The anoxic corrosion of metals can be indirectly measured through the release of hydrogen gas. This method is typically used in the waste management field due to its applicability to the very low corrosion rates expected in certain deep geological repository concepts (in the range of a few nm/year or less) which make the use of e.g. electrochemical or gravimetric techniques difficult. The measurement of released hydrogen can be done through either direct analysis of the hydrogen itself, using a hydrogen detector coupled with a purge and collect methodology [1–3], or measurement of pressure changes due to the accumulation of hydrogen [4–6]. This paper discusses the advantages and disadvantages of both techniques as applied in the measurement of the corrosion of copper in bentonite and of steel in cement. Also included in this paper is a discussion of the cell design, used in ongoing work, which has evolved to minimise hydrogen loss and oxygen ingress.
The direct analysis of hydrogen, using a technique such as mass spectrometry or a solid state hydrogen sensor, requires removal of analyte from the test cell. This permits good sensitivity, as hydrogen can be left to accumulate over a period of days or even months; uniform averaged corrosion rates as low as 0.01 nm/year can be monitored. However, this approach has a number of draw-backs. These include resetting of corrosion equilibria that include hydrogen, such as 2Cu + 2H+ ⇌ 2Cu+ + H2 and the potential for the introduction of trace amounts of oxygen. Other considerations include the recovery efficiency of hydrogen from the cell, which is a function of the relative volume of purge gas used to evacuate the cell, concomitant with the dilution of the cell hydrogen by the purge gas. This may reduce the hydrogen concentration to the limits of the detector, so recovery efficiency and concentration always require a compromise to maximise the quality of the data. This approach is also limited if changes are being made to a cell, such as dosing hydrogen sulfide to a copper-containing cell; the sulfide must be given sufficient time to react before the cell is purged, otherwise sulfide is removed and the mass balance calculation cannot be completed. This results in significant averaging of the copper corrosion rate across the hydrogen accumulation period so the true maximum corrosion rate cannot be known. Another consideration is that the purge and collect method may involve the collection of cell gas in a chromatography bag, which itself is not entirely impermeable to hydrogen. This approach also only analyses the hydrogen accumulated in the head-space; hydrogen dissolved in the electrolyte, trapped in the cement pores, or absorbed by the test specimens does not feature in these calculations.
Pressure monitoring can be performed directly of the cell head-space [5], or of the hydrogen partial pressure directly [7], using a palladium membrane that will only permit the transfer of hydrogen. The former method is very sensitive to the water vapour pressure, which itself can vary with fluctuations in water bath temperature, or changes in room temperature or direct sunlight, of cell components exposed to air.
Regarding cell design, the current cells use borosilicate glass with standard taper ground glass fittings. Cell components are glued together with epoxy for sealing, creating a long path length for diffusing hydrogen to escape the cell. Materials such as nitrile or viton are avoided as being too hydrogen and oxygen permeable. Stainless steel fittings are glued directly onto glass for sealing purposes, and 3-way ball valves incorporate a tube bridge which accumulates any hydrogen diffusing through the PTFE ball-valve lining, both minimising hydrogen loss through equilibration and permitting direct measurement. Cells are leak-tested using helium at twice the water vapour pressure of the experiment, before use and at various intervals during cell operation. Examples of hydrogen loss are provided from blank cell and valve tests.
[1] T. Sakuragi, S. Yoshida, O. Kato, T. Tateishi, “Study of stainless steel corrosion by hydrogen measurement under deoxygenated, low-temperature and basic repository conditions,” Prog. Nucl. Energy. 87 (2016) 26–31. doi:10.1016/j.pnucene.2015.10.008.
[2] N. Senior, R. Newman, S. Wang, N. Diomidis, “Understanding and quantifying the anoxic corrosion of carbon steel in a Swiss L/ILW repository environment,” Corros. Eng. Sci. Technol. 52 (2017) 78–83. doi:10.1080/1478422X.2017.1303102.
[3] N.A. Senior, R.C. Newman, D. Artymowicz, W.J. Binns, P.G. Keech, D.S. Hall, “Communication—A Method to Measure Extremely Low Corrosion Rates of Copper Metal in Anoxic Aqueous Media,” J. Electrochem. Soc. 166 (2019) C3015–C3017. doi:10.1149/2.0031911jes.
[4] P. Kreis, J.P. Simpson, “Hydrogen Gas Generation from the Corrosion of Iron in Cementitious Environments,” (European Federation of Corrosion, published by the Institute of Materials, 1992.
[5] N.R. Smart, B. Reddy, A.P. Rance, D.J. Nixon, N. Diomidis, “The anaerobic corrosion of carbon steel in saturated compacted bentonite in the Swiss repository concept,” Corros. Eng. Sci. Technol. 52 (2017) 113–126. doi:10.1080/1478422X.2017.1316088.
[6] N.R. Smart, D.J. Blackwood, G.P. March, C.C. Naish, T.M. O’Brien, A.P. Rance, M.I. Thomas, “The Anaerobic Corrosion of Carbon and Stainless Steels in Simulated Repository Environments- A Summary Review of Nirex Research.,” Harwell, UK), 2004, AEAT/ERRA-0313.
[7] G. Hultquist, M.J. Graham, P. Szakalos, G.I. Sproule, A. Rosengren, L. Gråsjö, “Hydrogen Gas Production During Corrosion of Copper by Water,” Corros. Sci. 53 (2011) 310–319. doi:10.1016/j.corsci.2010.09.037.
anoxic, corrosion, pressure, hydrogen, measurement, sensitivity