Age Verification

You must be of legal drinking age to enter this site.

Are you 19 years of age or older?

By entering this site, you agree to our Terms of Service and Privacy Policy.

Cans vs Bottles Research: What the Evidence Actually Shows

By Aura Editorial August 20, 2026
Cans vs Bottles Research: What the Evidence Actually Shows

You are standing at the cooler and the same style of drink is sitting there twice: once in a can, once in a bottle. Which package actually protects what is inside, and which one is the better environmental call? That is a practical question with measurable answers, so here is what the cans vs bottles research actually found — and, just as importantly, what it did not.

The short version: on light, cans win decisively. On oxygen, the deciding component is usually the closure, not the container material. On carbon, the answer flips depending on recycled content. And in Canada, both aluminium and glass are recovered at high rates — with glass ahead of aluminium in British Columbia's most recent published figures.

Light exposure: the clearest finding in the cans vs bottles research

Light damage in beverages is well characterised in brewing science. Riboflavin absorbs light in the 300–500 nm range and passes that energy along, kicking off reactions that produce 3-methyl-2-butene-1-thiol (MBT) — the "skunky" or light-struck compound. It is perceptible at astonishingly low levels: Gabriel and colleagues, writing in Kvasny prumysl in 2022, cite a detection threshold in beer of roughly 4–35 ng/L.

That team measured transmission spectra for commercial bottles and built a Packaging Riboflavin Index (PRFI), where a package that transmits everything scores about 100 and a package that transmits nothing scores 0. Their measured values:

  • Clear glass: PRFI 82.6 — essentially no protection.
  • Blue glass: 69.7.
  • Green glass: 18.1 to 57.5, depending on the shade. Some light green bottles transmitted about 45% of light at 450 nm; dark green under 10%.
  • Brown glass: 1.0 to 3.6 — between 23 and 83 times more protective than clear glass.

Note what is missing from that list: cans. The study tested glass and PET, not aluminium. But the authors define the scale themselves — if no critical radiation passes through the wall, PRFI is 0. An opaque aluminium can is that limiting case by construction. It is the only common beverage package where light exposure is not a variable you have to manage at all. A brown bottle is very good; a can is the boundary condition.

Oxygen permeability and shelf life: blame the closure, not the glass

Glass itself is not meaningfully permeable to oxygen. The seal is. Kajetan Müller's review in BrewingScience (2007) reports that a PVC-foamed crown cork passes 4.0 µg of oxygen per closure per day, which works out to 2.16 mg/L of oxygen ingress into a 0.33 L glass bottle over half a year (1.44 mg/L for a 0.5 L bottle). Barrier liners cut that by roughly an order of magnitude, to 0.22 and 0.14 mg/L respectively. Plain monolayer PET is far worse: about 10 mg/L in four months, against a practical brewing benchmark of 1 mg/L over six months.

So the honest reading is this: the glass wall is an excellent barrier, and the crown cap is the leak. A can's body and ends are a continuous metal barrier with one seamed joint, which is why cans are generally treated as the tighter package — but Müller's paper does not publish a comparable ingress figure for cans, so we will not invent one. What the data supports is narrower and more useful: if you are comparing sealed packages, compare the seals.

One caveat that does not apply to every drink: carbon dioxide loss through the closure is a genuine shelf-life problem for carbonated beverages. It is a non-issue for a still drink. Our vodka water cocktails have no carbonation to lose in the first place — the flavour and the alkaline water are what the package has to protect.

Beverage packaging life cycle: where the carbon argument gets complicated

Alice Brock and Ian Williams assessed six container types across three beverage categories in Detritus (2020, vol. 13, pp. 47–61). For pressurised drinks, their conclusion was that 100% recycled aluminium cans were the least impactful option assessed, with glass performing poorly because of the energy required to produce it.

Read that conditional carefully, because it carries the whole result: 100% recycled. Aluminium's advantage is entirely a recycling-loop story. Natural Resources Canada's Aluminum facts page states that secondary aluminium production requires 95% less energy than primary production. Canada is unusually well placed here: NRCan reports 10 primary smelters (one in Kitimat, BC and nine in Quebec) producing about 3.3 million tonnes in 2024, and credits Canadian producers with the lowest carbon footprint among major producers thanks largely to hydroelectricity. A can made from Canadian metal and fed back into the loop is a genuinely different object from a can smelted on coal power.

Aluminium can recycling Canada: the provincial numbers

Deposit-return performance is where the beverage packaging life cycle stops being theoretical. From Encorp Pacific's 2023 Annual Report for British Columbia: 1.375 billion containers recovered, a 79.6% overall recovery rate, with aluminium at 84.1% and glass at 89.9%. Encorp also reports that 100% of the aluminium it collected was sold to re-melt facilities to make sheet stock for new cans — a closed loop, not downcycling.

Notice that glass out-recovered aluminium in BC that year. That cuts against the tidy "cans are greener" story and it belongs in the record.

In Alberta, the Beverage Container Management Board reported a 2024 return rate of 83.1%, which it describes as the highest in Canada and second highest in North America.

What the research does not show

Several honest limits are worth stating plainly.

  • Cans do not always win on flavour stability. Fromuth and colleagues (ACS Food Science & Technology, 2023) tracked metabolite profiles in amber ale and IPA in cans and bottles over six months and found a strong package-type effect for the amber ale — which held up better in bottles — and little difference for the IPA, possibly because hop polyphenols buffer oxidation. Package effects are product-specific, not universal.
  • Small, narrow samples. The PRFI work characterised a handful of bottle types; the beer stability study used two beer styles. Neither is a general law about every beverage.
  • LCA results are assumption-driven. Change recycled content, transport distance or end-of-life collection rate and rankings move. The Detritus result is conditional, not absolute.
  • None of this is a health claim. Packaging affects light exposure, oxygen ingress and carbon footprint. It does not change the alcohol. The same standard drink is the same standard drink in either package — no version is healthier, safer or less intoxicating.

What to do with this today

  • If a drink will sit in daylight — a patio, a boat, a shop window — choose the opaque package. Light is the one variable where the evidence is unambiguous.
  • If you are buying glass, buy brown over green, and green over clear. The measured gap between brown and clear was 23–83x.
  • Store either package cool and dark. Temperature drives oxidation chemistry regardless of what the drink is in.
  • For sustainable cocktails at home, the highest-leverage step is not the material — it is returning the empties. An unreturned can forfeits the entire 95% energy saving that makes the LCA case work.
  • Check your province's deposit system: Encorp/Return-It in BC, the BCMB-regulated depot network in Alberta.

This is the reasoning behind our own format, and you can read more about it on why Aura: a 355 mL aluminium can, no light exposure, no carbonation to go flat, and a package that goes back into the deposit stream and returns as another can.

Ready to try it? Browse our vodka water cocktails in Cucumber Lime, Coconut Pineapple and Mango Peach, or find a shelf near you with where to buy — we're stocked across British Columbia, Alberta and Ontario. Please enjoy responsibly, and only if you are of legal drinking age in your province.

This article summarises published research for general interest and is not medical advice.

Sources

Back to Blog