Older Date Codes: How to Qualify Solderability on Aged IC Stock

The question behind every old date code #

Two offers arrive for the same part number. One carries a 2025 date code at a firm price; the other is a 2019 lot at two thirds of the money. The usual reflex is to screen on the year and stop there, which quietly does two bad things at once: it rejects stock that would have soldered without complaint, and it waves through lots whose real problem was never the calendar.

Age is a proxy, and a weak one. What decides whether a termination wets in the oven is the finish on that termination, the packaging that shielded it, the air it sat in, and whether anyone has been at the leads since the parts left the assembly site. Those four things are all verifiable. The date code is simply the trigger to go and verify them.

What component makers actually say about component age #

The three published positions below look contradictory at first glance and are not.

Microchip's document is worth reading closely because it separates the two ideas most RFQs confuse. It attributes the disappearance of age-related solderability failures to qualification, handling and packaging per J-STD-020, JEP160, JESD22-A113 and J-STD-033, and it still keeps a formal aging policy: four years for microcontrollers and other integrated circuits, with a recertification route (inspection, retest, rescan, repacking) as parts approach 24 months that permits another 24 months of shipment and ships with a Certificate of Conformance. Die in wafer or diced form carry no aging restriction under JEP160 storage guidance. Power management products, modules and discretes are recertified as they approach 24 months, and quartz frequency control products at 12 months, which is a reminder that "IC" is not one policy.

Infineon comes at it from the storage side. Its guideline ties any recommendation to storage time *and* storage conditions, and Infineon's own support has confirmed a three-year figure for an individual diode part number while declining to extend it to five years. The widely repeated "three to five years for SMD" is therefore a rule of thumb, not a specification you can quote back to a customer.

TI's study explains why the metallurgy usually outlasts the paperwork. After exposure to a Battelle Class 2 mixed-flowing-gas environment, described as an indoor environment without humidity control of the kind found in a warehouse, parts still passed surface-mount solderability testing. Units exposed loose in petri dishes developed thin corrosion films that were visible and confirmed by wetting-balance data, while the same devices left in tubes and in tape and reel showed no visible corrosion at all.

LimChip's reading of the three together: four years and three years are commercial and documentation boundaries a manufacturer will stand behind; eight-plus years is what the finish did under test *in its original packing*. Age triggers evidence. Packaging condition is what actually changes the outcome.

What genuinely degrades in storage #

Oxide on the termination. Tin and tin-lead finishes grow oxide and intermetallic layers over time. Flux has to remove that oxide during reflow, and past a certain thickness it cannot, which shows up as non-wetting or as solder that wets and then pulls back into islands.

Environment and packaging integrity. This is the variable that separates two lots of identical age. Original tubes, trays and reels attenuate corrosive gases and humidity, as the TI results show; loose parts, hand-repacked trays and split reels do not. Sulphur-bearing air, uncontrolled humidity and repeated temperature swings all accelerate the same mechanism.

Moisture, which is a separate axis entirely. A part can be perfectly solderable and still crack in the oven because the plastic body absorbed water. That is governed by MSL classification and floor life rather than by finish oxidation, and it is covered separately in the moisture sensitivity and baking guide. Do not let a passing solderability sample stand in for dry-pack control.

Tin whiskers on bare Sn finishes. Whiskers can bridge fine-pitch leads. JEDEC handles this with JESD201 for acceptance requirements and JESD22-A121 as the measurement method, and mitigation is a plating-process question rather than a storage question. TI, for example, publishes that it anneals matte-tin leadframe packages for one hour at 150 °C within 24 hours of plating, holds a minimum as-plated thickness of 7 µm, and applies at least 5 µm of SnAgCu after trim and form on hot-solder-dipped devices, referencing the mitigation practices in JEDEC/IPC JP002. For an aged lot, the practical question is which finish is on it and whether the supplier can state the mitigation, not whether whiskers grew in the warehouse.

Hands that have already been on the leads. Re-tinning, lead straightening and reballing can make an old part look better than it is, and they change the finish stack you thought you were buying. The visual signatures are covered in the counterfeit screening guide; the sourcing consequence is that a re-tinned lot is a reworked lot and should be graded as such.

Testing it properly: J-STD-002E #

The relevant standard is IPC/JEDEC/ECIA J-STD-002E, *Solderability Tests for Component Leads, Terminations, Lugs, Terminals and Wires*, published in November 2017. It answers one narrow question well: can these terminations be wetted now. It says nothing about how long the finished joint will last, and it is destructive, so the samples you test are gone.

Method selection follows the package, not the buyer's preference:

Preconditioning is where an aged-stock test is either meaningful or theatre. The standard defines coating durability categories that force oxidation before the dip: Category 1 requires no steam aging and suits parts to be soldered soon after test, Category 2 requires one hour ±5 minutes, and Category 3 requires eight hours ±15 minutes in a chamber at roughly 93 °C and near-saturated humidity. Category 3 is the default expectation for tin and tin-lead finishes that have been, or will be, stored a long time. A supplier report that passed with no steam aging on a 2019 lot has proved very little.

Acceptance is a coverage judgement: at least 95 % of the critical area covered by a continuous, smooth solder film, with non-wetting and dewetting bands failing the sample. Because it is destructive and sampling-based, agree the sample size, the aging category and who pays for the consumed units *before* the units ship.

What to ask for before you accept the lot #

• An explicit date-code range, not "recent" or "2xxx+". Range is what an engineering team can accept or reject against a written rule.
• Packaging state described in physical terms: factory-sealed moisture barrier bag with desiccant and an intact humidity indicator card, or original tube, tray or reel, or repacked. Repacked is not automatically bad, but it must be declared.
• Lot structure. Note that a single factory reel legitimately mixes date codes. Microchip's published packing rules for tube, tray and tape and reel allow one device with up to three date codes, one assembly site, one test site, one wire material and one die revision. So ask for the codes present rather than rejecting a two-code reel on reflex.
• Any recertification record, if the manufacturer's own age policy has been triggered. This is the cheapest evidence available and it usually comes with a Certificate of Conformance.
• A solderability report that states method, aging category, coverage result, test date and who performed it. A report without the aging category is not comparable to one with it.

How much of that you insist on should track the consequence of being wrong:

Where the line should be #

An older date code is a reason to ask for evidence, not a verdict. Aged stock is acceptable when the packaging is intact and traceable, when the termination finish is declared, and when a correctly preconditioned J-STD-002 sample confirms wetting. It should be refused when the storage history cannot be described, when the terminations show rework, or when the application cannot absorb a residual risk that nobody has measured. The manufacturers themselves draw the line in exactly this way: they set an age at which they re-examine and re-document a lot, not an age at which the silicon or the plating stops working.

For sourcing, the practical change is small and pays for itself. Put an acceptable date-code range and a packaging condition in the RFQ text, and ask what evidence travels with the lot. Buyers working from an approved BOM can check what is already published in the part catalogue and state the date-code requirement in the enquiry, which removes most of this argument before a quote is issued.

Sources #

• Microchip Technology, Component Age Policy and Integrated Circuits Packing Media Methods, 1 August 2024.
• Infineon Technologies, Storage of Products Supplied by Infineon Technologies, guideline V10.0, 10 May 2023.
• Texas Instruments, Shelf-Life Evaluation of Lead-Free Component Finishes (SZZA046).
• Texas Instruments, Lead finish / ball material and tin plating process, including JESD201 and JESD22-A121 whisker test practice.
• IPC/JEDEC/ECIA J-STD-002E, *Solderability Tests for Component Leads, Terminations, Lugs, Terminals and Wires*, November 2017 (test methods, coating durability categories and acceptance criteria).