Glass colours differ mainly in where their cut-off sits. Ordinary flint stops only the deep ultraviolet below roughly 300 to 320 nm, a UV-absorbing clear glass removes roughly everything below 380 to 400 nm, and standard or dark amber carries on into the visible blue with a cut-off commonly around 450 to 500 nm. Green and cobalt blue fall between those extremes but on opposite sides of the spectrum, so they cannot stand in for each other or for amber. Start from the wavelength band that harms your product, pick the lightest route that excludes it, and ask for a measured transmittance curve at your wall thickness before you rely on it.
Why wavelength, not darkness, decides the damage
A photon carries more energy the shorter its wavelength. Ultraviolet and violet photons have enough to break chemical bonds in a product directly, while red light mostly arrives as heat. Blue and violet visible light sits between the two and is far from harmless, because several important molecules absorb strongly there.
The bands have conventional names. UVC runs from roughly 100 to 280 nanometres, UVB from roughly 280 to 315 and UVA from roughly 315 to 400. Visible light starts at about 400 and ends at about 700, with violet and blue at the low end, green in the middle and red at the top. Daylight reaching a bottle contains almost no UVC because the atmosphere filters it, so glass is not designed around it. The working targets are UVB, UVA and the neighbouring blue-violet band.
Four reactions account for most commercial light damage, and they do not respond to the spectrum in the same way:
- Photo-oxidation. Light starts the reaction and oxygen feeds it, so a lit bottle with air in the headspace ages faster than a lit bottle without.
- Fading. The colour of the product, a natural extract or a pigment shifts or disappears. Short visible wavelengths drive this as much as ultraviolet does.
- Vitamin degradation. Riboflavin is the familiar case. Its absorption reaches into visible blue, so a bottle that passes blue light fails to protect it no matter how well it handles ultraviolet.
- Lipid rancidity. Oils and fats break down oxidatively, usually slowly, and an off note tends to appear before laboratory values have moved much.
Exposure matters as much as chemistry. In a closed carton in a dark warehouse a product ages by oxygen and temperature, and glass colour hardly counts. On an open retail shelf under display lighting the dose is continuous for weeks or months, and that shelf is often the brightest place the pack ever sits. Most real shelf-life claims are settled in the mixed case, part warehouse and part shelf, where the dose is above what a storage study implies and below what a shelf study implies. That is also where colour is most often over-specified.
What each glass colour and add-on route blocks
The evidence behind any colour claim is a spectral transmittance curve. A sample of known wall thickness goes into a spectrophotometer, and the fraction of light that passes is recorded one wavelength at a time from the deep ultraviolet through the visible range. Buyers read two things from it. One is the cut-off, the wavelength below which very little light passes. The other is the shape of the curve above the cut-off, since a colour can be nearly opaque in one visible region and quite open in another, and the open region keeps ageing a blue-sensitive product.
The figures in the table are indicative planning bands based on how these glasses ordinarily behave. They are not specification values for any particular bottle. What governs is the measured curve for the exact colour and wall thickness you plan to buy.
| Route | What it holds back | What still passes | Typical contents | Gap it leaves | Confirm with the plant |
|---|---|---|---|---|---|
| Flint (untinted clear) | Only the deep end, below roughly 300 to 320 nm | Essentially all UVA, most UVB and the whole visible range | Contents with no meaningful light sensitivity that are judged by eye: water, clear spirits, white vinegar, some syrups, decorative and food packs sold on looks | UVA and visible blue. Almost no useful protection for vitamin-rich, oil-based or natural-extract products under shelf lighting | The curve for that exact flint body. Base composition varies by plant, and the cut-off shifts with iron content and wall thickness |
| UV-absorbing flint (clear glass with an ultraviolet absorber) | Roughly everything below 380 to 400 nm | The full visible range, so the pack still looks clear | Clear juices, clear cosmetic fluids, waters and spirits where a clear pack is part of the brand | Visible blue and violet, so a blue-sensitive vitamin such as riboflavin is not fully covered. The effect depends on the absorber and is lost if a different glass is supplied | Whether the absorber is in the body or in a coating, the wavelength at which it is declared effective, and whether more than one protection grade exists for the body |
| Light amber | The ultraviolet range and into the violet, cut-off around 400 to 450 nm | Most visible light from blue-green upward. The bottle reads warm, not dark | Oils, tinctures and extracts of moderate sensitivity where a lighter look is wanted, plus some food and drink packs that accept a warm tone | The blue-violet edge of the visible range and the upper part of UVA, depending on shade | The shade band on offer and the curve for the quoted shade. Light and dark amber are optically different products that share a name |
| Standard and dark amber | The ultraviolet range and most visible blue-violet, cut-off commonly around 450 to 500 nm | Green, yellow, orange and red. Very little blue | Essential oils, herbal tinctures, pharmaceutical liquids, light-sensitive nutraceuticals, botanical extracts, beer and dark spirits | Longer visible wavelengths, heat and oxygen. Light can still enter through an unlined cap or an uncovered shoulder | The amber composition, because iron and sulphur levels set the cut-off, and the curve at your wall thickness |
| Light green | Part of the ultraviolet range and a slice of blue-violet | Green strongly, yellow and much of the red. Blue is reduced, not removed | Drinks and food packs wanting a green identity with low to moderate sensitivity; wine and oil with a short to medium shelf life | Most visible blue and a substantial part of UVA. Far weaker than amber for a sensitive product | How the green is made. A chromium green and an iron green behave differently, so ask for the curve of the shade offered |
| Emerald and dark green | The ultraviolet range and a large part of visible blue, cut-off nearer the amber region | Green through red. Blue and violet are strongly reduced | Wine, spirits, olive oil and other drinks where dark green is acceptable and an amber look is not wanted | Green and longer wavelengths, heat and oxygen | The shade band and its cut-off, since emerald covers a wide family, and that the quoted shade is the one the curve was measured on |
| Cobalt blue | The ultraviolet range plus the yellow, orange and red part of the visible spectrum | Blue and violet, and some near ultraviolet that is not absorbed | Pharmaceutical and cosmetic packs where blue is the brand, or where the product is sensitive to the warm end of the spectrum | Blue and violet light, which makes it a poor fit for products damaged by short visible wavelengths | Body colour or coating, the exact shade, and the curve. A deep-looking cobalt can still pass a useful amount of blue-violet |
| Any colour plus an external UV-blocking lacquer or spray | Whatever the body blocks, plus what the film absorbs | The body's own transmission minus the film's share. A colourless film leaves a clear body looking clear | Packs whose body must stay light for brand reasons but need some ultraviolet protection | Visible blue, unless the film is tinted to absorb it. The film brings its own adhesion, abrasion and food-contact questions | Whether the film is meant to stay on a food-contact surface, how performance is checked after abrasion, and whether a combined film-plus-body curve is reported |
| Any colour plus a full-coverage opaque label, shrink sleeve or carton | All light, where coverage is complete | Nothing in the covered area; the body's own transmission at every gap, shoulder, base and cap area left bare | Designs that must stay unchanged while shelf life is extended | Light entering through the neck, base, cap or any other bare surface | Actual coverage as a percentage of surface area, which surfaces are bare, and whether the label supplier can confirm opacity across the relevant wavelengths |
| Opaque glass or opaque external coating | Essentially all ultraviolet and visible light | No useful light. The contents cannot be inspected without opening | Extremely light-sensitive products, or formats where opacity is part of the brand | Heat and oxygen. It also removes visual inspection | Whether opacity is in the glass or a coating, how a coating is tested for pinholes and coverage, and how the pack is inspected on the line |
Two patterns come out of the comparison. Among transparent options flint has the lowest cut-off and amber the highest, with green and blue in the middle but facing different ends of the spectrum. And a surface treatment adds to the body's curve without altering it, so a lightly tinted body with a good ultraviolet film can beat an unfilmed dark body in the ultraviolet and still look lighter in the visible range.
What amber does well and where it stops
Amber became the standard barrier colour for reasons beyond being dark. It absorbs the entire ultraviolet range and continues well into visible blue, which covers both the bond-breaking short wavelengths and the blue-violet band that harms a wider group of molecules. The iron and sulphur chemistry that produces it is well understood, so the colour can be held within a defined band from one campaign to the next, and a protection level that drifts is worse than a modest one that stays put. It is also among the most widely available tinted glasses, and consumers already read an amber pack as functional.
Its limits are equally specific. Red, orange and much of the green pass through, so a product degraded by those wavelengths gains nothing from moving out of clear glass. Amber does not remove dissolved oxygen, slow a heat-driven reaction or make up for a closure that admits air. When a switch to amber fails to lengthen shelf life, the limiting reaction was usually oxygen or temperature all along.
Geometry is the third limit. Colour protects only where it stands between the light and the product, and light also arrives through the base, the shoulder, the neck and the closure. An unlined cap on a narrow neck can admit more light to the headspace than most buyers expect. A dark body under a light-transmitting cap is not a fully protected pack, so where light is the limiting factor the cap liner and label coverage belong in the protection specification.
Protection in the glass compared with protection on the glass
There are two ways to build a light barrier. It can be a property of the glass, through a tint or through an ultraviolet absorber that leaves the visible appearance alone. Or it can be applied afterwards as a lacquer, spray, shrink sleeve, label or carton. The optics are sound in both cases. The choice turns on appearance, cost and how the result will be verified.
| Attribute | In the body | On the body |
|---|---|---|
| Durability | Cannot wear off. The material is the same through the wall, so abrasion, washing and handling do not remove it | A film can abrade, a sleeve can shrink unevenly, and a lacquer can be affected by the product on the fill line or by washing |
| Coverage | Every surface, including shoulder, base and neck | Only where the film or label sits. The least covered surface sets the protection of the whole pack |
| Adjustment | Predictably thickness-dependent, so a slightly thicker wall raises blocking when the colour is fixed | Leaves the glass appearance free, which suits a clear or lightly tinted brand look |
| Main constraint | The protective tint is the tint the consumer sees. Strong ultraviolet blocking with a clear look needs a colourless absorber to be available for that body | Has to be verified on the finished, handled pack, not on a flat sample |
A sleeve that wraps the body completely and leaves the neck and cap clear is a partial barrier. The same applies to any design gap in a label.
The two routes work best as one system. A light body colour with a well-fitted ultraviolet sleeve may protect better than either alone, and if the sleeve was going to be printed anyway the pairing can be more economical than a very dark body. Specify and verify each part separately, though. A supplier's curve for a coloured body tells you nothing about the pack once a film or label has been applied.
Ranking the product by light sensitivity
Sensitivity is the input to everything else, and no glass supplier can provide it for your formula. It comes from your own stability work, from ingredient supplier data, or from a working assumption you test later. Four broad tiers are normally enough for a first decision.
| Tier | How light affects it | Typical products | Starting assumption |
|---|---|---|---|
| Extremely sensitive | Light is a primary failure route, and change is often visible within weeks under shelf lighting | Many essential oils and aromatic materials, certain vitamins and vitamin-rich preparations, some natural colourants, a number of pharmaceutical and nutraceutical liquids | Full light protection, meaning amber or an opaque route, with a verification plan |
| Highly sensitive | Light shortens an otherwise acceptable shelf life, more slowly | Alcohol-based tinctures and herbal extracts, botanical oils, some cosmetic actives, several food oils | Amber by default. The open point is whether the shade can be lightened or a film can carry part of the load |
| Moderately sensitive | Light contributes to ageing without dominating it | Many beverages, wine, olive oil, a wide range of cosmetics | Colour is often chosen for identity, with protection as a supporting benefit |
| Low sensitivity | Light is not a meaningful route | Products where colour is essentially a branding choice | Clear glass is acceptable, provided oxygen and temperature are handled |
Putting a product in the wrong tier is the costliest mistake in this subject. It usually happens when sensitivity is inferred from the category name and never measured on the actual formula.
Working from a target shelf life to a colour band
With the tier set, four steps turn a shelf-life target into a colour band. The sequence matters more than any single figure.
- State the shelf life and the storage behind it. A twelve month target and a twenty-four month target are different problems, and each changes again if the paperwork assumes a dark warehouse while the real route includes months on a lit shelf. Record the worst-case exposure, not the average.
- Isolate the light dose. A study run in the finished pack already contains the light contribution and cannot show it separately. A study run in the dark does not cover light at all. A short accelerated exposure with and without light, using the same fill and closure, replaces the guess with a number and usually costs less than years of over-specified glass.
- Match the wavelength to a band. Use the table and work from the band to be excluded, not from a colour name. Damage concentrated in the blue-violet region points to amber, or to an ultraviolet-absorbing body with a visible-blue treatment. Damage concentrated in the ultraviolet can sometimes be handled by a lighter colour with a film.
- Verify on the real pack. Combine a transmittance measurement with a stability check on the filled pack. Do not assume the band delivers the result.
Two principles prevent most over-specification. Darker and more opaque glass shows less of the product, so protection trades directly against shelf visibility and the darkest option is seldom the right one. And a longer shelf life does not automatically call for a darker colour. If oxygen is the limit, more colour achieves nothing, while removing headspace or improving the closure achieves a great deal. Colour should be the last variable changed in a shelf-life programme, and the right band is the lightest one that passes verification.
Light is one of three ageing variables, alongside oxygen and heat. A colour does nothing about dissolved oxygen in the liquid, air in the headspace or a warm pallet in a hot container, which is why a shelf-life problem is not automatically a light problem. The combined picture, including closure and temperature, is set out in our guide to total shelf life of a filled glass pack.
Balancing protection against a visible product
The hardest constraint is commercial. Clear spirits, wine in clear or lightly tinted glass, cooking oils, vinegars, syrups, honey, preserved foods and many beverages sell partly because the buyer can see them. For these, protection has to work around transparency. The practical options are a light tint, an ultraviolet-absorbing clear body, partial coverage from a label, or a shelf and pack strategy that limits exposure.
How hard the constraint bites depends on the channel:
- Bright retail shelf. Visibility is worth most and exposure is highest, so the two requirements collide here more sharply than anywhere else.
- Trade and professional channels. A pharmacy counter, a spa back-bar or a mail-order pack often carries weaker visibility expectations and lower sustained exposure, which makes room for a darker or more protective colour.
- E-commerce. The pack is judged from photographs under studio light, so a dark bottle presents well on screen. The physical pack spends its time in transit and in a fulfilment centre, with less exposure than a shelf.
Visibility is therefore a property of the channel more than of the product. When a product must stay visible and is also highly sensitive, one glass colour often cannot meet both demands and the answer is a combination. An ultraviolet-absorbing clear or lightly tinted body carries most of the ultraviolet protection, and a partial label or printed band adds cover where the design permits. Do not count on a thin branding label to protect by accident. Coverage decides whether it helps, and a decorative label rarely reaches 100 percent.
What a transmittance report must state
A claim of ultraviolet blocking cannot be checked unless the document behind it states its conditions. A report missing any of the following cannot be compared with another supplier's:
- The full curve over a stated wavelength range that starts below the ultraviolet boundary and continues into the visible, so the blue-violet behaviour shows. A single "percentage blocked" is nearly always quoted at one wavelength, and a product can be well protected there and exposed twenty nanometres away.
- Wall thickness at the point of measurement. Transmittance changes with the amount of glass in the light path, so a curve from a thin section does not describe a thick body.
- Sample geometry. A flat cut section, a formed body or a section of the actual bottle give different results because curvature and surface finish play a part.
- Instrument and method, including whether the value is total transmittance or a directional measurement.
- The finished pack wherever a coating or label is claimed to contribute, with the film or sleeve and any clear areas included. A film measured on a flat panel is not enough.
Two more details make the report usable in a specification. The cut-off needs an explicit definition, because suppliers describe it differently and a quoted value means nothing until the definition is fixed. The component tested must be identified, since base glass composition varies between plants and between campaigns, and one plant's curve does not automatically describe another's.
You do not need your own spectrophotometer for any of this. You need a specific request and a written answer, and it is reasonable to make the curve a condition of the order. The relevant test methods are public. Naming them in the specification keeps both sides on one definition, without either side claiming a certification it does not hold.
What to send for a colour recommendation
A colour name is a poor starting point for a brief. Three inputs let us answer in wavelengths and routes and check them against plants that can supply a measured curve:
- The contents, by type and, where known, by what degrades. An oil, an alcohol-based extract, a water-based beverage, a vitamin preparation and a cosmetic fluid each respond differently to light. Existing stability data shortens the work considerably.
- The target shelf life and the exposure it assumes: lit shelf, chiller, counter or carton.
- The channel, which sets how much the pack must show and how much light it will receive.
From there the discussion moves to whether the body alone is sufficient, whether an ultraviolet-absorbing clear body fits better, whether a light tint with an external film is more economical than a dark body, and which measurement the finished pack should be verified against. Any figure given at that stage is a planning orientation, to be confirmed against a real curve and a real fill. If the route includes a coating or sleeve, raise adhesion and durability in the same conversation, because a barrier that fails in handling is no barrier.
Several neighbouring decisions have their own pages:
- Which colour to buy for a given product and channel, with its effect on cost, availability and stock, is covered in amber versus clear as a purchasing decision.
- The shades that actually exist are listed in the colour range across clear, amber, green and specialty tints.
- Judging one delivery against the next is explained in specifying and checking batch-to-batch shade consistency.
- Whether colour should be melted into the glass or applied to its surface is compared in raw material colouring versus surface spraying.
Questions buyers ask about UV protection in glass bottles
Does clear glass block any ultraviolet?
A little. Ordinary flint stops the deepest ultraviolet and lets through most UVA and UVB along with all visible light, which is far too narrow an effect for a sensitive product. Two clear bottles from different plants need not behave the same, because iron content and wall thickness move the cut-off.
Is amber fully opaque to ultraviolet?
For practical purposes amber solves the ultraviolet problem, since its cut-off typically lies well above the ultraviolet boundary and UVA and UVB are largely excluded. It is not a perfect barrier at every wavelength and it passes longer visible light freely. Whether amber is enough depends on whether visible blue also damages the product, as it does some vitamins and natural colourants.
Can a clear bottle block UV without changing colour?
Yes, by either of two means: a body made with an ultraviolet absorber, or a clear external lacquer or film that contains one. Neither normally suffices for a product harmed by visible blue, because protection in the visible range requires a visible tint. Ask which wavelength the protection is declared to reach, since "ultraviolet blocking" on a clear pack can describe very different performance levels.
Does a dark label or shrink sleeve add real protection?
Only where it covers. An opaque full-body sleeve blocks the whole spectrum in that zone, while an open neck, a clear base or an unprinted gap passes whatever the body colour passes. If the sleeve is doing real work, have its coverage measured by surface area and the bare surfaces listed.
Do blue and green bottles protect as well as amber?
Usually not, where sensitivity spans the blue and violet region, although both improve on clear glass for many products. Green cuts into blue-violet and passes green strongly. Blue passes blue and violet and blocks the warm end. Compare measured curves, not names: shade families overlap, and a deep green can sit near an amber while a light green sits far from it.
Does a thicker wall improve light protection?
Yes, gradually, because more glass in the light path absorbs more of the wavelengths the colour already absorbs. It is seldom the efficient route, since thickness also adds weight, cost and forming difficulty. A colour change, an ultraviolet-absorbing body or a film usually gives more protection for the same spend. Where the colour is fixed and the design is already heavy, wall thickness is a legitimate secondary lever, to be confirmed by measurement.