Moisture, trace-metal, and anion results in lithium-ion battery raw-material quality control cannot be interpreted as universal impurity numbers. Their meaning depends on the chemical being tested, the physical or chemical form of the analyte, sample preparation, analytical method, reporting basis, and downstream process. Water can promote LiPF₆ degradation in carbonate electrolytes, while the same water concentration does not imply the same pathway in LiFSI systems. Iron may be a metallic contaminant in an NMC powder but an intended lattice element in LiFePO₄. Sulfate may be a process residue, whereas PF₆⁻ is an intentional electrolyte anion. Valid comparison therefore requires a matrix-specific measurand and a fit-for-purpose method.

A QC Number Is Meaningful Only When the Measurand Is Defined

A reported value such as “water: 30 mg/kg,” “Fe: 5 mg/kg,” or “sulfate: 20 mg/kg” is incomplete unless the result defines what was actually measured.

For battery raw materials, a defensible measurand normally includes:

• the material identity, grade, physical form, and sampling condition;
• the analyte’s relevant chemical or physical form;
• the extraction, digestion, dilution, or thermal-release procedure;
• the analytical technique, calibration range, recovery, and reporting limit;
• the reporting basis, such as mass of original solid, prepared solution, or dry material.

The 2025 edition of Eurachem’s Fitness for Purpose of Analytical Methods treats sampling, sample handling, validation, selectivity, working range, detection capability, precision, and recovery as parts of method fitness. This is directly relevant to battery QC: two laboratories can produce different results from the same material without either result being analytically invalid if they define or prepare the measurand differently.

The correct QC question is therefore:

What material-specific state does this result measure, and what conclusion can that measurement support?

Why Water Results Are Salt- and Sample-Form Specific

Water in LiPF₆ Carbonate Electrolytes Has a Specific Reaction Path

In carbonate electrolytes containing lithium hexafluorophosphate, water can participate in LiPF₆ degradation and the formation of HF and fluorophosphate species. The simplified relationship is:

Water exposure + LiPF₆-containing electrolyte → salt-degradation products → altered electrode-interface reactions

The pathway depends on solvent composition, water concentration, temperature, storage time, electrode chemistry, and electrochemical potential. A water value alone therefore cannot predict the magnitude of cell degradation.

A study of silicon anodes compared a Gen2 electrolyte containing 1.2 M LiPF₆ in ethylene carbonate/ethyl methyl carbonate at a 3:7 mass ratio, initially below 10 ppm water, with the same electrolyte after adding 50 ppm water. Measurements were performed at potentials from 1.0 to 0.01 V versus Li/Li⁺. The additional water changed the morphology and composition of the silicon solid-electrolyte interphase, promoted parasitic reactions, and produced a more heterogeneous fluorophosphate-rich interphase. The authors also associated LiPF₆ hydrolysis products with attack on the silicon surface oxide. These findings are reported in the original silicon-anode study.

The valid conclusion is narrow: water changed interphase formation in that LiPF₆/carbonate/silicon system under the tested conditions. The study does not establish 50 ppm as a universal failure threshold for graphite, lithium metal, different solvents, different additives, or commercial full cells.

LiFSI Does Not Inherit the LiPF₆ Water-Reaction Model

Lithium bis(fluorosulfonyl)imide has different anion chemistry from LiPF₆. An original physicochemical study reported substantially greater LiFSI stability toward hydrolysis than LiPF₆ in the evaluated systems (Journal of Power Sources, DOI 10.1016/j.jpowsour.2010.12.040).

This difference has two important boundaries:

1. A LiPF₆-derived HF formation model cannot be transferred unchanged to an LiFSI electrolyte.
2. Greater salt stability toward hydrolysis does not demonstrate that water is harmless in an LiFSI cell.

Water may still participate in solvent reactions, electrode-surface reactions, or interphase formation. The tolerable concentration remains dependent on electrode chemistry, electrolyte composition, potential window, temperature, and cell design. Consequently, a water specification developed for LiPF₆ cannot be adopted for LiFSI solely because both materials are lithium electrolyte salts.

The Water Method Must Match the Sample Form

ASTM E1064-24 covers coulometric Karl Fischer determination of water from 0% to 2.0% by mass in most liquid organic chemicals. It is particularly suited to low water concentrations in organic liquids. Its stated scope does not automatically cover electrode powders, cathode precursors, solid lithium salts, or polymeric binders.

Direct coulometric injection can be appropriate for a compatible liquid electrolyte or solvent. Solid powders commonly require an oven-transfer procedure in which evolved water is transported into the titration cell. A Metrohm application bulletin on water in lithium-ion battery materials describes this indirect approach for solid battery materials.

Oven temperature, hold time, sample mass, carrier-gas flow, blank correction, and vial handling then become part of the measurand:

• insufficient release conditions can leave some water unmeasured;
• excessive heating can cause sample reactions or generate volatile products that interfere with the determination;
• moisture uptake during sampling can raise the result before analysis;
• different thermal programs can produce values that are numerically similar but operationally different.

A powder result should therefore be reported with its release program and validation evidence. The phrase “measured by Karl Fischer” is insufficient for cross-lot or cross-supplier comparison.

Why Total Metal Concentration and Metal-Particle Risk Are Different Measurands

Trace-metal interpretation begins with a distinction between elemental identity and material role.

Iron is a principal lattice element in LiFePO₄. Nickel, manganese, and cobalt are principal elements in NMC materials. Reporting these elements as generic “metal impurities” would be chemically meaningless. In another matrix, the same element could be an unintended dissolved ion, a metallic fragment introduced by equipment wear, a surface deposit, a secondary phase, or an intentional dopant.

These forms can differ in mobility, electronic conductivity, surface reactivity, dissolution behavior, and spatial distribution. They also require different evidence.

An original study introduced elemental Al⁰, Cu⁰, Fe⁰, Mg⁰, and Si⁰ contaminants into lithium-ion electrode materials and found that their electrochemical and thermal effects depended on both the contaminant and the electrode in which it was present. The study investigated metallic contaminants associated with recycling streams; it did not test every ionic, oxide, or lattice-bound form of those elements. The open research record and journal DOI therefore support an electrode- and form-dependent conclusion, rather than a universal ranking of metal limits.

Sample Preparation Determines What ICP Measures

ICP-MS or ICP-OES measures elements introduced into the plasma after the sample has been converted into a suitable solution. The result depends on whether preparation achieved complete digestion, partial acid leaching, or extraction of a soluble fraction.

The US EPA SW-846 Compendium separates inorganic sample-preparation procedures from ICP determinative methods and distinguishes total, total-recoverable, dissolved, and matrix-specific digestion approaches. Although these methods were developed for environmental and waste matrices rather than battery-material release testing, they demonstrate a general analytical boundary: an ICP result cannot represent material that the preparation step did not bring into solution.

For an electrode powder, complete digestion can provide a bulk elemental inventory. It simultaneously destroys information about whether an element was present as a discrete metallic particle, surface deposit, oxide, or lattice constituent.

Localization Requires a Different Measurement

SEM-EDS can examine the morphology and elemental composition of selected regions or particles. NIST’s technical discussion of SEM X-ray microanalysis explains that spatial resolution depends on beam energy, specimen composition, and the electron–sample interaction volume.

SEM-EDS can therefore support the identification of an Fe-rich or Cu-rich particle in an electrode powder. It cannot, from a few selected fields, establish a representative bulk concentration across an entire production lot. That conclusion requires a sampling plan and a bulk quantitative method.

A strong metal-contamination investigation often needs both bulk and localized evidence. ChemicalCell’s analysis of where metal contamination can enter high-purity chemical systems provides the adjacent source-investigation path; the current question remains how the resulting analytical data should be interpreted.

Anion Identity Must Be Classified Before It Is Called an Impurity

“Anion content” does not describe one universal quality parameter.

PF₆⁻ in LiPF₆, FSI⁻ in LiFSI, and TFSI⁻ in LiTFSI are intentional components of electrolyte salts. Chloride or sulfate in a lithium hydroxide feedstock may instead represent residual process ions. Sulfate associated with an NMC precursor or resynthesized cathode can occupy yet another role, depending on whether it remains extractable, forms a surface species, or is present in a secondary phase.

The first distinction is therefore:

Intentional salt anion ≠ extractable residual anion ≠ total elemental content ≠ solid-phase anion-containing species

Ion chromatography with conductivity detection can quantify separated ionic species in a prepared solution. It does not automatically measure non-extracted material, total sulfur, total phosphorus, or the location of an anion-containing phase.

Lithium Hydroxide Demonstrates Why Matrix Treatment Matters

A Thermo Fisher application note for chloride and sulfate in saturated lithium hydroxide used a tenfold-diluted saturated LiOH solution, an OnGuard II H cartridge, an AS29-Fast-4µm column, suppressed conductivity detection, and a 5.0 µL injection.

Under those stated conditions, the reported method detection limits for the saturated solution were:

• 0.09 mg/L for chloride;
• 0.13 mg/L for sulfate.

Spike recoveries were 91–102%, with reported relative standard deviations of 1–6%.

These values describe that method, matrix preparation, calibration, and solution basis. They are not universal specification limits for solid battery-grade LiOH. Converting the result to a solid-material basis requires traceable sample mass, dissolution or saturation conditions, dilution factors, blank correction, and recovery appropriate to the actual material.

Sulfate Effects Depend on the Cathode Process and Chemical State

A study of resynthesized Li[Ni₁/₃Co₁/₃Mn₁/₃]O₂ compared material produced with and without excessive sulfate in recycled-battery leachate. XRD and XPS identified sulfate-related species, including an Li₂SO₄ impurity phase in the sulfate-containing material. The material also showed a more porous morphology and slightly lower discharge capacity across the tested 0.1–5 C rates. The conditions and results are reported in the original NCM111 study.

This evidence supports a specific chain:

Excess sulfate in the investigated leachate → retained sulfate-related species and changed particle growth → altered morphology and electrochemical behavior in resynthesized NCM111

It does not establish that every detectable sulfate concentration is harmful, that the same response occurs in fresh commercial precursors, or that an IC result alone predicts cathode performance. Establishing that connection requires evidence linking the extractable sulfate result to the relevant surface or solid phase and then to the intended cathode process.

How This Framework Applies to Lot Consistency and Supplier Change

Lot consistency and supplier change should be treated as comparability questions, rather than separate proof of quality.

A difference between two results can be attributed to the material only after controlling the major measurement variables:

Same material definition + comparable sampling + matched preparation + matched method + matched reporting basis

If one of these elements changes, the observed difference may arise from the material, the measurement system, or both. Conversely, two identical COA values do not prove material equivalence when impurity form, particle distribution, sampling exposure, or method recovery has changed.

A supplier change therefore creates a need to re-establish analytical comparability; it does not by itself prove deterioration. Detailed approval decisions belong in the separate ChemicalCell references on battery electrolyte raw-material qualification and battery electrode-material qualification.

What Can Battery Raw-Material QC Validly Conclude?

A well-defined moisture result can establish water content in the tested sample under the specified sampling and KF procedure. It cannot independently establish electrolyte stability, future package exposure, or cell life.

A digestion–ICP result can establish bulk elemental concentration for the material fraction brought into solution. It cannot independently identify a metallic particle, surface deposit, oxidation state, or lattice position.

An IC result can establish the concentration of selected extractable ionic species under the validated preparation and chromatographic conditions. It cannot independently establish total elemental content, solid-phase distribution, or electrochemical effect.

Lot data can demonstrate statistical behavior only when sampling and measurement remain comparable. A supplier-change bridge can show equivalence only for the attributes and application conditions actually evaluated.

The defensible conclusion is therefore conditional:

Battery raw-material quality results become comparable and technically useful when the material matrix, analyte form, sample preparation, analytical method, reporting basis, and downstream relevance are defined together.

That framework explains why a universal moisture, metal, or anion threshold cannot be derived from a grade name, one supplier specification, or one published experiment. It also defines the evidence required before a QC number can support lot comparison, source-change assessment, or application-specific qualification.