EDTA Titration of Limestone: A Persistent High Bias
A QC laboratory determines CaCO3 in limestone by EDTA titration. A certified reference limestone is 95.1% CaCO3. A new analyst obtains 97.8%, 97.5%, and 98.0% on three replicates, with good repeatability. The procedure is: 0.5000 g sample dissolved in HCl, diluted to 250.0 mL; 25.00 mL aliquot; pH adjusted to 12.5 with NaOH; murexide indicator; titration with 0.0100 M EDTA. The EDTA factor is 1.004 and was applied correctly. A 0.15 mL blank was not subtracted. The burette correction is -0.05 mL at 25 mL. The limestone contains 1.2% MgCO3. The analyst stops at the first permanent purple, although the endpoint fades after about 20 s. The lab must decide whether to release the result, restandardize, or investigate.
1. Rank the plausible causes of the +2.7 percentage-point bias and explain quantitatively which are large enough to matter.
Read the worked answer
The largest plausible cause is magnesium. If 1.2% MgCO3 is titrated as though it were CaCO3, the equivalent bias is 1.2% × (100.09/84.31) ≈ 1.42 percentage points of apparent CaCO3. That is large enough to explain half the bias. The next largest is the endpoint: a fading murexide endpoint encourages over-titration, and 0.5–1.0 percentage point is plausible. The unsubtracted 0.15 mL blank contributes roughly (0.15/48.7) × 97.8 ≈ 0.30 percentage point, using a titre near 48.7 mL. The burette correction of -0.05 mL would lower the result by only about 0.10 percentage point, so it cannot explain a high bias. The EDTA factor was applied correctly and is not a cause.
Magnesium should be titrated only if the pH is below about 12.5 or the pH electrode is mis-calibrated; at sufficiently high pH, Mg(OH)2 precipitates and does not consume EDTA. The combination of Mg interference, endpoint over-titration, and blank can easily reach 2.7 percentage points.
2. Design a short experiment to establish whether Mg is being titrated and whether the pH adjustment is adequate.
Read the worked answer
Prepare synthetic solutions containing Ca and Mg in the same ratio as the limestone, plus a Ca-only control. Adjust aliquots to pH 12.0, 12.5, and 13.0 with NaOH, measure the actual pH with a calibrated electrode before titration, and titrate under the same conditions. If Mg is being titrated, the apparent CaCO3 will fall as the pH is raised and Mg(OH)2 precipitation becomes complete. Centrifuge or filter a portion at each pH to confirm precipitate formation. Run the certified reference limestone with and without added Mg, and analyze the same solutions by ICP-OES as an independent reference.
Also test the indicator: murexide binds Mg and gives a fading endpoint, so compare with a sharper Ca indicator or a back-titration. If the CRM result moves to 95.1% at pH 13.0 but is high at pH 12.0, the pH adjustment, not the EDTA standardization, is the root cause. Record the endpoint criterion and require a stable endpoint for at least 30 s.
3. What should be reported for the production batch, and what corrective actions are needed?
Read the worked answer
Do not release 97.8% CaCO3. The CRM bias is about +2.7 percentage points, and the three replicate results are precise but inaccurate, which is exactly the failure mode a CRM is meant to catch. Invalidate the analyst's run, subtract the blank, verify the pH electrode and the actual pH of the aliquot, and repeat the CRM and sample with the corrected procedure. If the corrected CRM falls within the laboratory's acceptance limit around 95.1%, the sample can be reported with the corrected value and an uncertainty statement. If not, investigate the EDTA standardization and the Mg separation further.
Corrective actions include updating the SOP to require a blank, pH verification with a calibrated meter, a defined endpoint stability criterion, and a Mg check for this limestone source. Retrain the analyst on endpoint recognition and require CRM bracketing for each batch. Any previously reported results from the same analyst or same procedure should be reviewed for the same high bias.