Work out the baking-soda weight for your citric acid, how much CO2 the batch makes, and how many days it lasts at your bubble rate — with a yeast-bottle comparison on the side, so you mix the right ratio the first time instead of guessing. Everything runs in your browser; nothing is saved.
How it works
Citric acid and baking soda react to release CO2. The balanced reaction is C₆H₈O₇ + 3 NaHCO₃ → Na₃C₆H₅O₇ + 3 H₂O + 3 CO₂: one molecule of citric acid needs three of baking soda and makes three of CO2. The weights follow from the molar masses — citric acid 192.12, baking soda 84.007, CO2 44.01.
Baking soda. Baking soda (g) = citric acid (g) × 252.02 ÷ 192.12, or citric acid × 1.312. The 252.02 is three baking-soda molecules (3 × 84.007).
CO2 made. CO2 (g) = citric acid (g) × 132.03 ÷ 192.12, or citric acid × 0.687. The 132.03 is three CO2 molecules (3 × 44.01). Gas volume in litres = CO2 grams ÷ 1.84, because CO2 gas weighs 1.84 g per litre at 20 °C.
How long the batch lasts. Daily use (g/day) = bubbles per second × bubble size (mL) × 1.84 × hours per day × 3.6. The 3.6 converts seconds to hours and millilitres to litres (3,600 ÷ 1,000). Days = CO2 made ÷ daily use. You need a bubble rate and hours for the days figure — without them the recipe is shown but no days estimate.
Water. About 1 mL of water per gram of citric acid, and about 2 mL per gram of baking soda as a slurry.
Yeast comparison. A yeast bottle ferments sugar: sucrose → 4 CO2 gives 0.514 g of CO2 per gram of sugar at full yield (4 × 44.01 ÷ 342.30). Real bottles run at roughly half yield, so this tool uses 0.514 × 0.5 = 0.257 g of CO2 per gram of sugar, and says so.
The bubble-size assumption. No published standard fixes a bubble-counter bubble at 0.1 mL. Forum measurements put real bubbles anywhere from about 0.05 mL to about 0.16 mL, so bubble size is a user-entered input here, defaulting to 0.1 mL and labelled as an assumption. Change it and the days figure changes with it — the recipe (soda and CO2 made) does not depend on bubble size at all.
Worked example
200 g citric acid, 2 bubbles per second, 9 hours a day, 0.1 mL bubbles.
Citric acid 200 g. Bubble rate 2 bps. CO2 hours 9 per day. Bubble size 0.1 mL.
Baking soda = citric × 1.312. CO2 = citric × 0.687. Litres = CO2 ÷ 1.84. Daily use = bps × bubble mL × 1.84 × hours × 3.6. Days = CO2 ÷ daily use.
Baking soda = 200 × 1.312 = 262.4 g, so 262 g. CO2 = 200 × 0.687 = 137.4 g, so 137 g. Litres = 137.4 ÷ 1.84 = 74.7 L, so 75 L. Daily use = 2 × 0.1 × 1.84 × 9 × 3.6 = 11.9 g/day. Days = 137.4 ÷ 11.9 = 11.5 days. Water: 200 mL for the citric bottle, 525 mL for the soda bottle.
262 g baking soda + 200 g citric acid makes 137 g of CO2 (75 L), lasting about 11.5 days at 2 bps, 9 hours a day.
Working back from a target works the same way. To make 100 g of CO2, start from 146 g of citric acid + 191 g of baking soda (100 ÷ 0.687 = 145.5 g citric; 145.5 × 1.312 = 190.9 g soda). And the yeast comparison: a 2 L bottle with 400 g of sugar makes about 103 g of CO2 (400 × 0.257 = 102.8 g), with about ½ teaspoon of yeast and ¼ teaspoon of baking soda.
How to read your results
The headline is the recipe: grams of citric acid in, grams of CO2 out. The stat grid shows the three numbers that decide the batch: baking soda in grams (citric × 1.312, with the water for both bottles), CO2 made in grams and litres, and how many days the batch lasts.
If the days stat shows a dash, you left the bubble rate or the hours blank — the recipe is still right, there is just no days estimate. Add both and the estimate appears.
Read the days figure as an estimate, not a promise. It moves with the bubble-size assumption: if your batch runs out much faster than the estimate, your bubbles are bigger than 0.1 mL. The recipe itself never moves — soda and CO2 made do not depend on bubble size.
If you entered sugar, the yeast row shows what the same bottle would make as a yeast system — about 103 g of CO2 from 400 g of sugar — so you can compare the two methods on numbers, not habit.
Common mistakes
- Sealing a citric system without a safety valve. DIY CO2 runs under pressure — use pressure-rated bottles, tubing and fittings only, with a gauge and a safety valve.
- Using an ordinary drinks bottle. They are not pressure-rated and can fail.
- Skipping the check valve. When pressure drops, tank water can siphon back into the bottles.
- Guessing the ratio instead of weighing it. A wrong ratio leaves acid or soda unused — and unused acid keeps reacting unpredictably.
- Treating the days figure as exact. It depends on the bubble-size assumption (0.05–0.16 mL in practice), so it is an estimate either way.
- Expecting a yeast bottle to run steady. Yeast output rises and falls over 2–3 weeks, unlike a citric system, and fades as the sugar runs out.
- Filling a yeast bottle to the top. Fill it to about three-quarters full.
- Using non-food-grade citric acid or baking soda, or mixing without eye protection. Add the citric solution slowly.
Batch sizes at 2 bps, 9 hours a day, 0.1 mL bubbles
Every row is the same recipe scaled up, at the worked-example rate.
| Citric acid | Baking soda | CO2 made | Gas volume | Water (citric + soda) | Batch lasts |
|---|---|---|---|---|---|
| 50 g | 66 g | 34 g | 19 L | 50 mL + 131 mL | 2.9 days |
| 100 g | 131 g | 69 g | 37 L | 100 mL + 262 mL | 5.8 days |
| 200 g | 262 g | 137 g | 75 L | 200 mL + 525 mL | 11.5 days |
| 300 g | 394 g | 206 g | 112 L | 300 mL + 787 mL | 17.3 days |
Days scale with the batch: double the citric acid and the batch lasts about twice as long at the same bubble rate. Change the bubble rate, hours or bubble size and the days column changes — the recipe columns do not.
FAQ
Where does the 1.312 ratio come from?
From the balanced reaction and the molar masses. One citric-acid molecule (192.12) needs three baking-soda molecules (3 × 84.007 = 252.02), so the weight ratio is 252.02 ÷ 192.12 = 1.312. The CO2 ratio is the same idea: three CO2 molecules (3 × 44.01 = 132.03) over 192.12 gives 0.687.
Why does the days figure change when I change the bubble size?
Because daily use is bubbles per second × bubble size × hours. A 0.2 mL bubble carries twice the gas of a 0.1 mL bubble at the same bubble rate, so the batch lasts half as long — 5.8 days instead of 11.5 for the 200 g example. The recipe does not change: the same 200 g still makes 137 g of CO2.
Can I use any bottle for a citric CO2 system?
No. Use pressure-rated bottles, tubing and fittings only, with a gauge and a safety valve — never seal a citric-acid system without one. Ordinary drinks bottles and fittings can fail under pressure.
Citric or yeast — which lasts longer?
On the numbers, citric: 200 g of citric acid makes 137 g of CO2, while 400 g of sugar in a yeast bottle makes about 103 g. Citric also runs at a steadier rate until the batch ends; yeast output rises and falls over 2–3 weeks. Yeast needs no pressure-rated parts, which is its real advantage.
What if I only know how much CO2 I want?
Leave the citric acid blank and enter the target instead. To make 100 g of CO2 you need about 146 g of citric acid and 191 g of baking soda.
Does the water amount matter?
It is a recipe convention, not a stoichiometric rule: about 1 mL of water per gram of citric acid, and about 2 mL per gram of baking soda as a slurry. For the 200 g example that is 200 mL for the citric bottle and 525 mL for the soda bottle.
Does it have to be food-grade?
Yes — food-grade citric acid and baking soda only. Wear eye protection when mixing, add the citric solution slowly, and keep the bottles where a leak cannot drain into the tank.
Sources
The research notes for this tool record no URLs for these sources, so they are listed by name here rather than linked.
- Haynes (ed.), CRC Handbook of Chemistry and Physics — molar masses: citric acid 192.12, NaHCO₃ 84.007, CO₂ 44.01, sucrose 342.30; CO₂ gas density 1.84 g/L at 20 °C (Reference).
- Engineering Toolbox — CO₂ gas density 1.842 kg/m³ at 20 °C, confirming the 1.84 g/L used here (Reference).
- Balanced reactions: C₆H₈O₇ + 3 NaHCO₃ → Na₃C₆H₅O₇ + 3 H₂O + 3 CO₂; C₁₂H₂₂O₁₁ + H₂O → 4 C₂H₅OH + 4 CO₂, giving 0.514 g CO₂ per gram of sucrose at full yield (Chemistry).
- UKAPS forum measurements — real bubble-counter bubbles run roughly 0.05–0.16 mL, which is why bubble size is a user-entered assumption here (Hobby practice).