Potential drift

In potentiometry with ISEs, there are two significant causes of potential drift (other than variation in the analyte activity) variations in the liquid junction 

---

Unfortunately, in the presence of detectable polyions in the solution a strong potential drift is normally observed due to the instability of the ion concentration gradients. Moreover, the main disadvantage of polyion-selective potentiometric electrodes lies in the intrinsic irreversibility of the underlying response mechanism. The target polyions eventually displace the counter-ions in the membrane phase and consequently the sensor loses its response. 

The behavior of potentiometric and pulsed galvanostatic polyion sensors can be directly compared. Figure 4.11 shows the time trace for the resulting protamine calibration curve in 0.1 M NaCl, obtained with this method (a) and with a potentiometric protamine membrane electrode (b) analogous to that described in [42, 43], Because of the effective renewal of the electrode surface between measuring pulses, the polyion response in (a) is free of any potential drift, and the signal fully returns to baseline after the calibration run. In contrast, the response of the potentiometric protamine electrode (b) exhibits very strong potential drifts. 

Dioctyl sebacate (DOS) with relative permittivity e of 3.9 and 2-nitrophenyl octyl ether (NPOE) with e = 23.9 are the traditionally used sensor membrane plasticizers. The choice of a plasticizer always depends on a sensor application. Thus, NPOE appears to be more beneficial for divalent ions due to its higher polarity, but for some cases its lipophilicity is insufficient. Furthermore, measurements with NPOE-plasticized sensors in undiluted blood are complicated by precipitation of charged species (mainly proteins) on the sensor surface, which leads to significant potential drifts. Although calcium selectivity against sodium and potassium for NPOE-based membranes is better by two orders of magnitude compared to DOS membranes, the latter are recommended for blood measurements as their lower polarity prevents protein deposition [92],... 

The first and very simple solid contact polymeric sensors were proposed in the early 1970s by Cattrall and Freiser and comprised of a metal wire coated with an ion-selective polymeric membrane [94], These coated wire electrodes (CWEs) had similar sensitivity and selectivity and even somewhat better DLs than conventional ISEs, but suffered from severe potential drifts, resulting in poor reproducibility. The origin of the CWE potential instabilities is now believed to be the formation of a thin aqueous layer between membrane and metal [95], The dominating redox process in the layer is likely the reduction of dissolved oxygen, and the potential drift is mainly caused by pH and p02 changes in a sample. Additionally, the ionic composition of this layer may vary as a function of the sample composition, leading to additional potential instabilities. 

M. Fibbioli, W.E. Morf, M. Badertscher, N.F. de Rooij, and E. Pretsch, Potential drifts of solid-contacted ion-selective electrodes due to zero-current ion fluxes through die sensor membrane. Electroanalysis 12, 1286-1292 (2000). 

The integrated planar silver chloride electrode uses a thin layer of 150 pm polymer that consists of a heat curing epoxy resin poly-hydroxy-ethylmethacrylate (PHEMA) to immobilize the KC1 electrolyte. The potential drift of the reference electrode reduced to 59 pV/h after a conditioning phase of several hours. However, this reference electrode was only used for P02 measurement, while an external reference electrode was used for pH measurement. 

The ISE potential drift is connected with changes in the ISE surface structure and with dissolution of ion-exchanger ions or ionophores in test solutions. The values given in the literature differ considerably and may be as large as 2 mV per day. 

In general, the calibration curve method is suitable for all samples where the test substance is not bound in complexes or when it can be liberated from complexes by suitable sample pretreatment. Otherwise, the compositions of the samples and of the standard solutions must be as similar as possible to obtain results with acceptable accuracy. In view of the ISE potential drift, the calibration must be repeated often (at least twice a day). As mentioned above, the precision of the determination is not particularly high with a common precision of the potential measureihent at a laboratory temperature of 1 mV the relative error is 4% for univalent and 8% for divalent ions [58], However, this often suffices for practical analytical purposes. An advantage is that the same precision... 

The fluoride ion selective electrode is the most popular means of fluoride ion determination after sample destruction by any method but it does have limitations. It can be used either directly to measure the fluoride potential6 or as an end-point detector in a potentiometric titration with a lanthanum(l II) reagent as titrant.4,7 Problems can be experienced with potential drift in direct potentiometry, especially at low fluoride ion concentrations. Titration methods often yield sluggish end points unless water miscible solvents are used to decrease solubilities and increase potentia 1 breaks and sulfate and phosphate can interfere. End-point determination can be facilitated by using a computerized Gran plotting procedure.4... 

Catalyst deterioration due to gas poisoning is only avoided by careful gas cleaning. Anodic oxidation followed by dissolution of Pt and transfer to the cathode is a serious cause for Pt loss. It is potential dependent and accelerates as the cathode potential increases, for instance under partial load or in off-time, when the cathode potential drifts toward the oxygen equilibrium potential. Therefore it is of utmost importance that whenever the fuel cell is switched off, the oxygen in the cathode lumen is rapidly exchanged by inert nitrogen and that the cell voltage under operation does not surmount 0.8 V. 

This point may be located analytically by virtue of the fact that a semiconductor powder in a solution of adsorbing ions acts as a buffer for those ions everywhere but at the PZZP. Thus, potential drift or differential potentiometric titrations( 7,8 ) can be employed to determine the PZZP as illustrated in Figure 2 for CdS. Once the PZZP is determined in this fashion, a direct comparison of EA and is possible and has been done for a variety of semiconductors. (jO Figure 3 illustrates the vs. pH data for p-GaP and shows good agreement between the predicted at the PZZP from electronegativity calculations and the observed value.(9)... 

It is good practice to check carefully the electrochemical potential of the embeddable reference electrode against an accurate reference (SCE or Ag/AgCl), preferably in a laboratory, before the electrode is embedded in concrete. Normally, a saturated Ca(OH)2 solution is used as a test solution. By prolonging the exposure time in the solution, the magnitude of shortterm potential drift can be detected (be aware of temperature dependence). Potential values should always be compared with data provided by the supplier of the reference electrode. It is recommended that the functional and/or calibration check procedures given by the supplier are followed. 

The most common problem with hydrocarbon diffusion pump oil is its fractionation into multivapor pressure components. As pump oil breaks down, it develops both lower and higher vapor-pressure characteristics. Oils with high vapor pressures can potentially drift into the system, although they are more likely to be effectively removed from the system by being trapped in the alembics of the central vertical tube, in the cold trap between the system and the diffusion pump, or in the cold trap between the diffusion pump and the mechanical pump. If not trapped, they are free to travel into the vacuum line itself or into the mechanical pump. Diffusion pump oils that collect in a mechanical pump are not likely to have any significant performance effects (as opposed to the degrading effects of mechanical pump oil collected in diffusion pumps). 

Figure 9.2 Potential drift of pesticides onto children near schools or playgrounds.

However, most of the time, the corrective term to E° is unfortunately unknown and beyond the error associated with potential measurements on an absolute (NHE) scale, mainly because of junction potential and reference electrode potential drifts. Thus in actual experiments Ej/2 and E° (or E°) are generally considered identical [94]. 

A word of caution should be mentioned with respect to derivative methods. The derivatives tend to emphasize noise or scatter in the data points, being worse for the second derivative. Hence, if a particular titration is subject to noise or potential drift, a direct plot may be preferred. 

---