5. NOBLE GASES IN WATER – THE COLLABORATION WITH EAWAG

But once again we wondered for a short while whether RoKi and the rest of us might contribute to solve the world’s energy problem. On my way back from the 1989 LPSC, Otto Eugster of the University of Bern told me about a potentially sensational discovery. Fleischmann and Pons (1989) claimed to have observed nuclear fusion of deuterium atoms in an electrochemical reaction on a palladium electrode at room temperature. Although probably pretty much the entire scientific community was sceptical, the ensuing Cold Fusion hype was overwhelming, something I had never experienced before and have never experienced since. Reprints of their article were circulated worldwide by fax machines. The fax that I received after returning to Zürich was barely decipherable, so must have been a perhaps fifth or tenth generation copy. The idea of fusing hydrogen atoms at low temperature was not new. So called muon-catalysed fusion works well, although so far it is not a useful energy source because the energy required to produce muons exceeds the fusion output. At the Paul Scherrer Institute (PSI) near Zürich, muon-catalysed cold fusion had been studied for about ten years. Therefore, the group of Claude Petitjean decided to repeat the Fleischmann and Pons experiment, as did many other groups worldwide. We were asked to measure the ³He and ⁴He supposedly produced in the palladium electrode. To ensure that any ³He present resulted from fusion reactions and not simply from the decay of tritium originally present in the heavy (deuterium-enriched) water, we had insisted that the experiment be performed with tritium-free heavy water. This was not an easy requirement, since deuterium enrichment most likely also leads to even stronger tritium enrichment, unless one had started with “old” water, such as from glacial ice. Well, obviously this was not done, as the palladium sample analysed by RoKi showed a huge ³He signal, clearly from tritium which must have been highly enriched in the heavy water. Our extraction line subsequently delivered for months a very high ³He blank left over from that unfortunate sample, preventing many critical analyses. In any case, the PSI experiment gave no indications of fusion reactions à la Fleischmann and Pons (Blaser et al., 1989). The upper limit of ⁴He was at least six orders of magnitude below expectations for a neutron-free fusion rate as proposed by Fleischmann and Pons (1989). So, for a second time, we failed to do our part to solve the world’s energy problem.

In the late 1980s, Friedrich Begemann brought together researchers working in different fields of noble gas cosmo- and geochemistry at one of the meetings at Ringberg Castle in Bavaria (see Box 3.1). Peter Schlosser from the University of Heidelberg introduced us to noble gas hydrology. He told us about his passion shared with Dieter Imboden in using physics to solve environmental problems. Dieter was setting up his research group at the Swiss Federal Institute of Aquatic Science and Technology near Zürich (better known as Eawag) and the newly formed Department of Environmental Science at ETH. However, a close collaboration between Peter Schlosser and Dieter Imboden was not possible at that moment, since Peter was about to move to the Lamont-Doherty Earth Observatory of Columbia University in New York to establish a research team for noble gas hydrology with focus on oceanography. Peter therefore recommended that we contact Dieter. This proved to be an excellent and remarkably fruitful idea, which led to our group’s involvement in noble gas hydrology for more than 30 years, until this day. The most important person in this story is Rolf Kipfer who, thanks to Peter Schlosser’s advice and Dieter Imboden’s commitment, finally found his dissertation topic and his lifelong passion for noble gases in all kinds of waters.

Together with Heiri Baur, Markus Hofer and Urs Menet, RoKi set up the analysis techniques for determining noble gas concentrations and isotopic compositions in water samples. First, RoKi and his fellow doctoral student Werner Aeschbach-Hertig, together with their supervisor Dieter Imboden, investigated water dynamics in lakes. Over the years, RoKi’s research group at Eawag broadened their interests to include groundwater dynamics, palaeoclimate reconstructions using groundwater aquifers and speleothems, contamination of groundwater with arsenic, gas monitoring in the field with portable mass spectrometers, and other topics, many of which have direct impact on society. From this large research programme I will discuss a few select studies in the following, mostly projects done in collaboration with our group and biased towards work where I was personally involved.

The dynamics of water bodies in lakes can be studied by tritium-helium dating, based on the radioactive decay of ³H to ³He with a half-life of 12.3 years. Atmospheric nuclear bomb tests in the 1950s and early 1960s delivered large amounts of tritium to the atmosphere, oceans, and lakes such that until today T concentrations are above their natural (pre-bomb) levels. Analysis of the ³H and ³He concentrations in a water sample allows one to determine its “tritium-³He (T-³He) age”, which in the simplest case represents the time elapsed since the respective water parcel was separated from gas exchange with the atmosphere. Igor Tolstikhin developed the method in 1969 in Russia (Tolstikhin and Kamenskiy, 1969). Tritium concentrations in water are often measured by two successive ³He analyses. A sample that has been completely degassed for a first analysis of ³He is stored for a few months to build up tritiogenic ³He in the laboratory and is then analysed a second time, which yields the tritium concentration of the water. In our laboratory, the high sensitivity Tom Dooley mass spectrometer (Section 3.1.8) is very well suited to measure even low concentrations of tritium in relatively small water samples. Actually, nominal T-³He ages often do not represent true times when one water parcel got isolated from the atmosphere, but rather reflect mixing of multiple water masses with different histories. The T-³He dating technique is very powerful precisely for studying such mixing processes. A fine example is provided in the dissertation of Werner Aeschbach-Hertig, who with his colleagues studied the water mixing dynamics in Lake Lucerne, one of the most beautiful lakes in the Alps (Fig. 5.1; Aeschbach-Hertig et al., 1996a). It has a remarkably complex topography, consisting of six distinct basins divided by sills. In winter, significant density differences between individual basins drive deep water exchange between them. Werner dubbed Lake Lucerne the “Swiss Miniature Ocean”, alluding to work by others who study ocean mixing processes by T-³He dating. Tritium-He dating also allowed Roland Hohmann in his dissertation to determine the renewal rates of deep water in Lake Baikal, the deepest and by volume largest lake on Earth (Hohmann et al., 1998). In several further studies, RoKi, Werner Aeschbach-Hertig, and colleagues determined the fluxes of mantle-derived noble gases, e.g., in Lake Van in Anatolia, and Laacher See in the Eifel in Germany (Kipfer et al., 1994; Aeschbach-Hertig et al., 1996b). A contribution from mantle gas is most conspicuous in helium, since the ³He/⁴He ratio in the mantle differs strongly from the atmospheric value.

Noble gases in suitable aquifers provide information on palaeotemperature, because their solubilities are temperature dependent, especially for Ar, Kr, and Xe (Fig. 5.2). A precise determination of the concentration ratios of noble gases in groundwater samples with known infiltration ages therefore allows one, for example, to determine mean annual temperature differences between the Last Glacial Maximum (LGM) and the Holocene. Dissolved noble gases are particularly useful palaeotemperature recorders because they are insensitive to chemical and biological processes. After completing his doctoral thesis at ETH/Eawag, Werner Aeschbach-Hertig worked on this topic as a postdoc in Peter Schlosser’s group at Lamont, before he returned to Europe, first to Eawag and then to the University of Heidelberg. Werner developed an improved scheme to correct for noble gases often present in water in excess of concentrations in equilibrium with the atmosphere (Aeschbach-Hertig et al., 2000). Urs Beyerle and co-workers applied a similar correction for such “excess air” to data from an aquifer in Switzerland and concluded that the mean annual air temperature in central Europe during the last ice age was at least 5 ºC below the Holocene average (Beyerle et al., 1998). By compiling data from many aquifers, Seltzer et al. (2021) reported similar temperature differences between the LGM and the Holocene for mid-latitudes and tropical regions, arguing against earlier claims that mean air temperatures near the equator were barely cooler during the LGM than today.

I had so far participated in only a few groundwater studies, but I became involved in the decoding of an even more spectacular palaeoclimate archive: speleothems (Fig. 5.3). When visiting caves we are fascinated by the long calcite structures that grow slowly due to precipitation of CaCO₃ from meteoric water. Especially the stalagmites that grow from the cave floor are excellent palaeoclimate recorders (McDermott et al., 2005). Oxygen isotopes measured along a profile parallel to the growth axis of a stalagmite are used to reconstruct the mean annual temperature during its growth phase (Fleitmann et al., 2004). However, the interpretation of the oxygen record is not easy, because the O isotopic composition of the calcite is not determined by the cave temperature alone. Therefore, an independent temperature recorder in stalagmites is highly desirable. RoKi’s hope was that noble gases could provide such a thermometer, based on the same physical principle that has proven so successful in aquifers, with the caveat, of course, that the amounts of water available in a stalagmite sample are very much smaller than those in a groundwater sample. Regardless, RoKi approached Dominik Fleitmann, a speleothem researcher then at the University of Bern. The resulting collaboration yielded two doctoral theses, by Yvonne Scheidegger and Elaheh Ghadiri, and also Nadia Vogel was heavily involved in this work. Nadia had found her way back to ETH and Eawag after postdoctoral positions in Paul Renne’s group at UC Berkeley and the University of Bern.

The development of analytical capabilities to use stalagmites as reliable noble gas palaeothermometers truly deserves the word “challenging”. There were two big problems. First, a typical calcite sample of a few hundred milligrams contains no more than a few mg of water in fluid inclusions, compared to the typical 45 grams we use in groundwater studies. Second, and even more challenging, stalagmites not only contain water-filled inclusions but also inclusions of air (see Fig. 5.3). The noble gases in the latter have atmospheric element ratios and, if included in the gas inventory, severely disturb the temperature signal represented by the noble gases dissolved in the water inclusions. Also problematic are atmospheric noble gases adsorbed during crushing of the calcite sample cubes. To some extent, atmospheric contamination can be corrected for by the same methods as “excess air” in groundwater. However, a substantial reduction of atmospheric noble gases during sample preparation was first required. Yvonne Scheidegger developed an analysis protocol that largely achieved this (Scheidegger et al., 2011). She first crushed the calcite cubes in a noble gas-free atmosphere and then selected only grains of a certain size range. Because air inclusions tend to be larger than water inclusions, this reduced contamination by the former. In favourable cases this allowed her to correct for the remaining atmospheric noble gases and determine the cave temperature at the time the water-filled inclusions formed.

However, Yvonne’s work also showed that an improved way to reduce contributions from noble gases from air inclusions was called for. We therefore approached our workshop technician Andreas Süsli with a truly difficult request: could Andreas build a device with which we could crush calcite cubes of a few mm in size and at the same time sort the crushed material into three different grain size fractions. This would all have to be done in ultrahigh vacuum and in such a way that the noble gases in the different fractions could be released and analysed without breaking the vacuum. It turned out that we had not overestimated Andreas’ ingenuity. The device he constructed is shown in Figure 5.3c. The “combined vacuum crushing and sieving system” allowed Nadia Vogel to reduce atmospheric contamination in the best sieve fraction by up to two orders of magnitude, and the noble gas temperatures she derived reliably reproduced known growth temperatures of several stalagmite test samples (Vogel et al., 2013). Elaheh Ghadiri then used the technique to determine the temperature reconstruction of the last glacial-interglacial transition (Younger Dryas) and the altitude gradient of this temperature with stalagmites in the Swiss Jura (Ghadiri et al., 2018, 2020).

The vacuum crusher and siever once again demonstrated the enormous advantage of having highly qualified and well trained technical personnel fully integrated into the research group, enabling close daily interaction with the scientists. This is a big asset offered by ETH. Other examples already mentioned include the Closed System Stepped Etching devices, especially the “gold and platinum line”, and the “Compressor ion source” in the Tom Dooley mass spectrometer.

The last topic I would like to raise is unrelated to my own work, but I mention it because noble gas analysis directly in the field has a promising future. The classic way isotope geochemists and environmental scientists analyse noble gases in fluid or gaseous samples is to collect them in the field in some type of vessel. The samples – usually in very limited numbers – are then transported to a laboratory to be analysed there with more or less sophisticated equipment. In many cases direct on site analysis of gas concentrations and perhaps also key isotopic ratios would be highly beneficial, as this would reduce logistics such as sample storage and transport, and, most importantly, allow on line monitoring or preliminary investigations that would help select the most appropriate samples for subsequent detailed laboratory analysis. The dynamics of gases or fluids in environmental systems often can only be adequately studied with a large number of samples taken at multiple points in space and time. A promising path toward this goal is to bring the analysis system into the field. Around 1980, John Reynolds and co-workers in Berkeley had developed a movable mass spectrometry system that was self-contained except for electrical power. The system was hosted in a truck and was applied, e.g., to study noble gases in fluids at Yellowstone (Kennedy et al., 1985). However, continued progress in the miniaturisation of mass spectrometers is now leading to the development of much smaller portable systems, which allow field analyses of noble gases but also of biogeochemically active gases (Mielczarek et al., 2020). Matthias Brennwald, Lars Mächler, Rolf Kipfer, and colleagues at Eawag have developed a commercialised portable mass spectrometry system the size of a large suitcase (Brennwald et al., 2016). It measures the partial pressure of noble gases, N₂, O₂, CO₂, and CH₄ in gaseous and aqueous matrices in environmental systems. The system allows essentially maintenance-free and autonomous operation, with power supplied by, e.g., two conventional car batteries. Yama Tomonaga and co-workers used this device to monitor the free gas phase in an underground rock laboratory in Switzerland in the context of a large scale radioactive waste management test (Tomonaga et al., 2019). The system has also been applied, for example, at a CO₂ capture facility in Norway (Weber et al., 2021), to analyse noble gases, CO₂, and N₂ at the East African Rift system in Tanzania within a few hours of sampling (Mtili et al., 2021), and to study geogenic arsenic contamination of groundwaters with noble gases (Lightfoot et al., 2022).