A lot of boards arrives, half of it goes to the line, and the rest goes on a shelf. Six months later the second half comes out and the first panel fails on wetting at a fine pitch pad. Nothing was done wrong in production. The finish aged in a way nobody tracked.
PCB shelf life is a property of the surface finish and the storage, not of the board as a whole. Two panels from the same lot, one in a sealed bag with a humidity indicator and one in an open carton, are not the same material after a season in a warehouse.
What actually ages on a stored board
Four things move while a board sits. The pad finish oxidises or diffuses, which is a solderability problem at the joint. The laminate absorbs moisture, a blister problem at reflow. Flux and plating residues keep working on the surface, which shows up as staining and, on a fine pitch part, as leakage. And the packaging degrades: adhesive migrates, paper interleaves hold humidity, and a vacuum bag punctured in handling stops being a barrier without anyone noticing. Moisture is the one that turns a good board into scrap in one pass through the oven, because absorbed water flashes to steam faster than the resin relaxes.
Planning life by surface finish
The numbers below are the planning values used across the industry, and they are the ones to open a discussion with. The material datasheet and the actual storage fix the final number for a lot.
| Finish | Planning life in sealed storage | What fails first | What the storage has to do |
|---|---|---|---|
| OSP | Shortest of the common finishes, commonly planned at 6 months | Wetting at fine pitch pads; the coating is consumed by the first reflow | Sealed with desiccant, opened only for the run, used first in first out |
| Immersion tin | Around 6 months | Whisker growth and oxide at the pad edge | No sulphur bearing packaging, no rubber bands against the surface |
| Immersion silver | 6 to 12 months | Tarnish and sulphide staining, which reads as poor wetting | Sealed, low sulphur environment, no paper against the pads |
| ENIG | 12 months or more | Black pad on a badly controlled bath; corrosion stain in humid storage | Dry storage, and a bath that is controlled rather than assumed |
| ENEPIG | 12 months or more | Same as ENIG, with palladium slowing diffusion | Dry storage, same discipline as ENIG |
| Lead free HASL | 12 months or more | Thickness variation across a pad, a coplanarity issue rather than ageing | Dry storage; the finish itself is the most forgiving |
The point is not which finish is better. OSP suits a board built within a month and not one held a year as a spare, so matching finish to build schedule costs nothing and prevents the most common late lot argument.
The storage controls that hold up
Sealed packaging with desiccant and a humidity indicator card is the first control, and the card is the part that matters, because it makes the state of the bag visible instead of assumed. A bag opened for a partial build is no longer sealed, so it needs a reseal and a date.
The room matters as well. The published environment at the assembly end is 22 plus or minus 3 degrees Celsius and 45 to 65 per cent relative humidity, comfortable for people and stable for boards. A warehouse that swings with the season is not the same thing, which is why the bag exists. The third control is the date code, applied when the boards are packed rather than received, because a lot without a pack date can only be aged by guess, and the guess always favours keeping the boards. That discipline is set out under date code management.
Baking, and when it does more harm than good
Bake is the standard answer for a board that has taken up moisture, and it is right when done to the right schedule. The bake used here follows J-STD-033 at 125 degrees Celsius, with time set by board thickness and how long the material has been out of the bag, not by a fixed hour count.
Bake before assembly, never after. Baking a board that already carries parts drives flux residue to cure and can push moisture out through a joint rather than out of the laminate. On a board baked repeatedly the intermetallic layer at the pads grows, and a pad that has seen three bake cycles plus two reflows is a weaker joint than a pad that saw one.
When a lot is past its date
The decision is a test, not a judgement. A solderability check on a sample pad, wetting balance or a dip and spread, answers the question the finish was always about. Then a bake to the J-STD-033 schedule if moisture is the concern. Then a first article on one panel with the real profile, checked at the joints that were at risk: fine pitch pads, a large ground plane, a press fit hole. If the sample wets and the first article passes, the lot can be built with the bake recorded against the date code. If wetting is marginal, the lot is not scrap, but it is not a production lot either; it belongs in a build where rework is possible.
How SUPERB PCBA keeps a stored lot traceable
Boards and components are held in a controlled store at 22 plus or minus 3 degrees Celsius and 45 to 65 per cent relative humidity, with ESD control to ANSI/ESD S20.20 and a bake route to J-STD-033 at 125 degrees Celsius for anything that has been out of its packaging. Every lot carries a pack date and a location, so a partial build leaves a record of what was opened and when.
Assembly starts from that record. Boards are reflowed on a twelve zone nitrogen oven at under 500 ppm oxygen, inspected optically after reflow and by X-ray where a joint is hidden, and released to IPC-A-610 Class 2 or Class 3 as the drawing states. A bake or a solderability check is in the traceability report that ships with the lot, the same route described under component traceability report. Moisture sensitivity at the component level, including the floor life clock, is a separate control on MSL and bake control.
Questions buyers ask
Does a sealed bag stop the clock?
It slows it. A sealed bag with desiccant keeps moisture out, but the finish still ages, only more slowly. The bag is a control, not a reset; the date code keeps running.
Is a stained ENIG pad unusable?
Not automatically. A stain is a surface condition and the question is whether it wets. A sample check decides it: a stain that wets is cosmetic, while a pad that is dull under the stain is a bath problem that repeats on the next lot.
Next step
Send the Gerber set, the BOM and the build schedule you are planning, including any lot that will be held as a spare. The reply comes back with a finish recommendation matched to that schedule, the packaging and bake plan, and the reports that travel with each lot when it ships.