Argon vs CO₂ vs Tri-Mix: How to Choose a Shielding Gas

What a Shielding Gas Actually Does in the Arc

In any gas metal arc welding (GMAW) or gas tungsten arc welding (GTAW) process, the shielding gas has one non-negotiable job: displace atmospheric oxygen and nitrogen from the molten pool so they cannot form porosity, oxides, and brittle nitrides. But that is only the baseline. The gas also becomes the conductive plasma column through which the arc runs, so its physical properties directly shape how the arc behaves.

Two properties dominate. The first is ionization potential, the energy required to strip electrons from the gas atoms and make them conduct. Argon ionizes easily (about 15.8 eV), which produces a smooth, easily started, forgiving arc. Helium is much harder to ionize (about 24.6 eV), so a helium arc carries more energy and runs hotter and broader for a given current, but it needs higher voltage and is less stable. The second property is thermal conductivity, which governs how heat spreads from the arc column into the plate. CO2 and helium conduct heat aggressively, widening penetration; argon conducts poorly and concentrates heat in a narrow finger.

The last lever is chemistry. Argon and helium are inert, they do not react with the weld metal. CO2 and oxygen are oxidizing, they dissociate in the arc and feed reactive oxygen into the pool. A controlled amount of that oxidation is actually desirable in steel welding: it stabilizes the arc, wets the toe of the bead, and reduces the ropey, high-crowned bead that pure argon produces on carbon steel. Too much of it, and you burn out alloying elements and generate spatter. Reading a gas selection chart is really reading a balance between these forces.

• Ionization potential: Lower (argon) means an easy, stable, low-voltage arc; higher (helium) means a hotter, broader, harder-to-start arc that needs more voltage.
• Thermal conductivity: High-conductivity gases (CO2, helium) broaden the heat pattern and deepen penetration; argon concentrates it into a narrow finger.
• Oxidizing potential: CO2 and O2 additions feed reactive oxygen into the pool, stabilizing the arc and improving wetting on steel, at the cost of some spatter and alloy loss.

Pure CO2: Cheap, Deep, and Spattery

One hundred percent carbon dioxide is the workhorse of production steel fabrication where cosmetics do not matter, and it is the cheapest common shielding gas by a wide margin. Because CO2 dissociates in the arc and conducts heat well, it delivers the deepest, broadest penetration of any common gas, which is exactly what you want on thick, dirty, or rusty structural steel. Fillet welds on heavy plate with an E70T or ER70S wire under straight CO2 fuse into the root with confidence.

The trade-offs are real. Pure CO2 will not support the smooth axial spray transfer mode, so you are limited to short-circuit transfer (thin material, out of position) or a coarse, hot globular transfer at higher currents. Globular means large, irregular droplets crossing the arc, which produces heavy spatter, a rougher bead, and more post-weld cleanup. CO2 is also a heavier-than-air asphyxiant that pools in pits, tanks, and low confined spaces, so OSHA 1910.146 confined-space procedures and ventilation matter.

Where CO2 shines economically: structural steel shops, farm and equipment repair, and any high-deposition job on 1/4 inch and thicker material where a wire brush or a grinder handles the spatter. Solid ER70S-3 and ER70S-6 wires run fine on straight CO2, and self-shielded and gas-shielded flux-cored wires (E71T-1C is classified specifically on CO2 per AWS A5.20) are often optimized for it. Just do not expect a chrome bead.

• Best for: Thick structural steel, heavy fillets, rusty or mill-scale plate, high-deposition production where appearance is secondary.
• Transfer modes: Short-circuit and globular only, no true spray. Expect noticeable spatter at globular currents.
• Watch out: Heavier than air, accumulates in confined spaces. Follow OSHA 1910.146 and ensure ventilation. Not for aluminum or stainless cosmetic work.

C25 and Argon-Rich Blends: The Everyday Sweet Spot

C25, meaning 75% argon / 25% CO2, is the default MIG gas for carbon steel in the majority of general fabrication and job shops across the US, and for good reason. The 75% argon component enables a stable, low-spatter arc and cleaner bead appearance, while the 25% CO2 keeps enough penetration and arc stabilization to weld real-world steel that is not laboratory-clean. On thin sheet and light structural work it is hard to beat for a balance of quality, speed, and cost.

As you lean the CO2 down, the arc gets softer and the transfer changes. Blends around 90/10 or 92/8 (often sold as C10 or C8) will support true spray transfer above the transition current, giving flatter beads and higher travel speeds on thicker steel, and they are common on robotic and high-volume lines. Add-oxygen blends replace CO2 with a small oxygen percentage: 98% argon / 2% oxygen and 95/5 are staples for spray-transfer and pulse welding on carbon and low-alloy steel, producing a fluid puddle and excellent bead wetting. For austenitic stainless steel, a tri-component blend of roughly 90% helium / 7.5% argon / 2.5% CO2 (a stainless tri-mix) or a low-CO2 argon blend keeps carbon pickup low so you do not sensitize the corrosion resistance.

The practical rule counter staff give: if you run a single steel MIG gas for a mixed shop, buy C25. If you move to spray or pulse for production or you care about minimal spatter and flatter beads, step to a 90/10 argon/CO2 or an argon/oxygen blend. Match the wire too: ER70S-6, with its higher silicon and manganese deoxidizers, is the forgiving default under all of these blends.

Pure Argon and When Steel Buyers Still Need It

One hundred percent argon is mandatory in two situations. First, virtually all gas tungsten arc welding (GTAW/TIG), regardless of base metal, runs on pure argon or an argon/helium mix, because the tungsten electrode and the fluid TIG puddle need a fully inert atmosphere. Second, MIG welding aluminum and most other non-ferrous metals requires straight argon; any CO2 or oxygen would oxidize the aluminum and wreck the weld. If you are running a 4043 or 5356 aluminum wire through a spool gun or push-pull, argon is the answer.

What straight argon is not good for is MIG welding carbon steel. Without any oxidizing component, the arc on steel becomes unstable, the puddle does not wet out, and you get a narrow, high-crowned, ropey bead with poor tie-in at the toes. That is exactly why the steel blends above add 2 to 25 percent of an active gas. Buyers occasionally try to run their aluminum argon on steel to save a cylinder, and the results are consistently poor.

For thick aluminum, distributors often recommend an argon/helium blend such as 75/25 or even higher helium to add heat and improve fusion and travel speed on heavy sections, since aluminum's high thermal conductivity otherwise robs the arc of heat. That extra helium comes at a cost, which we cover next.

• TIG / GTAW: Pure argon (or argon/helium) on essentially all metals; the tungsten and fluid puddle require full inertness.
• Aluminum MIG: Straight argon for most thicknesses; add 25 to 75 percent helium on heavy aluminum for more heat and better fusion.
• Do not: Run pure argon for steel MIG. Without an oxidizing addition the arc wanders and the bead is ropey with poor wetting.

Tri-Mix (Helium / Argon / CO2): Specialty Performance

Tri-mix means a three-gas blend, and the term is used two different ways depending on the base metal, which trips up a lot of buyers. For stainless steel, tri-mix usually means a helium-rich blend around 90% He / 7.5% Ar / 2.5% CO2. The high helium content adds heat for good fusion, the small argon fraction stabilizes the arc, and the tightly controlled 2.5% CO2 gives just enough arc stability and wetting while keeping carbon pickup low enough to protect the alloy's corrosion resistance. It excels at short-circuit and out-of-position stainless work on 300-series material with an ER308L, ER309L, or ER316L wire.

For carbon and low-alloy steel spray and pulse welding, a different tri-mix family exists: argon-rich blends with small helium and CO2 additions (for example around 85 to 90% Ar with helium and 2 to 5% CO2) that boost deposition and travel speed on thicker sections while keeping spatter low. High-nickel alloys and some duplex stainless procedures also call out specific tri- and quad-mixes in the welding procedure specification (WPS).

The reason tri-mix costs more is helium. Helium is a finite, extraction-limited gas subject to real supply volatility, and it is expensive to compress and ship. A tri-mix is also a precision blend, so it is typically mixed to tolerance rather than pulled off a bulk tank, which adds cost. The payoff is measurable: faster travel, better fusion on conductive or thick material, and, for stainless, metallurgical protection you cannot get from C25. Always weld to the gas your WPS or the filler manufacturer specifies, because on code work (AWS D1.1 for steel, D1.6 for stainless, ASME Section IX for pressure work) the shielding gas is an essential variable that can require requalification if you change it.

• Stainless tri-mix: ~90% He / 7.5% Ar / 2.5% CO2. Heat from helium, stability from argon, minimal carbon pickup from tightly held CO2. Great out-of-position on 300-series.
• Steel spray tri-mix: Argon-rich with small He and CO2 additions to raise deposition and travel speed on thicker carbon/low-alloy steel.
• Cost driver: Helium supply is constrained and precision blending adds cost, so tri-mix runs well above C25 per cubic foot.

Cost, Cylinders, and Buying Smart

Shielding gas is sold two ways: you own the cylinder (fill-and-return) or you lease/rent it from the distributor. Most small shops and hobbyists rent; high-volume users often own or move to bulk liquid. A common industrial size is the 251 cubic foot high-pressure cylinder (often labeled a size 300 or 'Q' bottle, roughly 2,200 to 2,400 PSI full), with 125 and 80 cubic foot bottles common for smaller shops and portable rigs. Cylinder rent typically runs about $80 to $250 per year depending on size and region, and that lease cost, not the gas, is what quietly drives your true cost per cubic foot.

As a rough national guide for 2026, straight CO2 is the cheapest, on the order of roughly $0.02 to $0.05 per cubic foot delivered in a bottle; C25 and argon/oxygen steel blends fall around $0.05 to $0.12 per cubic foot; pure argon runs roughly $0.08 to $0.16; and helium-bearing tri-mixes commonly land in the $0.20 to $0.45 per cubic foot range or higher, driven by helium markets. These are order-of-magnitude figures, your local distributors set actual pricing and it varies across the US, but the ranking almost never changes: CO2 < steel blends < argon < tri-mix.

Two buying levers save real money. First, right-size your flow. Most MIG runs well at 20 to 35 CFH (cubic feet per hour); cranking the regulator to 50-plus CFH does not shield better, it creates turbulence that pulls air into the pool and wastes gas. In a draft-free shop, 25 to 30 CFH is a fine default. Second, if you burn more than roughly a couple hundred cubic feet a week, ask your distributor about bulk liquid cylinders (dewars) or micro-bulk, where the per-cubic-foot cost of argon and blends drops substantially and you stop swapping bottles constantly. Also confirm your cylinders carry current DOT hydrostatic test dates (typically every 5 or 10 years by cylinder type) and that fittings meet the correct CGA connection (CGA-580 for inert argon/helium/nitrogen, CGA-320 for CO2), since the distributor will not fill an out-of-date or mismatched bottle.

Matching Gas to Process and Metal: A Decision Framework

Start with the metal, then the process, then the position, and the gas usually selects itself. Carbon steel with a MIG gun in a general shop: C25 for short-circuit on thin and medium plate, step to 90/10 or argon/oxygen if you run spray or pulse for production. Carbon steel where cost and penetration beat looks: straight CO2. Stainless steel MIG: a stainless tri-mix (90/7.5/2.5) or a low-CO2 argon blend to protect corrosion resistance; stainless TIG uses pure argon. Aluminum MIG: pure argon, adding helium on heavy sections; aluminum TIG uses pure argon (AC).

Transfer mode is the second filter. Short-circuit transfer (low voltage, thin material, all positions) tolerates higher CO2 content and is where C25 and CO2 live. Axial spray transfer (high current, flat and horizontal, flat clean beads) requires at least about 80% argon, so it needs C10, argon/oxygen, or argon-rich tri-mix, never straight CO2. Pulsed spray (pulse MIG) gives spray-like quality at lower average current for out-of-position and thinner material, and likewise wants an argon-rich blend. If you do not know your transfer mode, look at your voltage and wire, or ask your welding supply rep to match gas to your machine's synergic program.

Finally, respect the paperwork. On any code or inspected work, the shielding gas is spelled out in the WPS and is an essential variable under AWS D1.1, D1.6, and ASME Section IX, so substituting a cheaper blend can invalidate qualification. Cross-check the filler metal too: AWS A5.18 (solid carbon steel wire like ER70S-6), A5.28 (low-alloy wire), A5.9 (stainless wire), and A5.20 (flux-cored) classify wires against specific shielding gases, and running off-spec gas can move your mechanical properties out of the classified range. When in doubt, the filler manufacturer's data sheet names the qualifying gas.

• Pick order: Metal first, then transfer mode, then position. The gas usually falls out of those three answers.
• Spray transfer rule: True axial spray needs roughly 80%+ argon. Pure CO2 cannot spray; it stays short-circuit or globular.
• Code work: Shielding gas is an essential variable in the WPS under AWS D1.1 / D1.6 and ASME Section IX. Do not swap blends without checking requalification.

Frequently Asked Questions