principles
Things worth remembering
None of this is complicated. All of it is routinely got wrong, usually by trusting a number further than the method that produced it.
Significant figures
Multiplying or dividing keeps the fewest significant figures of the inputs. Adding or subtracting keeps the fewest decimal places. Rounding at the end, not at every step, avoids compounding the error.
Accuracy is not precision
Accuracy is how close a result sits to the true value; precision is how closely repeats agree with each other. A method can be reliably, repeatably wrong — tight scatter around the wrong number.
Every method has a blind spot
Each technique answers one narrow question well and is silent on the others. Most misreadings come from treating one method's answer as a general verdict.
Beer–Lambert has a ceiling
Absorbance rises linearly with concentration and path length until it does not. Above roughly A = 1 the relationship bends as the detector runs out of light to lose. Dilute into range rather than extrapolating out of it.
Detection and quantification differ
The limit of detection is the smallest amount distinguishable from noise. The limit of quantification — the smallest amount you can put a reliable number on — is higher, usually by around three times.
Blanks and controls
A result without a blank is a number without a baseline. Controls are what separate a real signal from an artefact of the solvent, the vessel, or the instrument itself.
Solutions
Concentration, pH and buffers
Four ways of saying how much is in the bottle, and why a buffer stops working outside a narrow window.
Henderson–Hasselbalch
pH = pKa + log([A−]/[HA]). When the acid and its conjugate base are present in equal amounts the log term is zero, so the pH simply equals the pKa. That is the whole reason buffers are chosen by their pKa.
Buffer capacity
A buffer resists pH change only within roughly one unit either side of its pKa. Outside that range one component is nearly exhausted and the pH moves almost as freely as unbuffered water.
Molarity vs molality
Molarity is moles per litre of solution and shifts with temperature, because volume expands. Molality is moles per kilogram of solvent and does not, since mass is temperature-independent.
ppm and ppb
In dilute aqueous solution 1 ppm is 1 mg per litre and 1 ppb is 1 µg per litre, because a litre of water weighs about a kilogram. That shortcut fails in any solvent that is not roughly water-dense.
Serial dilution
Repeated fixed-ratio steps reach very low concentrations with volumes you can actually pipette. Errors multiply rather than add, so a 1% error at each of six steps is about 6% at the end, not 1%.
Order of addition
Acid into water, never the reverse. Dilution is exothermic, and adding water to concentrated acid puts the heat into a small volume at the surface, which can boil and spit.
Temperature
Rate, temperature and storage
Why cold storage works, and why the number on the freezer matters more than the number of days.
The Arrhenius relationship
k = A·e−Ea/RT. Reaction rate depends on temperature exponentially, not linearly — which is why a modest drop in temperature buys a disproportionate amount of stability.
The Q₁₀ rule of thumb
As a rough guide, reaction rate roughly doubles for each 10 °C rise. It is an approximation, not a law, and it breaks down across phase changes and wherever the mechanism itself changes.
Freeze–thaw
Damage accumulates per cycle, not per unit of time frozen. Aliquoting so each vial is thawed once is usually worth more than a colder freezer.
Lyophilisation
Removing water by sublimation under vacuum. Water is what most degradation routes need, so a dry solid is generally far more stable than the same material in solution.
Light and oxygen
Photodegradation and oxidation act independently of temperature. Amber glass and a sealed headspace address problems that a freezer does not.
Shelf life is conditional
A stability figure is only meaningful alongside the conditions it was measured under. “24 months” with no temperature attached states nothing testable.
Measurement
Describing a set of results
A single number is a claim. A number with a spread and a sample size is a measurement.
Mean and standard deviation
The mean says where the results sit; the standard deviation says how tightly. Reporting one without the other hides whether the mean means anything.
Relative standard deviation
RSD is the standard deviation as a percentage of the mean, which makes spread comparable between measurements of very different magnitudes.
n matters
Three replicates and thirty replicates can produce the same mean and the same SD while deserving very different confidence. Sample size belongs in the report.
Blank vs control
A blank contains everything except the analyte and measures the background. A control contains a known amount and checks the method still works. They answer different questions and neither substitutes for the other.
Systematic vs random error
Random error scatters results around the true value and averages out with more replicates. Systematic error shifts every result the same way and does not — more measurements make it look more precise, not more correct.
Outliers
Discarding an inconvenient result without a documented reason is how bias enters quietly. Either the exclusion has a cause you can state, or the point stays in.
Common questions
Analytical methods FAQ
What does HPLC actually measure?
It separates a mixture by how strongly components partition between a mobile and a stationary phase, then measures what comes off the column. It answers how much of what was detected is one substance. It does not identify that substance, and anything that does not elute or does not absorb at the detector's wavelength is invisible to it.
What does a purity percentage not tell you?
A figure from a chromatogram means that proportion of the signal the detector saw was one peak. It is silent on anything the detector could not see — residual solvent, water, counter-ions, inorganic salts. Purity, identity and content are three separate questions and one number answers only one of them.
How does mass spectrometry differ from NMR?
Mass spectrometry ionises molecules and sorts them by mass-to-charge ratio, answering what mass is present — strong evidence of identity, though isomers share a mass. NMR reads how nuclei behave in a magnetic field and reports on their chemical environment, answering how the atoms are connected. NMR is far less sensitive and needs much more material.
Why is content found sometimes higher than the labelled amount?
Containers are often deliberately overfilled. A figure above the label means more material is present than stated, which covers residue left behind during reconstitution and transfer so the usable amount still meets the label.
What is the difference between molarity and molality?
Molarity is moles per litre of solution and shifts with temperature, because volume expands. Molality is moles per kilogram of solvent and does not, since mass is temperature-independent. For careful work across temperatures, molality is the safer measure.
Why do exact constants have no uncertainty?
Since the 2019 redefinition, seven constants are fixed by definition rather than measured. The uncertainty moved: instead of measuring the Planck constant against a kilogram, the kilogram is now realised from a Planck constant that is exact by fiat. The error bars sit on the realisation, not on the number.
Why is the molar volume of a gas sometimes 22.4 L and sometimes 22.7 L?
Because the definition of standard conditions changed. IUPAC's current STP is 0 degrees Celsius and 100 kPa, giving 22.711 litres per mole. The 22.414 figure belongs to the older definition at 101.325 kPa, one standard atmosphere. Both are correct for their own conditions, which is why the conditions have to be stated.
What is the difference between accuracy and precision?
Accuracy is how close a result is to the true value. Precision is how closely repeated measurements agree with each other. A method can be precisely wrong: tight, repeatable results clustered around the wrong number. Only a reference standard reveals that, which is why calibration is not optional.
Why does a buffer stop working outside a narrow pH range?
A buffer works because both a weak acid and its conjugate base are present in useful amounts. Move more than about one pH unit from the pKa and one of the two is nearly exhausted, so there is nothing left to absorb further acid or base and the pH moves almost freely.
Does storing something colder always make it more stable?
Usually, but not always, and not without limits. Rate falls exponentially with temperature, so cold helps a great deal. But freeze-thaw damage accumulates per cycle, some materials are damaged by freezing itself, and light and oxygen degrade material at any temperature. A freezer addresses one route, not all of them.