Alkalinity determination by potentiometry: intercalibration using three different methods

Seawater was sampled from different depths in the North Atlantic Ocean (Canary Islands region) and distributed among three different labs for the determination of titration alkalinity. Analysis was performed by potentiometric methods, involving titration in a closed cell, titration in an open cell a...

Full description

Bibliographic Details
Published in:Ciencias Marinas
Main Authors: L Mintrop, FF Pérez, M González-Dávila, JM Santana-Casiano, A Körtzinger
Format: Article in Journal/Newspaper
Language:English
Spanish
Published: Universidad Autónoma de Baja California 2000
Subjects:
Online Access:https://doi.org/10.7773/cm.v26i1.573
https://doaj.org/article/37fc5c78c9df42f782916c58525704a0
Description
Summary:Seawater was sampled from different depths in the North Atlantic Ocean (Canary Islands region) and distributed among three different labs for the determination of titration alkalinity. Analysis was performed by potentiometric methods, involving titration in a closed cell, titration in an open cell and a two end-point acid addition method. The precision, which is the sample reproducibility taken from the mean standard deviation for replicate measurements, was between 0.45 and 0.90 µmol · kg–1 for the individual labs. Accuracy, here taken as the deviation for the values of a lab from the mean of all three, was mostly below 1 µmol · kg–1 and never exceeded 0.1% of the sample value. Mean standard deviation for all labs and all samples was 0.87 µmol · kg–1, once the individual methods were calibrated using certified reference material (CRM). Without CRM calibration, the mean standard deviation would increase to 2.8 µmol · kg–1. The conclusion is that current high precision methods for alkalinity measurements calibrated with CRMs are able to reach similar accuracy as the measurement of total dissolved inorganic carbon by coulometry and therefore allow for the precise determination of the oceanic carbon dioxide system by using the two measured parameters.