Premium
Conductimetric Measurement of CO 2 Concentration: Theoretical Basis and Its Verification
Author(s) -
Baker John M.,
Spaans Egbert J. A.,
Reece Clive F.
Publication year - 1996
Publication title -
agronomy journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.752
H-Index - 131
eISSN - 1435-0645
pISSN - 0002-1962
DOI - 10.2134/agronj1996.00021962008800040029x
Subject(s) - reflectometry , conductivity , calibration , chemistry , electrical resistivity and conductivity , analytical chemistry (journal) , thermodynamics , time domain , environmental chemistry , statistics , physics , mathematics , computer science , electrical engineering , computer vision , engineering
Atmospheric CO 2 is important to plant growth and also plays a key role in the global energy balance. Thus, there are needs for reliable methods for measuring CO 2 concentrations, [CO 2 ]. One approach has been the conductimetric method, where sampled gas is bubbled through deionized water. Some of the CO 2 in the air dissolves and ionizes, causing an increase in solution electrical conductivity. The method is inexpensive relative to other techniques, but usage has been limited, possibly due to its apparently empirical basis and suggestions that it must be frequently checked to correct for shifts due to temperature and other effects. We have derived the fundamental basis for the method and equations that allow [CO 2 ] to be directly determined from measurements of solution electrical conductivity and temperature, with no empirical calibration. The equilibrium constants and ionic conductivities that are used are temperature dependent, but those dependences are well known and easily computed. The approach was tested with a system in which the conductivity was measured with time‐domain reflectometry (TDR), using a coaxial cell through which the aerated water was circulated. To maximize sensitivity, a long cell (1 m) of low impedance (19.3 Ω) was used. The system was compared against an infrared gas analyzer (IRGA) over a range of [CO 2 ] from 0 to 1000 μmol mol −1 , at three different temperatures (5, 19, and 34°C). Regression of [CO 2 ] calculated directly from the conductivity measurements against the IRGA measurements produced a slope of 1.00, r 2 of 0.997 and a standard error of estimate of 16.1 μmol mol −1 . Resolution was approximately 1 to 2 μmol mol −1 , too large for micrometeorological flux measurements but sufficient for many monitoring applications. The approach should work with any accurate device for measuring solution conductivity.