Weighing a reference standard: minimum weight and moisture

The weighing is a measurement with its own limit

Pharmacopoeial methods say that a substance is accurately weighed without saying how much of it counts. The definition lives in the balance chapter: a weighing is acceptable when the balance meets stated limits for repeatability and accuracy, and both limits are 0.10 per cent. Repeatability is assessed by weighing one test weight at least ten times; twice the standard deviation of those readings, divided by the smallest net weight the laboratory intends to use on that balance, must not exceed 0.10 per cent.

Rearranged, that criterion stops being a preference and becomes a floor. The smallest net weight must be at least 2000 times the measured standard deviation. A balance that repeats to 0.015 mg cannot support a 5 mg weighing, whatever its display resolution suggests.

A second floor sits underneath the first. Where the measured standard deviation falls below 0.41 of the scale interval, the chapter substitutes 0.41d on the grounds that no balance repeats better than its own resolution allows. Substitute that value and the floor becomes 820d:

The floor is a limit, not an achievement. A balance with a theoretical floor of 8.2 mg that repeats to 0.015 mg with the air conditioning cycling has a working minimum of 30 mg, and that is the figure that belongs in the method.

Why the limit bites harder for a standard than for a sample

A weighing error on a sample moves one result. A weighing error on a standard moves the assigned concentration of every solution made from it, and the error is fixed rather than random, so it does not average out across a batch.

At 5 mg the weighing is the dominant uncertainty in the whole chain, and at 1 mg the number on the certificate is decoration. The repair is usually not a better balance but a bigger portion: 25 mg into 50 mL followed by a dilution puts the weighing term near 0.1 per cent and costs a little extra standard, which is cheaper than repeating an analysis whose error cannot be recovered.

Weighing by difference, and where the material goes

Standard is weighed by difference: tare the closed vessel, add material, close, reweigh. The direct route, weighing into an open flask standing on the pan, hides three losses at once - powder left on the spatula, powder left in the funnel, and powder that leaves as dust when a static charge discharges. Weighing by difference records only what actually entered the vessel, and that is the quantity the calculation needs.

Two habits keep it honest. Nothing goes back into the original container; the remainder is discarded, because a spatula that touched the bench has contaminated the batch. And the vessel is closed before the reading is taken, since an open hygroscopic portion gains mass while the display settles.

Static deserves its own line. Dry, finely divided solids pick up charge and either cling to the receiver or scatter across the pan; the reading wanders for tens of seconds and never quite settles. Low relative humidity, plastic receivers, borosilicate glass and nitrile gloves all make it worse. An ionising bar or an antistatic boat fixes it, and a grounded metal holder helps by shielding the charge rather than dissipating it.

Use as is is an instruction, not a suggestion

Reference standard leaflets are unusually direct on this point: the closed container is allowed to equilibrate at ambient temperature before opening, so that moisture does not condense on cold material; the contents are used as is; they are not dried or desiccated before use; and once the container is opened, the entire content is used, because storage and re-use after opening are not covered by the producer's data. A reference standard is established for use with a named monograph, and the leaflet is part of how that use is defined.

The instruction follows from how the value was established. When a value is assigned on the material as received, the water that material already holds is inside the number. Dry it and a component the value was measured against has been removed; the assay then reads high by the amount of water driven off, which for a hydrate or a hygroscopic salt is commonly 0.5 to 5 per cent. Where a monograph does require drying before use, the certificate states the value on the dried basis and the drying condition becomes part of the method. The two practices are not interchangeable, and the document, not the habit of the laboratory, decides which one applies.

Equilibration is a real quantity. A vial taken from 5 degrees Celsius and opened at once collects condensate on the material; twenty to thirty minutes closed at room temperature is the usual allowance, and a vial from a freezer needs longer. Cold chain and stability covers the arrival side; the weighing room is where a handling mistake turns into a number.

Moisture, counted in micrograms

A deliquescent solid exposed in a weighing room at 50 per cent relative humidity can gain tenths of a per cent per minute in the first minutes after the container is opened. On a 10 mg portion, 0.5 per cent is 50 micrograms, five times the scale interval of a 5-place balance and a 0.5 per cent bias in the concentration of a 1 mg/mL solution. The practical answer is a time limit: open, transfer, close, record, with the open interval kept under a minute. If the procedure needs longer, weigh into a closed vessel and use the closed-vessel technique instead of accepting the drift. Hygroscopic peptide standards supplied as lyophilisates are the extreme case, and the leaflets for those materials say so explicitly.

The same physics runs in reverse for a warm container. A flask warmer than the room heats the air around it, that air rises along the wall, and the apparent weight falls by viscous friction. A cold vial does the opposite: it cools the air, the air sinks, and the reading creeps up. Neither is a balance fault, and neither is fixed by taring again.

Buoyancy: the 0.085 per cent nobody writes down

Balances are adjusted to read conventional mass: the mass that would balance the object in air of density 1.2 milligrams per cubic centimetre against weights of density 8.0 grams per cubic centimetre. Those reference conditions describe a stainless steel weight and not an organic powder. A solid near 1.2 grams per cubic centimetre displaces roughly seven times the air volume of the weight that balances it, so its true mass is about 0.085 per cent higher than the reading.

In a ratio assay the bias largely cancels, because standard and sample are weighed on the same balance in the same air. In absolute gravimetry, where a concentration is stated from the mass alone, it does not cancel, and 0.085 per cent sits close to the 0.10 per cent the balance chapter allows. It is worth knowing which situation you are in before adding a correction nobody else applies.

What the worksheet should carry

• Balance identifier, date of last calibration, and the standard deviation used to set its minimum weight.
• The minimum weight, and the net weight actually taken, so a reviewer can see that the second exceeded the first.
• Room temperature and relative humidity at the time of weighing.
• Whether the material was used as received, and how long it was equilibrated.
• Container type, and confirmation that the weighing was by difference.
• Any drift observed, and what was done about it.

None of this is exotic, and all of it is the difference between a result that can be defended two years later and one that has to be repeated.

Buyer checks before the balance becomes the problem

Most sub-minimum weighings start with an order rather than a balance. A 10 mg unit cannot be weighed on a 4-place balance, so buying the smallest pack to economise buys a problem at the bench.

• Ask for the unit quantity, and compare it with your balance's minimum weight, not with its readability.
• Ask whether the assigned value refers to the material as received or to a dried basis, and what the drying condition is if there is one.
• Ask for the storage condition and the equilibration instruction in writing, on a document that survives the packaging.
• Ask at what sample size homogeneity was demonstrated; a value established on 100 mg portions is not automatically the value of a 5 mg portion.
• Ask whether a larger unit is available for a hygroscopic material, since the extra material usually costs less than repeating the analysis.

Producers working to the reference material standard have to characterise the material, assess homogeneity and establish stability over the period they claim; the 2016 edition of that standard remains current, with a revision now in committee. None of it substitutes for the weighing, because the value on the certificate is a property of the material while the concentration in the flask is a property of your balance.

Symptoms and what they usually mean

Closing the loop

The certificate describes a material. The solution in the flask is a separate measurement, built from a net weight, a dilution and a balance that either supports the weight or does not. Reading the certificate settles the basis, qNMR against mass balance settles the purity, and salt, hydrate and free base conversion depends on both. When a result cannot be reconciled, check the order: portion size, then basis, then balance, then chemistry.