I still remember the audit that changed how I buy vials. A state inspector pulled our method file for EPA 8260 and asked one question: “Show me the certificate that ties this lot of 40 mL vials to the volatile organic analysis you ran last Tuesday.” We had the data. We did not have the paper trail, because the vials came from a distributor who repackaged them in unmarked cases. We passed, barely, after three days of scrambling. I have specified certified vials for every regulated method since, and I have watched more labs get tripped up by the same gap.
COD vials sit at an awkward intersection. They are consumables, so procurement treats them as commodity items. They also carry analytical results, so regulators treat them as part of the measurement system. If your lab runs chemical oxygen demand by any of the standard methods, the vial is not a container. It is a reaction vessel, a spectrophotometric cell, and in some methods, a sealed pressure device. Get the specification wrong and you do not get a warning. You get a low bias, a failed QC check, or a broken vial in a block digester at 150 °C.
Three method families dominate COD work, and each treats the vial differently. Understanding the differences before you buy saves you from the most common mistake: assuming a vial qualified for one method is qualified for all of them.
EPA 410.4 and the closed reflux colorimetric approach
EPA method 410.4 is the standard colorimetric COD method for wastewater. The sample is digested in a sealed tube with potassium dichromate and sulfuric acid at 150 °C for 2 hours, then read at 600 nm (or 420 nm for low-range work). The method specifies a culture tube or ampule that can withstand the digestion pressure, and it assumes a defined optical path length. Most commercial COD vials are built to a 16 mm outer diameter with a 10 mm light path, which matches the standard spectrophotometer cell holders on Hach, Thermo, and similar instruments.
The method does not name a specific vial part number. It names performance. That is the loophole that gets labs in trouble. A vial can be 16 mm and still fail if the wall thickness varies enough to change the path length, or if the glass transmits differently at 600 nm from lot to lot. I have seen absorbance drift of 0.008 AU between vial lots, which sounds small until you are reading a low-range sample at 0.030 AU.
ISO 6060 and the titrimetric reference
ISO 6060 is the international reference method for COD, and it is titrimetric rather than colorimetric. The digestion happens in a reflux apparatus or a sealed tube, and the residual dichromate is titrated with ferrous ammonium sulfate. The vial requirement here is less about optics and more about chemical inertness and seal integrity. Borosilicate glass is the default. Soda-lime glass will leach sodium into the digestate at 150 °C and skew the titration, especially at low COD values where the dichromate excess is large.
ISO 6060 does not mandate a specific vial, but it does require that the digestion vessel contribute no measurable COD. That is a blank requirement, and it is where cheap vials fail. I have run blank digests on imported vials that read 8 mg/L COD equivalent because of residual organic contamination on the glass or in the cap liner. On a sample reading 30 mg/L, that is a 27% error.
AOAC and the food and agricultural angle
AOAC methods for COD are less common but show up in food processing wastewater and agricultural runoff work. The vial requirements track EPA 410.4 closely, but AOAC method validation adds a wrinkle: you often need to demonstrate equivalence between your vial-based method and the reference titrimetric procedure. That means running both on the same samples, which means your vial blank and your digestion recovery have to be documented, not assumed.
If you are validating under AOAC guidelines, keep the certificate of analysis for every vial lot you use during the study. Auditors will ask for it. I keep a binder with lot certificates going back five years, and it has saved me twice.
Glass, caps, and the failure modes I have actually seen
Most COD vial failures are not dramatic. They are slow drifts and quiet biases that show up as QC failures weeks after the vial lot changed. Here is what I watch for.
Glass composition and thermal history
Borosilicate 3.3 is the right glass for COD work. It has a low coefficient of thermal expansion (about 3.3 × 10⁻⁶ /K), which matters when you go from a 150 °C block digester to a room-temperature cooling rack in under five minutes. Soda-lime glass at 9 × 10⁻⁶ /K will survive a few cycles, then crack. I have seen it happen on the third or fourth digestion, which is worse than failing on the first because you have already reported the earlier results.
Wall thickness matters too. A vial with a nominal 10 mm path but a ±0.15 mm wall tolerance will give you a path length that varies by 3% across the lot. On a mid-range sample that is a 3% bias you will never see unless you run a path-length check. I ask suppliers for path length verification on a sample of at least 10 vials per lot, and I reject lots that exceed ±0.05 mm.
Cap liners and the seal problem
The cap is where most leaks start. COD vials use a screw cap with a liner, usually PTFE-faced silicone or a bonded PTFE septum. The liner has to do two things: seal against 1.5 to 2 bar of internal pressure at 150 °C, and contribute no organic carbon to the digestate.
Silicone-backed liners are common because they seal well, but the silicone face can leach at temperature if it contacts the liquid. The PTFE face is the barrier. I have seen liners where the PTFE was so thin (0.05 mm) that it developed pinholes during autoclave-style digestion, letting the sample touch the silicone. The result was a positive bias of 15 to 40 mg/L COD on affected vials. You will not catch that with a visual inspection.
Torque matters here. Hand-tight is not a specification. I specify 1.5 to 2.0 N·m for the cap torque on 16 mm COD vials, checked with a torque screwdriver on a sample of five caps per case. Under-torque and you leak at pressure. Over-torque and you crack the cap or deform the liner. Both failure modes are silent until they are not.
Volume and headspace
COD vials are typically 16 × 100 mm with a working volume of 10 mL (2 mL sample plus reagents) or 5 mL for micro-scale work. The headspace is not incidental. It is the pressure buffer during digestion. A vial filled too full will build pressure faster and can vent or fail. A vial with too much headspace will lose volatile organics to the gas phase, which biases low.
If you are running a method that specifies a fill volume, respect it. If you are adapting a method, validate the headspace effect before you commit. I have a customer who switched from 10 mL to 5 mL vials to save reagent and saw a 12% low bias on samples with high volatility, because the headspace-to-liquid ratio changed. They went back to 10 mL vials within a month.
Certification, traceability, and what the paperwork should say
This is the part procurement hates and auditors love. A COD vial used in a regulated method needs documentation that ties the physical object to the analytical result. Here is what I require from any supplier.
- Lot-specific certificate of analysis with the glass composition, path length verification, and blank COD value.
- Blank COD result below 2 mg/L for the lot, tested by the digestion method you will use.
- Path length tolerance stated numerically, not as “nominal 10 mm.” I want to see the measured range.
- Cap liner material specification with the PTFE thickness and the silicone durometer.
- Pressure test data showing the vial and cap assembly held at 2 bar for the digestion duration without leakage.
- Traceability statement linking the lot to the glass batch and the liner batch.
If a supplier cannot produce these, they are selling you a vial, not a qualified consumable. That is fine for a teaching lab. It is not fine for a method that reports to a permit.
This is also where the broader consumables supply chain matters. A vial is only as good as the septum, the cap, and the insert that go with it. If you are running COD alongside other wet chemistry, it helps to standardize on suppliers who can document the whole assembly. The COD vial specification should include the cap and liner as a unit, not as three parts sourced separately.
Method-specific selection criteria with numbers
Here is the decision table I use when specifying COD vials for a new method or a new lab. The numbers are the ones I have validated in practice, not marketing claims.
Comparison: vial specification by method
- EPA 410.4 (colorimetric, standard range): 16 mm OD, 10 mm path, borosilicate 3.3, 10 mL working volume, PTFE-faced liner ≥0.1 mm, blank COD <2 mg/L, path tolerance ±0.05 mm. Digestion at 150 °C for 2 h.
- EPA 410.4 (low range): same physical spec, but path length tolerance tightens to ±0.03 mm and blank COD must be <1 mg/L. Low-range work at 420 nm is more sensitive to glass transmission variation.
- ISO 6060 (titrimetric): 16 mm OD or larger, borosilicate 3.3, 10 to 20 mL working volume, PTFE or glass stopper, no liner leaching. Blank COD <2 mg/L. Reflux or sealed tube at 150 °C.
- AOAC validation work: same as EPA 410.4, plus lot certificates retained for the full validation period and a documented equivalence study against the titrimetric reference.
Decision checklist before you buy
- Does the vial OD match your spectrophotometer cell holder? 16 mm is standard, but confirm.
- Is the path length verified per lot to ±0.05 mm or better?
- Is the glass borosilicate 3.3, and is that stated on the certificate?
- Is the cap liner PTFE-faced at ≥0.1 mm, with the silicone backing specified?
- Has the assembly been pressure tested at 2 bar for the digestion duration?
- Is the blank COD below 2 mg/L by your digestion method?
- Can the supplier provide lot traceability to the glass and liner batches?
- Does the cap torque specification fall in the 1.5 to 2.0 N·m range, and do you have a torque tool to verify it?
If any of those answers is “we think so,” get it in writing before you place the order.
Operational precautions in the digestion block
Even a well-specified vial can fail if the handling is wrong. These are the operational rules I enforce in my own lab and in customer labs I audit.
Loading and capping
Cap the vial immediately after adding reagents. The dichromate-sulfuric acid mixture is aggressive, and any splash on the cap threads will corrode them and compromise the seal on the next use. I use a dedicated capping station with a torque screwdriver set to 1.8 N·m. It takes an extra four seconds per vial and eliminates the leak failures we used to see at a rate of about 1 in 200.
Do not reuse caps. The liner deforms under torque and heat, and a reused cap will not seal to the same pressure. I know labs that reuse them to save money. They also report more QC failures. The math does not favor reuse.
Digestion block temperature and ramp
Set the block to 150 °C and let it equilibrate for at least 30 minutes before loading. A cold block will ramp the vials unevenly, and the vials at the edges will see a different thermal history from the center ones. I have measured a 4 °C spread across a 24-position block during ramp, which is enough to change the digestion kinetics on low-range samples.
Load vials in a consistent pattern and do not mix vial types in the same run. A borosilicate vial and a soda-lime vial in the same block will expand differently and one of them will crack. I have seen it.
Cooling and reading
Cool the vials to room temperature before reading. The absorbance of the dichromate solution is temperature-dependent, roughly 0.001 AU per °C at 600 nm. Reading hot vials gives you a bias that varies with your cooling time. I use a forced-air cooling rack and read after 20 minutes, which is long enough to equilibrate and short enough to keep throughput up.
Wipe the outside of the vial before reading. A fingerprint on the optical surface will scatter light and give you a positive bias. I keep a lint-free wipe and a bottle of methanol at the reader. It sounds trivial. It is not.
Troubleshooting checklist
When COD results drift or QC fails, work through this list before you blame the method. In my experience, the vial is the cause about a third of the time, and the rest is split between reagent age, digestion temperature, and reader calibration.
- Blank COD high (>5 mg/L): check vial lot certificate, check reagent age (dichromate degrades), check for liner leaching. Run a blank with a known-clean vial to isolate.
- Low bias across the board: check digestion temperature (a block running 5 °C low will under-digest), check vial fill volume and headspace, check cap seal for slow leaks.
- High bias on volatile samples: check headspace ratio, check cap seal integrity, check that vials are capped immediately after reagent addition.
- Absorbance drift between vial lots: check path length verification data, check glass transmission at your reading wavelength, run a path-length check on 10 vials from the new lot.
- Cracked vials in the block: check glass composition (soda-lime vs borosilicate), check ramp rate, check for mixing of vial types in one run.
- QC standard out of range: check standard preparation, check reagent lot, check reader calibration, then check vial lot. In that order.
If you are running multiple methods, keep a log of vial lot numbers against QC results. I have caught two bad lots in five years that way, and both suppliers replaced them without argument once I showed the data.
Where the vial fits in the broader consumables picture
COD work rarely happens in isolation. Most labs running COD also run spectrophotometric methods, and many run HPLC or GC alongside. The vial specification discipline carries over. A vial insert with a poorly specified glass or a leachable liner will give you the same kind of quiet bias in an HPLC run that a bad COD vial gives you in a digest. The 2 mL vial you use for autosampler work has its own path length and inertness requirements, and the same rule applies: get the certificate, verify the lot, keep the paperwork.
For labs standardizing across a portfolio, it helps to buy from suppliers who document the whole chain. The HPLC consumables market has matured on this point, and the COD vial market is catching up. If you want to see how the major instrument vendors handle consumable qualification, the Agilent chromatography consumables documentation is a useful benchmark, and Thermo Fisher’s chromatography and sample handling pages show the level of specification detail you should expect from a serious supplier.
For method-specific guidance, the primary sources are still the best. The U.S. EPA SW-846 test methods page has the current method text and updates for 410.4 and related procedures. Restek’s technical resources are useful for understanding how consumable choices affect analytical results, even though their focus is chromatography. Read the method, then read the certificate, then compare.
Recommendation
For any lab running COD under EPA 410.4, ISO 6060, or an AOAC-validated method, specify borosilicate 3.3 vials at 16 mm OD with a verified 10 mm path length (±0.05 mm), a PTFE-faced liner of at least 0.1 mm, and a lot certificate showing blank COD below 2 mg/L. Cap at 1.5 to 2.0 N·m with a torque tool, do not reuse caps, and keep lot certificates for at least five years. If you are doing low-range work at 420 nm, tighten the path tolerance to ±0.03 mm and the blank requirement to <1 mg/L. Pay for the certificate. It costs a few cents per vial and it is the difference between a defensible result and a three-day scramble when the inspector asks for the paperwork.
The vial is not the glamorous part of the method. It is the part that determines whether the method holds up when someone else looks at it. Specify it like it matters, because in a regulated lab, it does.
References & further reading
Related resources from our lab consumables network:
- Cod Vial at codvial.com
- Vial Insert at hplcinsert.com
- 2Ml Vial at 2mlsamplevial.com
- Hplc Consumables at hplcconsumable.com
External references and standards: