Glass Jar Closures Explained: Screw, Lug and CR Caps

Buyers spend hours on the jar and five minutes on the cap. That is backwards. The closure is what keeps the product sealed, compliant and alive — and it is where most packaging failures actually happen. Here are the three closure families and how to read the numbers on a spec sheet.

The three closure families

1. Continuous thread (CT) screw caps

The most common closure in packaging. A continuous spiral thread engages as you turn. It opens and recloses reliably dozens of times, which makes it the right choice for anything the customer will get into more than once — flower jars, concentrate jars, tincture bottles, food jars.

2. Lug / twist-off caps

A lug cap has short interrupted lugs rather than a full spiral. A quarter turn engages or releases it. The real reason to use lug closures is vacuum sealing after hot filling: the cap is applied while the product is hot, and as it cools the internal vacuum pulls the lid down and holds the seal. That makes lug caps standard for jams, sauces and preserves — and a poor fit for products that need to be reopened often.

3. Child-resistant (CR) caps

CR closures are two-piece: an outer shell that spins freely unless the user performs a deliberate action (push down and turn, or squeeze and turn), and an inner cap that actually engages the thread. They are mandated for certain cannabis and pharmaceutical products.

The critical point for buyers: CR performance is a property of the jar and closure together, not of the cap alone. A certified cap on an untested jar can fail. Always ask for the test report covering the exact combination you are buying.

How to read a neck finish code

You will see codes like 53-400, 58-400 or 18-415. They break down simply:

A 53-400 jar needs a 53-400 cap. It is not interchangeable with a 53-410 cap, even though both are "53 mm". Getting this wrong is one of the most common sourcing errors, and it usually shows up as leaks in transit rather than at the filling line.

Liner materials: the part that actually seals

For cannabis flower and concentrates, a foam liner with an inert facing is the usual answer: it compresses enough to seal against glass, and it does not interact with terpenes or oils.

Matching the closure to the product

How to check a seal actually works

1. Torque test. Measure the removal torque on a sample of filled, capped jars. It should fall inside the range the closure manufacturer specifies — too tight damages the liner, too loose leaks.
2. Inversion test. Fill, cap and invert for 24 hours. Any weeping at the thread is a liner or neck-finish problem.
3. Vacuum decay. For products that need a vacuum, measure whether pressure holds over time.
4. Transit simulation. Vibration and a drop test catch thread engagement issues that a bench test misses.
5. Repeat open / close. Cycle the closure the number of times a real customer would. CR mechanisms wear, and a cap that works on day one may not work on day fifty.

Closures for bottles: same logic, different hardware

Jars and bottles share the same closure families, but bottles add a few specialised options worth knowing before you specify:

Setting a torque target

Removal torque is the measurable proxy for a good seal. Rather than guessing, agree a numeric range with your supplier and test every run against it:

• Specify the application torque on the filling line, not just the removal torque.
• Re-measure removal torque after 24 hours, because liners relax once the initial compression is spent.
• Re-measure after transit simulation — vibration loosens caps over a pallet journey.
• Record the range in your specification so every future order is judged against the same number.

Common mistakes

• Ordering caps by diameter alone and ignoring the thread style
• Assuming a CR cap is certified for every jar it fits
• Using a PTFE liner where a cheaper foam liner would do — or the reverse, and watching the liner fail on an oily product
• Specifying a lug closure for a product that the customer opens repeatedly
• Never measuring removal torque, then discovering half a pallet leaks
• Forgetting that the closure colour has its own dye lot, separate from the glass