ART
BEAUTY & WELLNESS
CRAFT
CULTURE & HISTORY
ENTERTAINMENT
ENVIRONMENT
FOOD & DRINKS
REVERSE ENGINEERING
SCIENCES
SPORTS
TECHNOLOGY
WEARABLES
Building a Gas Chromatograph
Every separation so far has used a liquid to carry the sample. In 1952 Archer Martin and Anthony James — Martin again, eight years after paper chromatography — used a GAS instead, and the whole business got faster by a factor of a hundred.
A gas is thin. Molecules diffuse through it quickly, so they find the stationary phase, partition into it and come back out again thousands of times a second rather than thousands of times an hour. That speed is why a gas chromatograph resolves a mixture in minutes where a column takes an afternoon, and it is why the method now sits in every forensic, environmental and flavour laboratory on earth.
The catch is that the sample must survive being vaporised, so this works for what boils without decomposing: solvents, fuels, essential oils. You will build a working instrument from copper tube, an aquarium pump and a thermistor, and separate the components of a lavender oil you distilled yourself.
Advanced
2 days
Instructions
1
1
Understand the three parts before you build any of them
Understand the three parts before you build any of them
A gas chromatograph is an injector, a column and a detector, with a carrier gas flowing through all three and heat holding the column at a steady temperature.
The carrier gas does nothing chemically — it only pushes. The column is a long tube whose inside is coated with a high-boiling liquid: THAT liquid is the stationary phase and it does the separating. The detector reports when something other than pure carrier gas comes out the far end. Everything expensive about a commercial instrument is refinement of those three; nothing about them is beyond a workshop.
2
2
Make a packed column
Make a packed column
Crush firebrick, sieve it to a narrow size range, and soak the granules in silicone grease thinned with acetone so each one carries a thin film. Spread it out until the acetone has gone completely, then pack it into 2 metres of 4 mm copper tube, tapping constantly, and coil the tube.
The brick is only a support with a huge surface area; the grease film is the actual stationary phase. Sieving matters more than it sounds — mixed grain sizes pack unevenly, gas finds a fast channel, and every peak comes out broad. Plug both ends with a little glass wool so the packing cannot migrate into your pump or your detector.
Materials for this step:
Copper Tube2 metre
Firebricks1 piece
Vacuum Grease (Silicone)50 g
Acetone100 ml
Glass Wool Insulation1 packTools needed:
Digital Scale
Nitrile Gloves3
3
Heat the column evenly, and measure the temperature you actually have
Heat the column evenly, and measure the temperature you actually have
Wind nichrome wire evenly along the coiled column, insulate it, and drive it from a variable supply. Bond a thermocouple to the tube itself, not to the air near it.
Retention time depends steeply on temperature — roughly halving for every 20 to 30 °C you add — so a column that is hotter at one end than the other smears every peak, and a column whose temperature drifts between runs gives retention times you cannot compare. Even winding and a real measurement of the metal are what make the instrument repeatable rather than merely working.
Materials for this step:
Nichrome Wire5 metreTools needed:
DC Power Supply
K-Type Thermocouple with MAX6675 Module
Multimeter4
4
Build a thermal conductivity detector
Build a thermal conductivity detector
Mount two matched thermistors in small chambers, one in the gas leaving the column and one in carrier gas that has bypassed it, and wire them as two arms of a bridge. Feed the bridge a steady low voltage and read the imbalance.
Each thermistor is self-heated by the current through it and cooled by the gas around it. Pure carrier gas cools both equally and the bridge sits balanced. The moment a different molecule emerges, the gas around one thermistor conducts heat differently, that bead changes temperature, its resistance changes and the bridge tips. This is a universal detector — it sees anything that is not the carrier — and it was the detector on the first commercial instruments.
Materials for this step:
NTC Thermistor Kit1 kitTools needed:
DC Power SupplyMultimeter
Alligator Clip Test Leads5
5
Set a steady carrier flow — steady matters more than fast
Set a steady carrier flow — steady matters more than fast
Run an aquarium pump through a ballast vessel and a needle valve into the injector end, and measure the flow at the outlet by timing a soap bubble up a graduated tube.
The ballast is what turns a pump's pulsing into a smooth stream, and a pulsing carrier makes the detector baseline oscillate so hard that small peaks disappear into it. Air works as a carrier for this build; it is not ideal, because oxygen slowly degrades the stationary phase at temperature and its thermal conductivity is close to many analytes, which costs sensitivity. Nitrogen is better if you have it, helium better still.
Tools needed:
Air Pump
Stopwatch
Borosilicate Beaker6
6
Inject fast, inject tiny
Inject fast, inject tiny
Fit a septum at the column inlet, heat that inlet above the boiling point of your sample, and inject a fraction of a microlitre with a fine syringe in one quick motion.
Everything about injection is about making the sample enter as a narrow plug. A slow injection feeds the column over several seconds and every peak is that many seconds wide before separation even begins. Too much sample overloads the stationary phase and gives leaning, tailing peaks whose retention times shift with how much you put in — which quietly destroys the identification the whole instrument exists to provide.
Tools needed:
Syringe - Laboratory7
7
Run a lavender oil, and read the chromatogram
Run a lavender oil, and read the chromatogram
Inject a dilute solution of lavender essential oil and log the detector voltage against time. You should see a solvent peak early, then a series of peaks over several minutes: linalool and linalyl acetate are the two largest components of real lavender oil.
Retention time identifies; peak area quantifies. Run a known pure compound to find its retention time, then look for that time in your unknown. This is exactly how a laboratory tells adulterated essential oil from genuine — a cheap oil cut with synthetic linalool shows the right peak in the wrong proportion to its neighbours.
Materials for this step:
Ethanol - Lab Grade, 95%50 mlTools needed:
Syringe - LaboratoryMultimeterStopwatch8
8
Van Deemter: why there is a best flow rate
Van Deemter: why there is a best flow rate
Loading Jupyter Notebook...
Tools needed:
StopwatchMaterials
8- 2 metrePlaceholder
- 1 piecePlaceholder
- Placeholder
- 100 mlPlaceholder
- 1 packPlaceholder
- 5 metrePlaceholder
- 1 kitPlaceholder
- Placeholder
Tools Required
10- Placeholder
- Placeholder
- Placeholder
- Placeholder
- Placeholder
- Placeholder
- Placeholder
- Placeholder
- Placeholder
- Placeholder
Related Blueprints
These blueprints share knowledge with this one — techniques, materials, or principles that connect them in the learning graph.
CC0 Public Domain
This blueprint is released under CC0. You are free to copy, modify, distribute, and use this work for any purpose, without asking permission.
Support the Maker by purchasing products through their Blueprint where they earn a Maker Commission set by Vendors, or create a new iteration of this Blueprint and include it as a connection in your own Blueprint to share revenue.