Deviant data? Your instrument might not be to blame

Environmental Conditions in the Lab

There’s always some variation between measurements, but when it goes beyond the norm, it can be a challenge to identify the cause. Small problems might go unnoticed only to be magnified at the end of a project. Error is compounded as it propagates through an experiment – into a screwed up spectrum, lanes lost from a gel, or a puzzling percent yield.

In the real world, most scientists focus on troubleshooting their technique. A logical first step, it accounts for the most of the wiggle room directly within your control. But the most overlooked (and underestimated) source of error has nothing to do with you or your instruments. Every measurement made in your lab is influenced by the environmental conditions: the temperature, humidity, and barometric pressure. We also include a discussion of electrical power for good measure.

The temperature directly influences the density, viscosity, reaction rate, enzyme activity, and other unique factors depending on the experiment. It’s important to maintain constant pressure in the lab for balance accuracy, sample consistency, and safety concerns. But few are aware of the widespread influence of humidity. It creeps into your samples, affecting long-term sample stability, and even catalyzing microbial growth. Moisture from the air can change the mass, density, and chemical structure of your supplies. Stockroom managers everywhere are likely shaking a fist at their lithium chloride, which deliquesces on a hair-trigger.

Micropipette as a case study

Temperature and humidity affect the density of the air inside the tip and plunger. Since the volume of fluid drawn into the tip depends on plunger displacement, the difference between a brisk morning and a rainy afternoon can mean a difference of up to 50%.

To further complicate matters, micropipettes commonly dispense samples ranging from 4 °C to 37 °C. The fluid’s temperature affects its own density, but there’s another hidden consequence: it will change the temperature of the air inside the pipette on contact. The resulting expansion or contraction introduces wide margins of error if lab conditions are unstable from day to day.

The sample’s temperature also affects its surface tension. This can have an unpredictable impact while dispensing the liquid, and creates inconsistency between experiments – especially when tip pre-wetting is practiced. And all of these inconsistencies would go completely unnoticed unless you were pipetting onto a microbalance and calculating the density at every step. Nobody has time for that.

Searching for Stability

Environmental conditions are a major consideration for any lab work – especially GLP work for publication or consumer products. But the good news is that stable conditions are within reach for any lab.

If you’re in a large facility with maintenance staff that already knows how to do this, it should be as simple as placing a work order. If not, there will be a learning process for you and your maintenance staff. Their role will involve a series of trial-and-error adjustments to the HVAC system. Throughout the process, you’ll need to ensure that any fume hoods continue to receive higher pressure than the rest of your lab (if the flow rate does drop, your fume hoods will make sure everyone knows about it).

Your role will be to collect time-stamped temperature, humidity, and pressure data for your lab so the maintenance staff can make better adjustments. While you can take the measurements yourself, we recommend an automated logger. You may eventually need to investigate a problem that happened after-hours, with a spike on the graph as your only clue. We can make an affordable recommendation if you don’t have one; and our clients can request courtesy loaner devices.

If this sounds like a lot of work, it certainly can be. And you may hit technological and interpersonal hurdles along the way. But Verdant can back you up every step of the way, or coordinate the entire process on your behalf. There’s no need to take on another headache just to have consistent data.


While not a true environmental factor, your electrical supply is a critical part of lab stability infrastructure. Your instruments have to contend with noise in the power source, interference from other devices, and power surges/outages.

Instruments which make sensitive electronic measurements should be connected to a power conditioner or conditioned wall outlet. These include mass spectrometers, TGAs, DSCs, certain types of balances – anything that reads a signal on the order of microvolts (µV). Power conditioners are special filtering circuits that smooth out the fluctuations present in grid power. Some instruments can’t function correctly without power conditioning, while others simply have a noisy baseline.

At a minimum, all instrument power and data cables should have a ferrite choke (below). These should look familiar – they come standard on AC power adapters. Ferrite chokes cost just a few dollars, and clamp onto the cable in seconds. Without one, every cable is essentially a radio antenna carrying interference into your instrument. These will eliminate most electromagnetic interference (EMI), but extreme sources may need to be tracked down with a spectrum analyzer.

Ferrite Choke

Shortcuts & Alternatives

If your facility can’t produce the special conditions you require (dry room, cold room, etc.), other options do exist. House nitrogen on a high purge flow through a draped work area can provide low humidity or low temperature conditions. We’ve also implemented very low cost glove-boxes (<$10,000) that maintain target conditions within 0.1% RH and 0.1 °C using self-contained environmental PID control. Fully enclosed devices such as this automatic pipette by Eppendorf can also provide exceptional reproducibility by eliminating the need to interact with the sample.

Environmental Controlled Automatic Pipette

When in doubt, reach out!

Every situation is unique, and we’re here to help. Verdant always provides Q&A support free of charge – even if you don’t have a service contract.