Dissolved H2 in groundwater & formation water

Summary

Every other technique in this compendium measures hydrogen as a gas — molecules arriving at a sensor surface from soil pore space, a wellbore headspace, or a lab gas stream. Dissolved H2 work starts one step earlier: the hydrogen is still in the water, at concentrations so low (nanomolar to micromolar, not the parts-per-million-to-percent range this wiki otherwise deals in) that pulling a meaningful number out of a well or spring without wrecking it in the act of sampling is the central challenge.

This matters even at global-inventory scale: the largest published synthesis of natural-gas H2 occurrences (6,246 samples across 16 geological habitats) reports a mean H2 concentration of 3.5% but a median of only 0.01%, singling out dissolved-phase transport as one of the under-exploited pathways by which H2 migrates and accumulates away from the handful of high-concentration habitats that dominate the mean [31].

The reason the concentrations seen in groundwater specifically are so low is not just that water doesn’t dissolve much H2 to begin with — it’s that a dense, active microbial community living in almost every aquifer treats dissolved H2 as its preferred food source, and holds the standing concentration down to whatever level the locally dominant microbial metabolism can tolerate.

This is useful: the exact concentration of dissolved H2 in a water sample tells you which redox process — denitrification, iron reduction, sulfate reduction, methane-making — currently dominates that part of the aquifer, information no gas-phase soil or wellhead reading can give you.

It is also one of the most fragile measurements in Hestia’s toolkit: pulling the sample to surface, exposing it to air for even a few minutes, or drilling the well itself with a steel casing can each move the number by ten to a thousand times — enough to erase the redox distinction the measurement is meant to reveal. With rigorous sample-handling discipline, dissolved H2 shows whether a site’s subsurface biology is consuming a geologic H2 flux or letting it pass; without that discipline, the number is close to meaningless.

Mechanism & physics

Henry’s law and the solubility ceiling

At equilibrium, dissolved H2 concentration is set by Henry’s law,

$C_{aq} = k_H \, p_{H_2}$

where $C_{aq}$ is aqueous molar concentration, $p_{H_2}$ is the H2 partial pressure in contact with the water, and $k_H$ is the Henry’s-law solubility constant. NIST’s compiled value for H2 in pure water at 25 °C is $k_H \approx 7.8\times10^{-4}\ \mathrm{mol\,kg^{-1}\,bar^{-1}}$, consistent with an independently cited 8.13×10⁻⁴ M/atm figure [1].

H2 is a poorly soluble gas by the standards of everything else in this wiki: its dimensionless (concentration-ratio) Henry’s constant, $H_{cc}=C_{gas}/C_{aq}=1/(k_H RT)$, works out to $H_{cc}\approx52$ at 25 °C [1] — roughly 52 times more H2 sits in a gas phase than in an equal volume of water at equilibrium with it. This is why dissolved-phase H2 exploration is intrinsically a trace-analysis problem even directly above a percent-level free-gas anomaly.

Applying $C_{aq}=k_H p_{H_2}$ across the exploration-relevant pressure range gives the anchor values worth memorizing:

$p_{H_2}$	Physical meaning	$C_{aq}$
5.3×10⁻⁷ atm (0.53 ppm)	Global tropospheric H2 background	≈0.41 nM (≈0.8 ng/L)
1×10⁻⁶ atm (1 ppm)	Minimum soil-gas anomaly threshold	≈0.78 nM
1×10⁻² atm (1 vol %)	Modest free-gas seep or reservoir gas cap	≈7.8 µM
1 atm (100%)	Pure H2 gas phase / saturation	≈7.8×10⁻⁴ M ≈ 1.6 mg/L

Solubility anchor values

The saturation figure (≈1.6 mg/L at 1 atm, 25 °C) is independently corroborated by the “hydrogen water” biomedical literature [2], and the atmospheric floor (≈0.41 nM) agrees with an independently cited background of ≈8.65×10⁻⁷ mg/L in conventional water [2] — two calculations converging from different directions is a useful sanity check. The exploration-relevant consequence: even a reservoir-grade free-gas anomaly (1 vol%) yields an equilibrium dissolved concentration of only ~8 µM — three to four orders of magnitude below what a naive “ppm-level signal” intuition would suggest, before any biological sink (below) pulls the observed number down further.

Temperature: a non-monotonic complication

Unlike CO2 or CH4, whose aqueous solubility falls monotonically with rising shallow-groundwater temperature, H2’s Henry’s constant shows a solubility minimum near 320 K (~47 °C) at moderate pressure (~200 bar), after which solubility increases with further heating — captured in high-pressure solubility models fit across 273–423 K [7] and reproduced in molecular-simulation/EOS work [8].

At true deep-formation conditions the effect is large: a study of H2 in simulated pressurized-water-reactor primary water found the Henry’s constant decreased by ~40% as temperature rose from 290 °C to 330 °C [9] — H2 becomes markedly more soluble at reservoir temperatures in the hundreds of degrees.

In practice, the “colder water holds more gas” intuition, true for H2 across the shallow 0–50 °C groundwater range, inverts somewhere in the 50–300 °C window relevant to deep sedimentary basins. Any solubility-based mass balance for a deep reservoir must use a model validated at actual in-situ T/P — see Downhole deployment for the broader high-T/P instrumentation context.

Salinity: salting-out in formation brine

Dissolved-gas solubility in electrolytes follows the Setschenow (salting-out) relation,

$\log_{10}\!\left(\frac{S_0}{S}\right) = K_S\, m_{salt}$

where $S_0$ is pure-water solubility, $S$ is solubility at salt molality $m_{salt}$, and $K_S$ is a gas- and salt-specific coefficient. The foundational atmospheric-pressure dataset (Crozier & Yamamoto, 1974, spanning distilled water, NaCl solutions, and seawater from −2 to 30 °C) remains the reference baseline against which high-pressure brine studies are anchored [3].

High-pressure H2–NaCl brine measurements (to 5 mol/kg, 100–373 K, up to 230 bar) find that each additional 1 mol/kg NaCl reduces H2 solubility by roughly 15%, essentially independent of pressure and temperature [4][6], corroborated by an independent dataset to 2.5 mol/kg and 423 K/~40 MPa [5]; the salting-out order for common formation-water cations follows KCl < NaCl < CaCl2 [6]. Usefully, H2 shows a weaker salting-out response than N2, O2, or CH4 [6] — a hypersaline formation water suppresses dissolved H2 proportionally less than it suppresses those co-measured gases, which matters when interpreting gas ratios in produced water.

Applied to a real site: the FF4 natural-hydrogen seep in Western Australia has groundwater salinity ~77,000 ppm TDS [34], roughly 1.3–1.5 mol/kg NaCl-equivalent, implying a 20–25% reduction in H2 solubility relative to fresh water at the same T/P before any biological sink is considered. Neglecting the salting-out correction has been shown to introduce errors exceeding 200% in underground-hydrogen-storage mass-balance calculations [6], and the same sensitivity applies to back-calculating a subsurface free-gas H2 saturation from a produced-water dissolved-H2 measurement.

The redox-indicator mechanism — why this is not simply Henry’s-law equilibrium with a deep source

The key theoretical point distinguishing dissolved H2 from every gas-phase modality elsewhere in this wiki is that the dissolved H2 concentration observed in anoxic groundwater is almost never a passive Henry’s-law equilibrium value reflecting a deep source’s partial pressure; it is a microbially imposed steady state.

H2 is the shared intermediate electron donor (“interspecies hydrogen”) in essentially every anaerobic respiratory pathway in groundwater: denitrification, Mn(IV)/Fe(III) reduction, sulfate reduction, and methanogenesis all run, wholly or partly, on hydrogenase-catalyzed oxidation of H2 coupled to reduction of whatever electron acceptor is locally available. Because multiple competing microbial guilds draw on the same H2 pool, and because H2-utilizers are essentially the only organisms that can run that reaction, the standing concentration collapses to whatever level the currently dominant guild can still profit from [10][11].

The mechanism is thermodynamic. For a coupled H2-oxidation/electron-acceptor-reduction reaction, $\Delta G_{rxn} = \Delta G^{\circ\prime}_{rxn} + RT\ln Q$, and because H2 sits on the reactant side of every one of these half-reactions, lower ambient [H2] makes the reaction more favorable for a given acceptor. Direct laboratory measurement in anoxic marine sediment found sulfate-reducing bacteria operate down to an apparent minimum free-energy yield of −19.1 ± 1.7 kJ (mol SO4²⁻)⁻¹, while methanogenic archaea can be sustained on yields as small as −10.6 ± 0.7 kJ (mol CH4)⁻¹ [43], both consistent with the widely cited “biological energy quantum” of roughly −15 to −25 kJ/mol below which no known organism can conserve net energy [42][43].

Because standard free energies of the respective electron acceptors decline in the sequence O2 > NO3⁻ > Mn(IV) > Fe(III) > SO4²⁻ > CO2 (methanogenesis), each successive guild can only clear that same energy floor by tolerating a higher standing H2 concentration than the guild before it — nitrate/Mn(IV) reducers can scavenge H2 down to sub-0.1 nM, while methanogens, using the least favorable acceptor, only reach that threshold once ambient H2 has built to 5–15 nM [10][11][12][42][43].

Competitive exclusion — the guild with the lowest operating threshold wins the shared pool and holds it there — makes a measured dissolved-H2 concentration a field-usable redox indicator: it is diagnostic of which terminal electron-accepting process (TEAP) currently dominates that parcel of water, largely independent of how much H2 a deep source is actually generating.

This has a non-obvious exploration implication: a low dissolved-H2 reading in a monitoring well does not, by itself, argue against a strong deep source.

If the aquifer still has nitrate or reactive Fe(III)/Mn(IV) minerals available, an actively consuming microbial community can hold ambient H2 at sub-nanomolar levels even while a substantial upward flux is passing through — this is the anaerobic, subsurface counterpart of the aerobic soil “high-affinity hydrogen-oxidizing bacteria” (HA-HOB) sink documented on Geologic context and Environmental covariance & cross-sensitivity. Dissolved-H2 campaigns must be interpreted jointly with the aquifer’s broader redox chemistry (dissolved O2, NO3⁻, Fe²⁺, sulfate, sulfide, CH4, alkalinity/pH), never as a standalone number.

Mass balance for field interpretation

For a monitoring well or spring, a useful lumped steady-state framework is

$\frac{d[H_2]}{dt} = \underbrace{\frac{J_{geo}}{V}}_{\text{deep flux}} - \underbrace{k_{bio}[H_2]}_{\text{microbial consumption}} - \underbrace{v_{d}\,\frac{A}{V}\,\big([H_2]-[H_2]_{atm\text{-}eq}\big)}_{\text{degassing/atm. exchange}} = 0$

At steady state in a closed, biologically active system, $[H_2]_{ss}\approx J_{geo}/(V k_{bio})$ — the observed concentration scales with the ratio of source strength to consumption rate, not source strength alone. This is why dissolved H2 is a poor flux quantification tool on its own (a given concentration is consistent with a wide range of $J_{geo}$ depending on the unmeasured $k_{bio}$) but an excellent qualitative redox and source-confirmation tool combined with independent flux estimates from soil-gas surveys or mud-gas logging.

Variants & subtypes

Dissolved-H2 work divides along two axes: how the sample is obtained, and how the extracted or in-situ gas is analyzed.

Sampling methods

1. Bubble-strip / gas-stripping. A small, fixed-volume gas bubble is equilibrated with actively flowing groundwater inside a sampling cell (the glass “Chapelle cell,” ~230 mL, or the smaller, faster-equilibrating commercial “Microseeps cell,” ~70 mL) plumbed in-line at the wellhead or downhole; water is pumped through until the bubble reaches gas–liquid equilibrium, and its H2 content is read out and back-calculated via Henry’s law [13][14]. This is the method Chapelle’s USGS group developed and validated through the 1990s specifically for hydrogen, and per a 2005 optimization study it remains “the only method that exists for measuring dissolved hydrogen concentrations as low as 0.1 nM” with field practicality [15].
2. Downhole (in-situ) sampler. A sealed device is lowered into the well and allowed to equilibrate with formation water in place before retrieval, avoiding the pumping and surface-transit steps that risk degassing; validated as concordant with bubble-strip results, at the cost of a much longer (~2 h) equilibration time [13].
3. Diffusion (passive) samplers. A gas-permeable membrane — silicone tubing over a sintered-metal support, or a plunger-less syringe sealed behind a double layer of LDPE membrane — is left in place in a well, piezometer, or fractured-rock interval a pump cannot reach, and allowed to reach diffusive equilibrium (typically ~12 h) before retrieval and headspace analysis [16][17]. Slowest but least invasive, and the only field-validated approach that works where pumping is impossible.
4. Headspace-equilibration (“agitation”) method. A discrete grab sample is sealed in a vessel with a defined headspace, vigorously agitated to drive gas-phase equilibrium, and the headspace analyzed with a portable multi-gas meter or GC. Architecturally simplest, and the method increasingly used in current natural-hydrogen field campaigns precisely because it works on groundwater from dedicated shallow boreholes where dry soil-gas methods cannot reach — a 2024 Western Australian study on the FF4 seep measured headspace H2 up to 1,684–1,694 ppm in agitated groundwater with an iBrid MX6 portable meter (0–2,000 ppm electrochemical H2 channel, 1 ppm resolution) [33][36], far higher than any co-located dry soil-gas reading and interpreted as evidence that water itself was trapping and transporting H2 that dry-soil methods missed entirely [33].
5. Downhole formation-fluid (“wireline”) sampling at reservoir pressure. For deep formation water, oilfield-style tools — MDT-class wireline formation testers or U-tube-style borehole fluid samplers — capture formation water while deliberately keeping it above its bubble point, avoiding the phase separation a large pressure drop would otherwise cause. The first published application to a natural-hydrogen target (a Colombian basement/basin play) reported reservoir formation water with dissolved H2 content of ~0.9 g/m³ at surface (STP) conditions, correcting for H2 partitioning into drilling-mud filtrate, and explicitly targeted intervals where dissolved H2 “has reached concentrations beyond the bubble point” as an indicator of nearby free-gas accumulation [32].

Analytical back ends

• Headspace/stripped-gas GC with reduction-gas detection (GC-RGD/RCP) — the historical and still-dominant back end for both bubble-strip and headspace-equilibration methods. See Gas chromatography.
• GC-PDHID — the modern, more linear alternative RGD/RCP detector chemistry, architecturally interchangeable on the same stripped/headspace-gas front end.
• Membrane-inlet mass spectrometry (MIMS) — first developed as a general dissolved-gas technique using a gas-permeable membrane to admit liquid-phase gas directly into an MS vacuum system [21], now available in bench (Hiden HPR-40 DSA [25]), field-portable (Gasometrix miniRUEDI lineage), continuous-flow (CF-MIMS [23]), and pressure-tolerant gas-equilibrium (GE-MIMS, rated to 100 bar [24]) variants — the only technique family here that measures dissolved gas directly, with no discrete stripping or headspace step. See Mass spectrometry.
• Electrochemical (Clark-type, amperometric) dissolved-H2 microsensors — miniaturized, membrane-covered platinum-electrode probes (Unisense’s H2 series is the dominant commercial example) for direct insertion into water, porewater, or sediment, reading dissolved H2 continuously with no gas-phase extraction step at all [18][19][20]. See Electrochemical cells for the transduction physics family.
• Optical dissolved-H2 sensors — a thin, largely research-stage category. Palladium-film fiber-optic sensors are mature for gas-phase H2 (see Fiber optic sensors, Exotic optical), but no commercial fluorescence- or absorbance-based dissolved-phase H2 optode analogous to the well-established dissolved-O2 optical sensor product lines was identified — a genuine market gap flagged again below.

Performance envelope

The redox-indicator ladder

Two foundational quantitative datasets underlie the “dissolved H2 as redox indicator” concept, both traceable to Lovley and Chapelle’s work from the late 1980s through the mid-1990s. They report slightly different numeric bins because they synthesize different, overlapping sets of field and lab sediments/aquifers; both are reproduced rather than silently reconciled, because the existence of a small, literature-normal spread is itself useful context.

Terminal electron-accepting process (TEAP)	Lovley & Goodwin 1988 (aquatic sediments) [10]	Chapelle et al., multi-site groundwater synthesis [11][12]
Mn(IV) / nitrate reduction	<0.05 nM	<0.1 nM
Fe(III) reduction	~0.2 nM	0.2–0.8 nM
Sulfate reduction	1–1.5 nM	1–4 nM
Methanogenesis	7–10 nM	5–15 nM

Both ladders follow directly from the thermodynamics above: each row needs a higher standing H2 concentration because its electron acceptor yields less free energy per mole of H2 oxidized [10][11][12][42][43].

As a high-concentration anchor at the opposite end of the scale, dissolved H2 in actively H2-oxidizing Yellowstone hot-spring microbial mats — an aerobic/thermophilic ecosystem rather than the anoxic TEAP framework, but governed by the same “H2 as primary energy currency” logic — has been measured up to and exceeding 300 nM, with H2 oxidation identified as the dominant primary-production pathway for most of the biomass present [39]. The nanomolar ladder above is calibrated to temperate, near-neutral anoxic groundwater; geothermally influenced or seep-proximal settings can run one to two orders of magnitude higher while still being microbially, not Henry’s-law, controlled.

Method comparison

Method	Typical LOD (dissolved H2)	In-situ?	Equilibration/turnaround	Dominant artifact risk	Ref
Bubble-strip (Chapelle cell, 230 mL) + GC-RGD	~0.1 nM	No (wellhead flow-through)	~2 h to full equilibrium (30 min operationally accepted)	Slow equilibration allows atmospheric ingress/microbial consumption during the run	[13][14]
Bubble-strip (Microseeps cell, 70 mL) + GC-RGD	~0.1 nM (same detector, faster kinetics)	No (wellhead flow-through)	95–100% equilibrium in 10–15 min at >100 mL/min flow	Same class of risk, substantially reduced by shorter run time	[14][15]
Downhole (in-situ) sampler	Comparable to bubble-strip when cross-validated	Yes	~2 h	Long borehole dwell time; well-casing H2 generation (below)	[13]
Diffusion/passive sampler (silicone- or LDPE-membrane)	Sub-nM range reported feasible; not independently re-verified	Yes	~8–12 h	Slowest method; incomplete equilibration if retrieved early	[13][16][17]
Headspace-equilibration (“agitation”) + portable electrochemical multi-gas meter	~1 ppm gas-phase resolution; a much coarser dissolved-phase LOD than bubble-strip/GC once back-calculated	No (discrete grab + agitation)	Minutes	Agitation itself is the extraction mechanism — good for confirming a strong seep, poor for resolving the sub-15 nM TEAP ladder	[33][36]
RGD/RCP on stripped or headspace gas (bench/portable GC)	Manufacturer-quoted 25–50 ppb (gas-phase); a 50 ppb dissolved-H2-headspace LDL is factory-quoted for the AMETEK ta3000R	No (surface analytical step)	Minutes per injection	HgO-bed consumable depletion, drift; blind to non-reducing co-eluting species	[27], cross-ref Gas chromatography
Membrane-inlet MS, bench (Hiden HPR-40 DSA)	Species-dependent; realistic H2 floor likely ppm-level in gas-equivalent terms owing to vacuum-system H2 background	No (bench, flow-through membrane probe)	Seconds–minutes	Vacuum-system H2 outgassing background; no published H2-specific groundwater calibration found	[25], cross-ref Mass spectrometry
Field-portable / continuous-flow MIMS (miniRUEDI, CF-MIMS, GE-MIMS)	1–5% RSD precision reported for He/Ar/N2/O2/Kr; H2 listed in GE-MIMS’s gas suite but not separately validated	Yes (field-deployable, some autonomous/solar)	Seconds (CF-MIMS) to ~12–20 min (base miniRUEDI)	H2’s low relative sensitivity factor and fast membrane permeation complicate a dedicated calibration; no H2-specific field validation located	[22][23][24]
Unisense H2/H2-X Clark-type electrochemical microsensor	0.3 µM (300 nM) manufacturer-specified LOD; 0–2,000 µM range	Yes (direct insertion)	t90 <15 s (standard tips); <0.3 s optional fast-response tip	Stirring sensitivity <2%; 2–3%/°C temperature coefficient; H2S cross-sensitivity on standard (non-X) tips	[18][19][20]

Read together with the TEAP ladder above, one number in that table stands out: the Unisense electrochemical microsensor’s manufacturer-quoted detection limit (0.3 µM = 300 nM) sits above the entire Lovley–Chapelle TEAP ladder (0.05–15 nM) [10][11][12][18].

A Clark-type microsensor is a field-proven tool for porewater, biofilm, and microbial-mat H2 work where concentrations commonly run into the hundreds-of-nM-to-µM range [39] — fast, continuous, and in-situ — but out of the box it is not sensitive enough to resolve the nanomolar-scale redox zonation that is the main geochemical payoff of dissolved-H2 measurement in groundwater. Bubble-stripping to a GC-RGD/RCP interface, despite being slower and more artifact-prone in execution, remains the only field-validated route to the sub-nanomolar sensitivity the TEAP framework requires [13][14][15].

Physical-chemistry quantities worth keeping on a reference card

Quantity	Value	Conditions	Ref
Henry’s constant, H2–H2O	7.8×10⁻⁴ mol L⁻¹ atm⁻¹ (≈0.00078 mol kg⁻¹ bar⁻¹)	25 °C, dilute solution	[1]
Dimensionless Henry’s constant $H_{cc}$	≈52	25 °C	derived
Saturation dissolved H2 at 1 atm pure H2	≈0.8 mM ≈ 1.6 mg/L	25 °C, 1 atm	[1][2]
Atmospheric-equilibrium dissolved H2 floor	≈0.4 nM (≈8×10⁻⁷ mg/L)	25 °C, $p_{H_2}$ = 0.53 ppm	derived, cross-checked against [2]
Setschenow-type salinity effect	≈−15% solubility per +1 mol/kg NaCl	100–373 K, up to 230 bar, up to 5 mol/kg NaCl	[4][6]
Salting-out order (strength)	KCl < NaCl < CaCl2	Formation-brine-relevant cations	[6]
H2 solubility minimum (isobaric)	~320 K (~47 °C) at ~200 bar	High-pressure H2–H2O system	[7][8]
Henry’s constant change, high-T reservoir water	≈−40% (solubility increases)	290→330 °C at 20 MPa	[9]
Colombia downhole dissolved H2, formation water	~0.9 g/m³ (STP-equivalent)	~150 °C / 400 bar reservoir target	[32]
FF4 (WA) headspace H2 from agitated groundwater	Up to 1,684–1,694 ppm	Shallow (~1 m) borehole groundwater, agitated headspace	[33]

Figures of merit

The universal figures of merit — sensitivity, LOD, selectivity, response time, drift, and the rest — are defined once on the Figures of merit page; this table specializes them to aqueous-phase H2 measurement and adds the extraction/partition-specific axes (stripping efficiency, Henry’s-law back-calculation, sampling-artifact susceptibility) that have no analogue in a gas-phase sensor family. Concentration units and the arithmetic converting a gas-phase reading to a dissolved one follow Concentration units and Gas-phase vs. aqueous-phase H2 sensing; values below are pulled from the Performance envelope and Environmental covariance & cross-sensitivity sections above.

Figures-of-merit table for this modality (11 rows)

Figure of merit	Symbol / formula	What it means here	Typical value for this technology	How to read it
Sensitivity	$S \equiv \partial(\text{signal})/\partial C_{aq}$; for a Clark-type cell, $i_L \propto C_{aq}$ (same diffusion-limited architecture as gas-phase amperometric cells)	How much detector output moves per unit dissolved H2	No vendor-quoted current-per-concentration slope (e.g. nA/µM) was located for the Unisense microsensor in sources reviewed — only LOD, range, and $t_{90}$ are published (unverified). For extraction-based methods (bubble-strip, headspace), “sensitivity” belongs to the gas-phase detector (TCD/RGD/MS) feeding on the stripped gas, not to a standalone aqueous transducer	Aqueous “sensitivity” is meaningful mainly for direct-insertion sensors; for every stripping/headspace method it collapses into the gas-phase detector’s own sensitivity plus the accuracy of the Henry’s-law back-calculation (see partition row below)
Limit of detection (dissolved H2)	Method-dependent; see Performance envelope table above	Smallest dissolved H2 concentration a given method can resolve	~0.1 nM, bubble-strip (Chapelle/Microseeps cell) + GC-RGD [13][14][15]; sub-nM reported feasible for diffusion/passive samplers but not independently re-verified [13][16][17]; 0.3 µM (300 nM), Unisense Clark-type electrochemical microsensor [18]; no independently validated H2-specific LOD for MIMS in groundwater — likely ppm-level gas-equivalent given vacuum-system H2 background [22][23][24][25]; ~1 ppm gas-phase resolution (much coarser once back-calculated to dissolved terms) for headspace-agitation + portable electrochemical meter [33][36]	A headline LOD means little in isolation for this modality — always compare it against the diagnostic ladder in the next row
Diagnostic resolution vs. the TEAP redox ladder (aqueous-specific)	n/a — interpretive threshold, not a formula	Whether a method’s LOD sits below or above the ~0.05–15 nM Lovley–Chapelle terminal-electron-accepting-process ladder that gives dissolved H2 its unique interpretive value (see Performance envelope above)	Bubble-strip + GC-RGD (~0.1 nM) clears the entire ladder [13][14][15]; the Unisense electrochemical microsensor (300 nM) sits entirely above it [10][11][12][18], useful instead for porewater/microbial-mat H2 work at hundreds-of-nM-to-µM concentrations (Yellowstone mats up to >300 nM) [39]	This figure of merit is what determines whether a method can do the job at all: an instrument can be “sensitive” by ordinary gas-sensor standards and still be unable to resolve the redox zonation dissolved-H2 measurement exists to reveal
Selectivity / cross-sensitivity	$K_{ij}=S_j/S_i$	How much a co-present species fools the H2 reading, once gas is extracted or the sensor is in direct contact with water	Unisense standard (non-X) tip is cross-sensitive to dissolved H2S, requiring the H2S-tolerant “H2-X” variant in sulfate-reducing/sulfidic settings [18][20]; MIMS inherits H2’s low relative sensitivity factor and m/z=1–2 water-vapor interference (see Mass spectrometry); RGD/RCP on stripped gas is near-specific to reducing gases (H2, CO) [13][14]	The worst-case interferent, H2S, is most likely to co-occur exactly in the sulfate-reducing TEAP regime (1–4 nM) where selectivity matters most
Response time ($t_{90}$), direct in-situ sensor	$t_{90}\sim l^2/D$ (membrane-diffusion time constant)	How fast the transducer itself catches up once in contact with water at a new concentration	Unisense: $t_{90}<15$ s (standard tips), $<0.3$ s (optional fast-response tip) [18][19]; bench/field MIMS: seconds to minutes, membrane-diffusion-limited [22][23][24][25]	Fast electrode/membrane kinetics are not the bottleneck for this modality — sampling and equilibration time (next row) dominates total time-to-result by one to three orders of magnitude
Stripping/equilibration efficiency & Henry’s-law partition (aqueous-specific)	$C_{aq}=k_H\,p_{H_2}$; dimensionless $H_{cc}=C_{gas}/C_{aq}\approx52$ at 25 °C [1]	The fraction of dissolved H2 that transfers to a gas phase (bubble or headspace) at equilibrium, and the Henry’s-law arithmetic used to back-calculate true $C_{aq}$ from a measured gas-phase reading — see Concentration units and Gas-phase vs. aqueous-phase H2 sensing for the general conversion	Microseeps cell reaches 95–100% of gas–liquid equilibrium in 10–15 min at >100 mL/min flow [14][15]; the original Chapelle cell needs ~2 h for full equilibrium (30 min operationally accepted) [13][14]. Every gas-phase back end (RGD, GE-MIMS’s equilibrator, headspace agitation) inherits this same conversion step	95–100% is a kinetic completeness statement, not a guarantee the $k_H$ used for back-calculation matches actual in-situ temperature/salinity — applying the 25 °C freshwater constant to hot, saline formation water biases the back-calculated $C_{aq}$ by 20–40%+ (see Mechanism & physics above)
In-situ vs. discrete/ex-situ	n/a	Whether the device reads water in place or requires bringing a sample (or extracted gas) to an analyzer	In-situ, continuous: Unisense Clark-type microsensor [18]; downhole (in-situ) sampler [13]; diffusion/passive samplers [16][17]; field-portable MIMS (miniRUEDI, CF-MIMS, GE-MIMS) [22][23][24]. Discrete/ex-situ: bubble-strip wellhead flow-through [13][14]; headspace-equilibration grab sample [33][36]; bench MIMS [25]; RGD/RCP as a surface analytical step [27]	“In-situ” does not imply adequate sensitivity — the one continuous, truly in-situ sensor reviewed (Unisense) is also the one whose LOD sits above the diagnostic TEAP range
Sampling-artifact susceptibility (aqueous-specific)	n/a	How much a method’s own extraction/handling process can move the reported number, independent of instrument precision	Degassing on ascent (open pumped/bailed sampling) under-measures dissolved gas relative to closed methods [29][30]; atmospheric ingress adds a ~0.4 nM-equivalent contribution per contamination event — comparable to the entire lower TEAP ladder [1][2]; well-casing corrosion (iron/steel) generates persistent, non-decaying spurious H2, mandating PVC or other non-metallic casing [13]; microbial consumption continues in the sample bottle post-collection unless arrested, with rising pH a documented proxy for H2 loss in progress [41]	This axis, not any headline LOD or sensitivity spec, is what actually decides whether a dissolved-H2 number is trustworthy — see Environmental covariance & cross-sensitivity immediately below for the full quantitative treatment
Dynamic / linear range	$N_{\text{decades}}=\log_{10}(C_{max}/C_{min})$	Span over which one method or sensor produces a usable reading	Unisense: 0–2,000 µM, roughly 3.8 decades above its own 0.3 µM LOD [18]; bubble-strip + GC-RGD: from the ~0.1 nM TEAP floor up to Yellowstone-mat concentrations >300 nM (>3 decades) before a different analytical strategy is typically warranted [10][11][12][39]	The two technologies’ useful ranges barely overlap — Unisense picks up roughly where bubble-strip’s practical ceiling ends
Drift & stability	$R_0(t)=R_0(0)+d_Z t$; $S(t)=S(0)(1+d_S t)$	Baseline/slope creep and consumable depletion over deployment time	Unisense: stirring sensitivity <2%; 2–3%/°C temperature coefficient requiring active compensation [18]; RGD/RCP: HgO-bed consumable depletion and drift [27]; MIMS: no H2-specific drift figures published for any variant reviewed [22][23][24]	A seasonal or depth-driven temperature swing shifts both the true equilibrium concentration (Henry’s law) and the Unisense sensor’s own transduction in the same measurement window — a confound, not two independent error sources
Operating envelope	$C_{\text{reported}}=C_{\text{raw}}\cdot f_T(T)\cdot f_P(P)$	The temperature/pressure domain over which stated LOD/accuracy actually hold	Unisense: 0–60 °C, pressure-tolerant to 20 bar [18]; GE-MIMS explicitly rated to 100 bar for deep/pressurized fluids [24]; no commercial downhole dissolved-H2 sensor validated at true reservoir T/P — the one deep-formation case located (Colombia MDT) is a bespoke fluid-sampling operation, not a sensor deployment [32]	Every continuous sensor reviewed here is a near-surface instrument; deep formation-water dissolved-H2 work today means bringing a pressure-preserved sample to surface, not deploying a downhole sensor (see Use for geologic H2 below)

What actually decides usability

What dominates. For dissolved-H2 work, the traditional sensitivity–response-time–power triangle is not what decides usability — response is fast (seconds, for any direct-insertion sensor) and power is a non-issue.

What actually decides whether a number is usable is, first, whether the achieved LOD clears the ~0.05–15 nM TEAP ladder at all (only bubble-strip + GC-RGD does, cleanly), and second, sampling-artifact susceptibility, which can move the reported concentration by one to three orders of magnitude — larger than the ~10× steps separating adjacent redox regimes — through degassing, atmospheric contamination, well-casing corrosion, or uncontrolled microbial consumption before the sample ever reaches an instrument.

A field program should budget for method discipline (closed sampling, minimal holding time, non-metallic casing) before it budgets for a lower-LOD instrument; a lower-LOD instrument cannot rescue a compromised sample.

Environmental covariance & cross-sensitivity

On the evidence reviewed, the sampling event itself, not instrument noise, is the dominant source of error in dissolved-H2 work.

Degassing on ascent/decompression. Any pumped or bailed sample loses dissolved gas as it rises through a pressure gradient toward atmospheric pressure at the wellhead, and H2’s low solubility and high volatility ($H_{cc}\approx52$) make it more prone to this loss than more soluble co-analytes like CO2. This is precisely the failure mode pressure-preserving downhole/wireline sampling is built to avoid, and the reason coal-seam-gas hydrogeology has developed dedicated “sample at in-situ pressure” protocols [30]. The general methane-in-groundwater sampling literature — directly transferable to H2 — has separately shown that “open” sampling methods systematically under-measure dissolved gas relative to fully closed methods (e.g., an evacuated flexible-container system filled via an in-line manifold that never exposes the sample to atmosphere) [29].

Atmospheric contamination. The inverse artifact: any air introduced to a sample or stripping cell — a poorly sealed pump fitting, an incompletely purged cell, a bailer breaking the water-table meniscus — adds atmospheric H2 (0.53 ppm, equivalent to a ~0.4 nM aqueous contribution at equilibrium) directly on top of the true in-situ concentration. Because the entire lower half of the TEAP ladder (nitrate/Mn/Fe(III) reduction, <0.05–0.8 nM) sits within roughly one order of magnitude of the atmospheric-equilibrium floor itself, even a small air-contamination artifact can erase the distinction between “actively denitrifying” and “iron-reducing” water.

Microbial consumption during sample handling. Because the whole redox-indicator framework depends on H2-oxidizing organisms being present and metabolically active, the same organisms that create the useful in-situ signal will keep consuming H2 in the sample bottle unless deliberately arrested (cooling, biocide, rapid analysis).

Recent standardization work on microbial H2-consumption testing found that pH increase is a reliable field/lab indicator of ongoing microbial H2 consumption, and the magnitude of pH increase correlates positively with the percentage of H2 lost [41] — a practical, low-cost cross-check Hestia could build into a sampling SOP (measure pH at collection and again at analysis; a rising pH trend flags a reading that should be treated as a lower bound).

Samples should be kept at or below in-situ temperature and analyzed with minimal holding time; standard dissolved-gas handling guidance (developed for methane, directly applicable) recommends storing sample bottles cold, inverted or on their side, to keep any exsolved bubble away from the septum and limit both degassing and continued microbial activity before analysis.

Well-casing artifacts

Well-casing artifacts — a groundwater-specific failure mode with no soil-gas analog. Chapelle, Vroblesky, Woodward, and Lovley’s foundational 1997 methods paper found, at two independently studied field sites, that iron or steel well casings actively generate H2 through anaerobic corrosion, “masking” the true in-situ dissolved H2 concentration, and that DC submersible pumps independently generate spurious H2 as well [13].

Their recommendation — PVC-cased wells, or wells cased with other non-H2-generating materials, are a requirement for credible dissolved-H2 groundwater measurement [13] — is a design constraint Hestia should apply retroactively to any candidate monitoring-well network: an existing steel-cased water-supply or oil-and-gas monitoring well is not usable for dissolved-H2 redox work without first establishing a non-metallic casing interval at the sampled depth.

This is mechanistically related to, but distinct from, the “drilling-induced hydrogen” artifact documented for mud-gas logging and soil-gas surveys on Geologic context — both are cases of well/hole construction generating H2 the subsequent measurement cannot distinguish from a genuine signal, but the drill-bit mechanism is a one-time, decaying artifact, whereas anaerobic steel-casing corrosion is a persistent, non-decaying H2 source for the life of the well.

Cross-sensitivity in the analytical back end. Once gas is extracted into a headspace or stripping bubble, all the interference chemistry documented in the sibling GC and MS pages applies unchanged:

• RGD/RCP detectors are essentially blind to the bulk matrix (N2, O2, Ar, CO2, CH4) but are consumed/degraded by cumulative reducing-gas load on their HgO bed.
• MIMS systems inherit H2’s notoriously low relative sensitivity factor (~0.46 vs. N2, meaning an uncorrected instrument under-reads true H2 partial pressure by roughly a factor of 2) and the fact that water vapor is one of the two gases virtually every vacuum system has in detectable amounts, directly interfering at m/z = 1–2 (see Mass spectrometry).
• The Unisense microsensor’s standard (non-X) tip is cross-sensitive to dissolved H2S, requiring the H2S-tolerant “H2-X” variant in sulfate-reducing or sulfidic settings [18][20] — a sensor-level analog of the H2S poisoning issues documented for gas-phase catalytic and electrochemical H2 sensors (see Catalytic pellistors, Electrochemical cells).

Temperature and salinity as measurement confounds, not just target-analyte physics. Beyond their role in setting the true equilibrium concentration, temperature directly affects the Unisense microsensor’s own transduction (2–3%/°C sensitivity coefficient requiring active compensation [18]) and affects bubble-strip equilibration kinetics (equilibration studies were explicitly run at both 4 °C and 21 °C because cold groundwater equilibrates more slowly [14]) — a seasonal or depth-driven temperature swing at a monitoring site can shift both the “true” number and the instrument’s read of it, echoing the general covariance framework on Environmental covariance & cross-sensitivity.

Use for geologic H2

Field assessment

Dissolved H2 in produced water and monitoring wells is an under-used, high-information-content exploration signal for one question no gas-phase measurement can answer directly: what is the microbial redox context here, and is it consuming or transmitting geologic H2? It is also one of the most methodologically fragile techniques in this compendium — whether a dataset is scientifically actionable or just a sampling artifact depends more on handling time and casing-material choice than on instrument specifications. Treat any dissolved-H2 dataset collected without an explicit, documented chain-of-custody against degassing, atmospheric ingress, microbial consumption, and well-casing generation as qualitative at best.

Where dissolved H2 adds value. No other measurement in Hestia’s toolkit reads the microbial redox context of a candidate site the way dissolved H2 does. A soil-gas or free-gas H2 anomaly (see Soil-gas surveys) tells you gas is present; a dissolved-H2 TEAP profile in the same water tells you whether the local subsurface biology is actively consuming that gas, and roughly how fast given the electron acceptors present, or whether conditions are permissive for H2 to migrate essentially unconsumed.

This is directly actionable for site screening: a water column stuck at sub-nanomolar, nitrate/Fe(III)-reduction-level H2 despite a known nearby source suggests a robust microbial sink is intercepting a meaningful fraction of the upward flux — a genuine negative for surface-detectability, independent of source strength — whereas a water column sitting at methanogenic-level H2 (5–15+ nM) alongside elevated dissolved CH4 suggests either a weak local sink or co-generated biogenic methane complicating source attribution (see Isotopes & fingerprinting for the companion isotopic discrimination tools).

Dissolved H2 is also, uniquely among the modalities in this wiki, directly measurable in produced water from existing oil-and-gas wells without any new drilling — a low-marginal-cost dataset Hestia should be actively requesting from operators in candidate basins, provided the well-casing caveat above is satisfied.

Where it falls short.

• Sampling-artifact dominance. Every artifact above — degassing, atmospheric ingress, microbial consumption in the bottle, well-casing corrosion — operates on the same order of magnitude as, or larger than, the diagnostic differences between adjacent rungs of the TEAP ladder. A dissolved-H2 number collected without rigorous method discipline (validated non-metallic well casing, closed sampling, minimal holding time, cold storage, ideally in-situ pressure preservation for deep formation water) is not just noisy, it can be meaningless relative to the ~10× concentration steps separating one redox regime from the next.
• No cheap, sensitive, continuous field instrument exists for the diagnostic (sub-15 nM) range. The one continuous, truly in-situ, field-rugged sensor family reviewed (Unisense Clark-type microsensors) has a detection limit (300 nM) sitting entirely above the TEAP ladder; the one method proven to 0.1 nM (bubble-strip + GC-RGD) is a discrete, labor-intensive, minutes-to-hours-per-sample laboratory-grade workflow, not a continuous monitoring-well instrument.
• MIMS has essentially no published H2-specific groundwater validation. Every MIMS deployment reviewed (miniRUEDI, CF-MIMS, GE-MIMS) validates precision for He, Ar, N2, O2, Kr; H2 appears in vendor gas-suite lists but is not independently calibrated or field-tested for dissolved groundwater work in any source located this session [22][23][24]. This is a closeable gap — a dedicated H2-spiked-water calibration campaign against a known-concentration standard, commissioned directly with Hiden or the Gasometrix/Eawag group — but it is not closed today.
• Downhole, high-T/P deployment of any of these techniques is essentially unproven. The one located example of pressure-preserved deep-formation dissolved-H2 sampling (the Colombian MDT wireline case [32]) is a bespoke, single-well petroleum-industry-style operation, not a repeatable, off-the-shelf monitoring-well tool; no commercial downhole dissolved-H2 sensor (as opposed to a fluid-sampling tool that brings a sample to surface) was identified.

Method discipline — non-metallic well construction, closed or pressure-preserved sampling, minimal holding time, and ideally paired pH and major-ion redox context — is the core of this technique, not an optional refinement. See Calibration & QA/QC for how to write this discipline into a field SOP, and Helium & co-tracers for the parallel co-tracer discipline (He, noble gases) that strengthens source attribution once a dissolved-H2 dataset is trusted.

Vendors & products

The table compiles vendors, labs, and standards actually named in the dissolved-H2 groundwater literature reviewed. Price class is generally not disclosed in vendor materials reviewed and should be obtained directly via RFI/RFQ.

Vendor/lab	Product/service	Category	Notes	Ref
Unisense A/S	H2-10/25/50/100/500, H2-N, H2-NP (and H2-X sulfide-tolerant variants)	Clark-type amperometric microsensor	Tip sizes 8 µm–600 µm plus needle formats; 0.3 µM LOD; 0–2,000 µM range; t90 <15 s (<0.3 s optional); 0–60 °C; pressure-tolerant to 20 bar; polarized at +100 mV	[18][19][20]
Hiden Analytical	HPR-40 DSA / DEMS	Bench membrane-inlet mass spectrometer (MIMS)	PDMS-membrane probe/flow-through/DEMS-cell inlets for dissolved-species analysis; mass range 1–200/300 amu; ~33 kg + 7 kg pump	[25]
Gasometrix (Eawag spin-off)	miniRUEDI	Field-portable MIMS	Base gas suite He/Ne/Ar/Kr/N2/O2/CO2; extended methods for CH4, H2S, H2 per vendor; 60×40×14 cm, 13 kg, 50 W, 1–3% uncertainty	[22][23]
Chatton et al. / Brennwald et al. (Eawag/ETH)	CF-MIMS (continuous-flow) / GE-MIMS (gas-equilibrium, to 100 bar)	Research-grade field/lab MIMS variants	CF-MIMS: <1% precision, up to 1 measurement/2 s; GE-MIMS: 1–5% RSD, explicitly rated for deep/high-pressure geological-fluid systems; H2 listed in gas suite but not independently validated	[23][24]
AMETEK Process Instruments (Trace Analytical lineage)	ta3000R / ta5000R Reduction Gas Analyzer	Process/lab GC-RGD	Factory-quoted LDL 25 ppb H2 (bulk gas) / 50 ppb dissolved H2 in water headspace; ±10 ppb or ±10% accuracy	[27]
Peak Laboratories	Peak Performer 1 (RCP-based)	Lab/field reducing-compound photometer	Manufacturer markets ppb-level H2/CO detection; specific dissolved-water-headspace LOD figures not independently retrievable this session	[28]
SRI Instruments	8610C/D-series with RGD add-on	Field-portable GC	Demonstrated 500 ppb ambient H2 on unoptimized bench setup; suitable back end for bubble-strip/headspace work	[26]
Microseeps Inc. (Pittsburgh, PA)	Bubble-stripping gas-sampling cell (~70 mL)	Sampling hardware	Field-optimized commercial successor to the original glass “Chapelle cell” (~230 mL); faster equilibration, lower cost	[14][15]
IsoTech Laboratories	DG-1/DG-2/DG-3 dissolved-gas water analysis packages, IsoFlask® closed-sampling container	Commercial dissolved-gas lab service	Explicitly covers methane/ethane/propane concentration and isotopes; hydrogen is not listed among IsoTech’s standard dissolved-gas analytes as of the service page reviewed — a notable market gap for a well-established environmental dissolved-gas lab	[35]
Industrial Scientific	MX6 iBrid multi-gas monitor (H2 electrochemical smart sensor)	Portable multi-gas field meter	0–2,000 ppm H2, 1 ppm resolution; used as the analytical back end for the FF4 agitated-groundwater headspace method	[33][36]
U.S. EPA	RSKSOP-175, “Sample Preparation and Calculations for Dissolved Gas Analysis in Water Samples”	Standard operating procedure	Unofficial-but-widely-used framework for headspace GC dissolved-gas analysis, reported to cover dissolved H2 alongside CH4, C2–C4 hydrocarbons, N2, N2O, O2; see Standards & regulations	[40]
USGS	Foundational method literature (McInnes & Kampbell 2000 [14]; Chapelle, Vroblesky, Woodward & Lovley 1997 [13]; Vroblesky diffusion-syringe sampler [16])	Foundational method literature/standards precursor	Not a formal ASTM/ISO standard, but the de facto reference basis for essentially all subsequent bubble-strip and diffusion-sampler H2 work	[13][14][16]

Open questions

The foundational science here — H2 as a thermodynamically governed redox indicator — was essentially settled by the late 1990s and has not required major theoretical revision since; the natural-hydrogen exploration community is now actively re-purposing that 1990s methods toolkit for a new question, exemplified by the Western Australian FF4 “agitated groundwater headspace” work [33][34] and the first published Colombian downhole natural-hydrogen sampling case [32].

Even a field as well characterized as Bourakebougou (Mali), whose shallow karstic reservoir produces ~98% pure H2 free gas, was characterized in part through sequential spring/borehole water sampling for gas geochemistry, underscoring that dissolved-phase sampling is a standard companion technique even at the highest-purity documented sites [38].

A related, still-open question runs through the Midcontinent Rift play in Kansas/Nebraska, where H2-bearing gas has been recovered from Precambrian basement wells for decades but where the balance between deep (mantle/serpentinization) and shallow (near-surface microbial or radiolytic) origins — and, by extension, the degree to which dissolved-phase transport through overlying groundwater has redistributed or masked the original source signature — remains an active geochemical debate rather than a settled question [37]. Specific open items for Hestia:

1. No source reviewed provides a validated, H2-specific field detection limit for any MIMS instrument applied to groundwater. Before proposing MIMS (miniRUEDI, CF-MIMS, or GE-MIMS) for continuous dissolved-H2 monitoring at a Hestia site, commission an explicit H2-spiked-water calibration curve directly with Hiden Analytical or the Eawag/Gasometrix group — do not extrapolate from the published He/Ar/N2/O2/Kr precision figures.
2. The Setschenow coefficient for H2 specifically was located only as an aggregate “~15% per mol/kg NaCl” empirical statement [4][6], not a precise, temperature-resolved numerical table. For quantitative dissolved-H2 mass-balance work at a specific candidate brine composition and reservoir temperature, obtain or re-derive the full Chabab et al. (2020) parameter set rather than relying on this rounded figure.
3. No commercial optical (fluorescence/absorbance) dissolved-H2 sensor analogous to mature dissolved-O2 optodes was identified. Determine whether this is a genuine technology gap — worth an internal R&D bet, potentially building on the Pd-film fiber-optic gas-phase H2 sensor lineage on Fiber optic sensors and Exotic optical — or simply an under-searched corner of the commercial market.
4. Quantify the practical field detection limit of the “agitated groundwater headspace” method [33] in dissolved-concentration (nM) terms, not just headspace ppm. The FF4 methodology reports headspace ppm directly from a portable electrochemical meter; back-calculating a defensible dissolved-phase LOD requires the exact water/headspace volume ratio and equilibration completeness achieved, which were not extracted in this session’s review.
5. Test whether the pH-based proxy for microbial H2 consumption during sample holding [41] — developed in an underground-hydrogen-storage brine context — transfers quantitatively to shallow, lower-salinity groundwater sampling conditions. If it does, it would be a near-zero-cost field QA/QC addition to any Hestia dissolved-H2 sampling SOP.
6. Directly test the well-casing-material requirement [13] against Hestia’s own candidate monitoring wells or any existing produced-water wells being considered for dissolved-H2 sampling, since this single artifact can invalidate an entire dataset and is fully avoidable with upfront well-design diligence.
7. Scope a bespoke downhole/wireline dissolved-H2 sampling campaign (MDT-class or U-tube-style pressure-preserving tool, per the Colombian precedent [32]) for any deep formation-water target where free-gas H2 has not yet been directly encountered but is suspected to be near its bubble point — comparatively low-technology-risk, high-diagnostic-value relative to the harder open problems above.
8. Investigate whether the aerobic soil HA-HOB sink and the anaerobic groundwater TEAP framework are populated by overlapping or distinct microbial communities, and whether a single integrated microbial-sink model spanning the vadose zone through the saturated zone could improve joint interpretation of soil-gas and dissolved-H2 datasets at the same site.

Related pages

• Why H2 is hard — the selection-rule and background-concentration framing this page’s nanomolar arithmetic builds on
• Reading the specs — how to interpret an LOD/precision spec sheet, directly relevant to comparing the method table above
• Geologic context — the HA-HOB soil sink and drilling-artifact backdrop this page’s anaerobic groundwater analog parallels
• Soil-gas surveys — the complementary shallow, gas-phase technique dissolved H2 is often deployed alongside or in place of
• Environmental covariance & cross-sensitivity — the cross-modality treatment of the temperature/salinity confounds discussed here
• Downhole deployment — high-T/P instrumentation context for deep formation-water sampling
• Calibration & QA/QC — the discipline that must govern any dissolved-H2 sampling program
• Helium & co-tracers — companion noble-gas tracers for strengthening source attribution once a dissolved-H2 dataset is trusted
• Isotopes & fingerprinting — δD and other isotopic tools for distinguishing microbial, thermogenic, and deep-abiotic H2 sources
• Gas chromatography — the GC-RGD/RCP/PDHID back end used on bubble-strip and headspace-equilibration samples
• Mass spectrometry — the MIMS variant family and the m/z=2 background problem it inherits
• Mud-gas logging — the drilling-artifact-generation parallel (drill-bit metamorphism vs. steel-casing corrosion)
• Electrochemical cells — the transduction-physics family behind the Unisense Clark-type microsensor
• Standards & regulations — EPA RSKSOP-175 and related dissolved-gas sampling frameworks
• Vendors and Glossary — consolidated vendor and terminology reference

References

Full reference list (0 entries)

[1] NIST WebBook, “Hydrogen” — Henry’s law constant data (k°H ≈ 0.00078 mol kg⁻¹ bar⁻¹; alternate literature value 8.13×10⁻⁴ M/atm; 1282 L·atm/mol reciprocal form). https://webbook.nist.gov/cgi/cbook.cgi?ID=C1333740&Mask=10

[2] Molecular Hydrogen Institute, “Concentration and Solubility of H2” (saturation ≈1.6 mg/L at 1 atm/25 °C; conventional water background ≈8.65×10⁻⁷ mg/L). https://molecularhydrogeninstitute.org/concentration-and-solubility-of-h2/

[3] Crozier, T.E., Yamamoto, S., “Solubility of hydrogen in water, sea water, and sodium chloride solutions,” Journal of Chemical & Engineering Data 19(3):242–244 (1974). https://pubs.acs.org/doi/10.1021/je60062a007

[4] Chabab, S. et al., “Measurements and predictive models of high-pressure H2 solubility in brine (H2O+NaCl) for underground hydrogen storage application,” International Journal of Hydrogen Energy 45(56) (2020). https://www.sciencedirect.com/science/article/abs/pii/S0360319920332195

[5] “Solubility of hydrogen in sodium chloride brine at high pressures,” Fluid Phase Equilibria (2021). https://www.sciencedirect.com/science/article/abs/pii/S037838122100087X

[6] “Solubility of H2 in water and NaCl brine under subsurface storage conditions: Measurements and thermodynamic modeling,” International Journal of Hydrogen Energy (2023–2024). https://www.sciencedirect.com/science/article/abs/pii/S0360319923054976

[7] “An Accurate Model for Estimating H2 Solubility in Pure Water and Aqueous NaCl Solutions,” Energies 15(14):5021 (2022). https://www.mdpi.com/1996-1073/15/14/5021

[8] “Transport Properties of NaCl in Aqueous Solution and Hydrogen Solubility in Brine,” Journal of Physical Chemistry B (2023). https://pmc.ncbi.nlm.nih.gov/articles/PMC10591480/

[9] “Effects of temperature and pressure on Henry’s law constant for hydrogen in the primary water of a simulated pressurized-water reactor,” Progress in Nuclear Energy (2017). https://www.sciencedirect.com/science/article/abs/pii/S0306454917300853

[10] Lovley, D.R., Goodwin, S., “Hydrogen concentrations as an indicator of the predominant terminal electron-accepting reactions in aquatic sediments,” Geochimica et Cosmochimica Acta 52(12):2993–3003 (1988). https://pubs.usgs.gov/publication/70014288; https://www.sciencedirect.com/science/article/abs/pii/0016703788901639

[11] Lovley, D.R., Chapelle, F.H., Woodward, J.C., “Use of Dissolved H2 Concentrations To Determine Distribution of Microbially Catalyzed Redox Reactions in Anoxic Groundwater,” Environmental Science & Technology 28(7):1205–1210 (1994). https://pubs.acs.org/doi/abs/10.1021/es00056a005

[12] Chapelle, F.H. et al., “Deducing the Distribution of Terminal Electron-Accepting Processes in Hydrologically Diverse Ground Water Systems,” Water Resources Research 31(2):359–371 (1995). https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/94WR02525

[13] Chapelle, F.H., Vroblesky, D.A., Woodward, J.C., Lovley, D.R., “Practical Considerations for Measuring Hydrogen Concentrations in Groundwater,” Environmental Science & Technology 31(10):2873–2877 (1997). https://pubs.acs.org/doi/abs/10.1021/es970085c; https://www.usgs.gov/publications/practical-considerations-measuring-hydrogen-concentrations-groundwater

[14] McInnes, D., Kampbell, D.H., “The Bubble Stripping Method for Determining Dissolved Hydrogen (H2) in Well Water,” Field Analytical Chemistry and Technology 4(6):283–296 (2000), EPA/600/J-01/088. https://cfpub.epa.gov/si/si_public_record_report.cfm?LAB=NRMRL&dirEntryID=64980

[15] Alter, M.D., Steiof, M., “Optimized method for dissolved hydrogen sampling in groundwater,” Journal of Contaminant Hydrology 78(1–2):71–86 (2005). https://www.sciencedirect.com/science/article/abs/pii/S0169772205000483; https://pubmed.ncbi.nlm.nih.gov/15949607/

[16] Vroblesky, D.A., “A simple pore water hydrogen diffusion syringe sampler,” U.S. Geological Survey. https://www.usgs.gov/publications/a-simple-pore-water-hydrogen-diffusion-syringe-sampler; https://www.ncbi.nlm.nih.gov/pubmed/17973759

[17] Spalding, B.P., Watson, D.B., “Measurement of dissolved H2, O2, and CO2 in groundwater using passive samplers for gas chromatographic analyses,” Environmental Science & Technology 40(24):7861–7867 (2006). https://pubmed.ncbi.nlm.nih.gov/17256539/

[18] Unisense A/S, “H2 Sensor for Hydrogen Research” (product specifications). https://www.unisense.com/H2/

[19] Unisense A/S, “Hydrogen Sensor Measurement Principles and Applications in Research.” https://unisense.com/hydrogen-sensor-measurement/

[20] Unisense A/S, “New and improved H2 microsensor.” https://unisense.com/new-and-improved-h2-microsensor/

[21] Singleton, M., Hudson, G., “Membrane Inlet Mass Spectrometry for Measuring Dissolved Gases,” Lawrence Livermore National Laboratory Technical Report, OSTI ID 877829 (2005). https://doi.org/10.2172/877829

[22] “Membrane Inlet Mass Spectrometer for the Quasi-Continuous On-Site Analysis of Dissolved Gases in Groundwater,” Environmental Science & Technology 46(15):8288 (2012). https://pubs.acs.org/doi/10.1021/es3004409

[23] Chatton, E., Labasque, T., de La Bernardie, J., Guihéneuf, N., Bour, O., Aquilina, L., “Field Continuous Measurement of Dissolved Gases with a CF-MIMS: Applications to the Physics and Biogeochemistry of Groundwater Flow,” Environmental Science & Technology 51(2):846–854 (2017). https://pubs.acs.org/doi/10.1021/acs.est.6b03706

[24] Brennwald, M.S., Rinaldi, A.P., Gisiger, J., Zappone, A., Kipfer, R., “Gas equilibrium membrane inlet mass spectrometry (GE-MIMS) for water at high pressure,” Geoscientific Instrumentation, Methods and Data Systems 13:1–16 (2024). https://gi.copernicus.org/articles/13/1/2024/

[25] Hiden Analytical, “HPR-40 DSA: Membrane Inlet Mass Spectrometers for Dissolved Species Analysis,” Brochure TDS 171/3. https://hideninc.com/wp-content/uploads/2018/10/TDS-171-3_HPR-40-DSA-Brochure-6pp-High-Res.pdf

[26] SRI Instruments, “Operating the SRI RGD (Reduction Gas Detector),” operation manual. https://srigc.com/public/storage/downloads/RGDOperation.pdf

[27] AMETEK Process Instruments (Trace Analytical lineage), “ta3000R Reduction Gas Analyzer — Specifications.” https://www.asap.nl/wp-content/uploads/2012/07/ametek-ta3000r.pdf

[28] Peak Laboratories, “Peak Performer 1 RCP (Reducing Compound Photometer)” product datasheet, RCP 910-150. https://peaklaboratories.com/wp-content/uploads/2022/09/RCP-910-150-CO-in-Hydrogen.pdf (original product-page URL dead; replaced with live datasheet for the same product)

[29] Molofsky, L.J. et al., “Effect of Different Sampling Methodologies on Measured Methane Concentrations in Groundwater Samples,” Groundwater 54(5):669–680 (2016). https://ngwa.onlinelibrary.wiley.com/doi/10.1111/gwat.12415

[30] Banks, E.W., Smith, J.M.C., Hatch, M., Burk, D., Suckow, A., “Sampling Dissolved Gases in Groundwater at in Situ Pressure: A Simple Method for Reducing Uncertainty in Hydrogeological Studies of Coal Seam Gas Exploration,” Environmental Science & Technology Letters 4(12):535–539 (2017). https://pubs.acs.org/doi/10.1021/acs.estlett.7b00457

[31] Milkov, A.V., “Molecular hydrogen in surface and subsurface natural gases: Abundance, origins and ideas for deliberate exploration,” Earth-Science Reviews 230:104063 (2022). https://www.sciencedirect.com/science/article/abs/pii/S0012825222001477

[32] Patiño, C., Strąpoć, D., Torres, O., Mullins, O., Bustos, U., Bermudez, O., López, A., Trujillo, M., Morales, H., “First downhole sampling of a natural hydrogen reservoir in Colombia,” Frontiers in Energy Research 12:1443269 (2024). https://www.frontiersin.org/journals/energy-research/articles/10.3389/fenrg.2024.1443269/full

[33] Davies, K. et al., “Advancing Natural Hydrogen Exploration: Headspace Gas Analysis in Water-Logged Environments,” Energy & Fuels (2024). https://pubs.acs.org/doi/10.1021/acs.energyfuels.3c04562

[34] “A natural hydrogen seep in Western Australia: Observed characteristics and controls,” Science and Technology for Energy Transition (STET), article stet20240139 (2024). https://www.stet-review.org/articles/stet/full_html/2024/01/stet20240139/stet20240139.html

[35] IsoTech Laboratories, “Water Analyses” service page (DG-1/DG-2/DG-3 dissolved-gas packages). https://isotechlabs.com/services/water

[36] Industrial Scientific, “MX6 iBrid 6-Gas Monitor” product/specification pages. https://www.indsci.com/en/gas-detectors/multi/mx6-ibrid

[37] Guélard, J. et al., “Natural H2 in Kansas: Deep or shallow origin?” Geochemistry, Geophysics, Geosystems 18(5) (2017). https://agupubs.onlinelibrary.wiley.com/doi/10.1002/2016GC006544

[38] “Characterization of the spontaneously recharging natural hydrogen reservoirs of Bourakebougou in Mali,” Scientific Reports 13 (2023). https://www.nature.com/articles/s41598-023-38977-y; https://pmc.ncbi.nlm.nih.gov/articles/PMC10363119/

[39] Spear, J.R., Walker, J.J., McCollom, T.M., Pace, N.R., “Hydrogen and bioenergetics in the Yellowstone geothermal ecosystem,” Proceedings of the National Academy of Sciences 102(7):2555–2560 (2005). https://www.pnas.org/doi/10.1073/pnas.0409574102

[40] U.S. EPA, RSKSOP-175, “Sample Preparation and Calculations for Dissolved Gas Analysis in Water Samples” (GC headspace-equilibration technique). https://archive.epa.gov/region1/info/testmethods/web/pdf/rsksop175v2.pdf

[41] “Towards standardized microbial hydrogen consumption testing in the subsurface: harmonized field sampling and enrichment approaches,” World Journal of Microbiology and Biotechnology (2025). https://pmc.ncbi.nlm.nih.gov/articles/PMC12464130/

[42] Hoehler, T.M., “Thermodynamic control on hydrogen concentrations in anoxic sediments,” Geochimica et Cosmochimica Acta 62(10):1745–1756 (1998). https://www.sciencedirect.com/science/article/abs/pii/S0016703798001069

[43] Hoehler, T.M. et al., “Apparent minimum free energy requirements for methanogenic Archaea and sulfate-reducing bacteria in an anoxic marine sediment,” FEMS Microbiology Ecology 38(1):33–41 (2001). https://academic.oup.com/femsec/article/38/1/33/538599