Isotopic & genetic fingerprinting

Summary

Every other page in this wiki is built to answer one question: “how much H₂ is here?” This page answers a different question, one that matters more for exploration economics: where did this H₂ come from? A 300 ppm H₂ reading in soil gas is equally compatible with a genuine deep geologic seep, a shallow microbial fermentation signal, a drill-bit-metamorphism artifact, or a leak from a nearby industrial gas line — and no field concentration sensor, however sensitive, can tell those apart (see Why H₂ Is Hard and Geologic Context). Isotopes can, at least partially.

Every natural or industrial process that makes H₂ — water splitting on hot ultramafic rock, radioactive decay of uranium and thorium in old crustal rock, bacteria fermenting organic matter, an electrolyzer splitting water with grid power — imprints a slightly different ratio of heavy hydrogen (deuterium, D) to ordinary hydrogen (protium, H) into the gas molecule, because heavier atoms vibrate more slowly and react more slowly.

Measure that ratio precisely, plus the equivalent signature carried by companion gases (helium, methane), send the sample to a specialist laboratory, wait days to weeks, and the result is strong evidence about a gas’s origin and, in favorable cases, the temperature at which it formed at depth. This is a laboratory technique fed by a physical sample, not a field or real-time tool; it is neither cheap nor fast; and a single isotope number is rarely conclusive by itself. It takes a bundle of two or three isotope systems, read together against geologic context, to make a defensible case, which is the workflow this page describes.

Mechanism & physics

Why isotopes discriminate origin when concentration cannot

Hydrogen has two stable isotopes: protium (¹H, ~99.9885 atom %) and deuterium (²H or D, ~0.0115 atom %). A third isotope, tritium (³H, half-life 12.32 yr), is radioactive and separately diagnostic (see 3H/3He below). Because absolute D/H ratios are hard to measure with useful precision, isotope geochemistry universally reports compositions as a per-mil (‰) deviation from a reference standard, Vienna Standard Mean Ocean Water (VSMOW):

$\delta D \equiv \delta^2H = \left(\frac{R_{sample}}{R_{VSMOW}} - 1\right)\times 1000\ (\permil), \qquad R = \frac{[\mathrm{D}]}{[\mathrm{H}]}$

A more negative δD means the sample is depleted in deuterium relative to ocean water; a less negative or positive δD means it is enriched. The underlying continuous-flow/dual-inlet IRMS measurement physics — H₂⁺ at m/z 2 vs. HD⁺ at m/z 3, the H₃⁺ interference problem, TC/EA pyrolysis — is developed in full in Mass Spectrometry §1.4 and is not repeated here; this page picks up at the interpretation of the resulting number.

Isotope fractionation between two coexisting phases or species A and B is quantified by the fractionation factor

$\alpha_{A-B} = \frac{R_A}{R_B} \approx 1 + \frac{\delta_A - \delta_B}{1000}, \qquad 1000\ln\alpha_{A-B} \approx \delta_A - \delta_B \ \text{(for small differences)}$

Two physically distinct mechanisms produce non-zero fractionation, and distinguishing them underlies everything downstream on this page:

Equilibrium (thermodynamic) isotope effects arise because isotopic substitution changes a molecule’s vibrational zero-point energy without changing its electronic structure — a heavier isotope (D) sits in a stiffer, lower-frequency vibrational well, so C–D, O–D, and H–D bonds have lower zero-point energy than their light-isotope counterparts. The formal treatment is the Bigeleisen–Mayer reduced partition function ratio (RPFR),

$\ln\beta = \sum_i\left[\frac{u_i^{*}}{2}-\frac{u_i}{2}+\ln\frac{1-e^{-u_i}}{1-e^{-u_i^{*}}}\right],\qquad u_i = \frac{hc\tilde\nu_i}{k_BT}$

(sum over vibrational modes $i$; starred quantities are the heavy-isotope frequencies), which reduces at high temperature to $1000\ln\alpha \propto 1/T^2$ — the textbook reason essentially every isotope geothermometer (H₂–H₂O, CH₄–H₂O, carbonate clumped-isotope, etc.) is fit to an empirical $A/T^2 + B/T + C$ form, and why fractionation magnitudes shrink toward zero as temperature rises [9].

Kinetic isotope effects (KIE) arise instead from a rate difference: the lighter isotopologue typically reacts faster because it faces a smaller activation-energy barrier, so $k_H/k_D > 1$ for most bond-breaking reactions. KIEs dominate irreversible, one-way processes — biological H₂ production/consumption, water electrolysis, water radiolysis — and typically produce much larger fractionations than equilibrium effects at low temperature.

Kinetic D/H fractionation during microbial H₂ metabolism has been measured to be roughly two orders of magnitude larger than the corresponding carbon KIE (unverified), and genuine equilibrium isotope exchange with ambient water can occur simultaneously with the kinetic effect during active H₂-oxidizing metabolism, so a measured δD-H₂ downstream of a microbial mat reflects a superposition of both mechanisms, not a clean end-member signature [43].

Progressive, one-directional consumption of a substrate — microbial oxidation of a finite H₂ pool, or diffusive loss from a leaking sample container — is described by a Rayleigh distillation model: if $f$ is the remaining fraction of the original H₂ pool and $\alpha$ is the fractionation factor of the removal reaction,

$R = R_0\, f^{(\alpha-1)} \quad\Longleftrightarrow\quad \delta \approx \delta_0 + 1000(\alpha-1)\ln f$

Because $\alpha < 1$ for most H-removal kinetics (the light isotope is removed preferentially), the residual, unconsumed H₂ pool becomes progressively enriched in D (less negative δD) as $f$ decreases: the longer a natural H₂ show sits in contact with an active microbial sink, or the more a stored sample leaks, the less negative (more “shallow-looking”) its δD becomes. This equation underlies two of this page’s field caveats (see Environmental covariance, below).

The H₂–H₂O geothermometer

δD-H₂ is diagnostic, not merely descriptive, because molecular H₂ exchanges hydrogen isotopes with formation water at depth, and the equilibrium fractionation factor $\alpha_{H_2O-H_2}(T)$ is a strong, monotonic function of temperature. Because formation-water δD is comparatively well constrained relative to H₂’s own enormous range, a measured δD-H₂ paired with an assumed or independently measured δD-H₂O lets the last equilibration temperature be back-calculated — formally identical in structure to the carbonate clumped-isotope paleothermometer and to classic quartz–water and calcite–water oxygen-isotope geothermometers.

Two quantitative anchors illustrate the physics. The Kansas Precambrian-basement H₂ seeps associated with serpentinization of mafic/ultramafic rock near the Mid-Continent Rift show δD = −740 to −836 ‰, among the most deuterium-depleted natural gases measured, which back-calculates to an implied near-ambient equilibration temperature of roughly 10 °C, consistent with low-temperature, near-surface serpentinization and closely resembling H₂-rich gases independently described from the Samail ophiolite in Oman [5].

Separately, recent experimental calibration work confirms the H₂O(l)–H₂(g) D/H equilibrium fractionation factor is well-behaved and directly usable as a geothermometer from 3–90 °C, while flagging that isotopic exchange kinetics become fast enough above roughly 300 °C that ascending or cooling gas can retain a “memory” of a higher formation temperature even after passing through cooler shallow water; surface or spring δD-H₂ measurements can therefore systematically underestimate the true subsurface source temperature if partial re-equilibration occurred during ascent [7][8].

This nuance separates a naive “look up δD in a table” workflow from a defensible geothermometric interpretation, and it is why the STET review states that “the δD of H₂ alone is not an efficient tool to discriminate one source from another” and must be combined with the assumed or measured δD of coexisting H₂O to function as a geothermometer rather than a standalone origin label [4].

A parallel, less-developed geothermometer exists in the H₂–CH₄ system: because both species can approach isotopic equilibrium with the same formation water, and at sufficiently high temperature with each other, the combination of δD-H₂, δD-CH₄, and δ¹³C-CH₄ constrains formation conditions more tightly than any single ratio.

This is the multi-proxy logic Milkov’s 2022 review builds around the H₂–CH₄–H₂O ternary system, and the logic a 2023 published comment argues was not carried through completely, flagging that the diagnostic reliability of isotopic signatures of H₂ and its associated gases is more limited than the original paper contends [2][3]. The exchange marks the current state of the art: isotopes are the best available tool, but not an infallible one.

The empirical δD-H₂ envelope by origin

A 2024 compilation of 5,677 individual δD-H₂ measurements drawn from global literature and industrial datasets gives the largest empirical picture of δD-H₂ by origin, spanning green (electrolytic/biological), grey/turquoise (fossil-fuel-derived), brown (biomass/wood-burning), and white (naturally occurring geologic) hydrogen [1].

Green hydrogen is “readily distinguished by its considerable depletion in heavy isotopic species,” ranging −831 to −555 ‰, versus −377 to +196 ‰ for grey/turquoise fossil-sourced H₂, and −379 to 0 ‰ for brown biomass-burning H₂, against a quoted analytical uncertainty of close to ±1 ‰ [1]. White (naturally occurring geologic) hydrogen spans −996 to −49 ‰, a range wide enough to overlap every other category, which the source paper attributes to “diverse overlapping origins” [1].

The practical consequence is that a single δD-H₂ value in isolation, especially one in the −800 to −400 ‰ range where green/electrolytic, biotic, and geologic-crustal white hydrogen all overlap, is not by itself conclusive; it must be triangulated against companion isotope systems and geologic/hydrological context.

Independently, Milkov’s 2022 review proposes a coarser but mechanistically grounded two-tier split within “white” hydrogen itself: broadly defined crustal H₂ (δD-H₂ < −650 ‰) versus mantle and primordial H₂ (δD-H₂ from roughly −650 to −100 ‰), noting either category can additionally reflect temperature-dependent re-equilibration with water rather than a pristine source signature [2][12].

Read together with the Kansas/Oman serpentinization anchor above (−740 to −836 ‰, squarely in Milkov’s “crustal” bin despite reflecting bedrock water–rock reaction rather than radioactive decay), the practical reading for exploration is that the more negative the δD-H₂, the more confidently it points to a low-temperature, crustal, water–rock or radiolytic process, while a less negative value could reflect either a higher-temperature or mantle source or a shallow biotic/anthropogenic contaminant. This is why the single number cannot stand alone and must be read alongside He and CH₄ isotopes.

Atmospheric H₂: an unusually heavy background

Tropospheric background H₂ has direct field relevance and runs opposite to intuition: unlike most natural H₂, it is strongly deuterium-enriched, at δD ≈ +130 ‰, one of the largest δD values known for any natural terrestrial material [9][10][11].

This arises because two removal pathways dominate the atmospheric H₂ budget — soil microbial uptake and reaction with the OH radical — both of which display kinetic isotope effects that preferentially remove the light isotopologue, progressively enriching the residual atmospheric H₂ pool in D (a real-world illustration of the Rayleigh mechanism above); mass-balance closure of the global H₂ isotope budget additionally requires the photochemical (CH₄+OH-derived) source of H₂ itself to be strongly D-enriched, δD ≈ +100 to +200 ‰ [10][11].

In practice, the atmospheric H₂ “blank” that any near-surface sample is diluted with or contaminated by is isotopically heavy, while essentially every geologic, biotic, and industrial H₂ source of interest is deuterium-depleted (δD from −49 down to −996 ‰).

Any field or laboratory handling step that lets atmospheric air leak into a supposedly pure gas sample pulls the measured δD-H₂ upward (less negative, toward +130 ‰), a directionally predictable and in principle detectable air-contamination artifact distinct from the biological/geologic mixing artifacts discussed below.

Radiogenic and radiolytic signatures

Radiolytic H₂ is generated when ionizing radiation from the decay of U, Th, and K in host rock splits water molecules. A foundational experimental study irradiated water samples (pH 4, 7, 10; saline and fresh; aerobic and anaerobic) with a controlled γ-ray dose and measured an H₂ production yield of 0.40 ± 0.04 molecules per 100 eV of absorbed dose across a 900–3,500 Gray dose range, explicitly testing whether radiolytic H₂’s isotopic composition could serve as a source diagnostic for deep subsurface biosphere studies [6].

Radiolytic H₂ production scales with local α/β/γ flux and water/rock ratio, and is a recognized, quantitatively significant contributor to H₂ budgets in ancient, radiogenic-element-bearing Precambrian Shield fracture waters — the multi-decade research program on kilometer-deep Canadian, Fennoscandian, and South African shield fracture fluids has repeatedly documented mM-level dissolved H₂ attributed jointly to radiolysis and serpentinization, alongside abiogenic hydrocarbon gases from Fischer–Tropsch-type synthesis [6][46].

The companion helium signature is diagnostic in its own right: ⁴He-rich seepage, without a correspondingly high ³He/⁴He ratio, indicates radiolytic/radiogenic crustal helium production rather than mantle helium, because radiogenic ⁴He from α-decay of U/Th accumulates in old, low-permeability crustal rock while ³He (predominantly primordial mantle or cosmogenic) does not.

The STET review states this relationship directly: “He enriched in ⁴He is known to originate from radiolysis. Thus, a ⁴He-rich seepage infers crustal H₂ formed via radiolysis” [4]. This is the direct bridge between H₂ isotopes and noble-gas isotopes (see Helium & Co-Tracers): radiolysis is simultaneously an H₂-generating process and a ⁴He-generating process, so a sample’s He isotope ratio independently corroborates or contradicts a radiolytic interpretation drawn from δD-H₂ alone.

Variants & subtypes

Isotopic fingerprinting is a family of complementary measurement systems, each diagnostic of a different aspect of gas origin, rather than a single measurement:

• δD-H₂ (primary fingerprint). Origin-dependent, overlapping ranges from ~+196 ‰ (heaviest grey/fossil) down to −996 ‰ (most depleted white/geologic); also functions as a formation/re-equilibration-temperature proxy via the H₂–H₂O geothermometer above [1][2][7][8].
• δ¹³C-CH₄ and δD-CH₄ (Bernard/Whiticar genetic classification of co-occurring methane). The classical genetic diagram plots δ¹³C-CH₄ against the molecular ratio C₁/(C₂+C₃) (the original Bernard diagram) or against δD-CH₄ (the Whiticar diagram), separating thermogenic gas (generally δ¹³C-CH₄ > −50 ‰) from microbial/biogenic gas (generally δ¹³C-CH₄ < −60 ‰), with microbial gas further split by pathway: the CO₂-reduction pathway (dominant in marine sediments and deep groundwater) produces especially ¹³C-depleted methane (δ¹³C-CH₄ commonly −110 to −60 ‰) with comparatively less depleted δD-CH₄ (roughly −250 to −170 ‰), while the acetate-fermentation pathway (dominant near-surface, freshwater/wetland settings) produces less ¹³C-depleted methane (δ¹³C-CH₄ roughly −65 to −50 ‰) but strongly D-depleted δD-CH₄ (roughly −400 to −250 ‰) — the two pathways separate more cleanly on δD-CH₄ than on δ¹³C-CH₄ alone [14]. Both diagrams were substantially revised and re-calibrated against a global dataset of more than 20,000 gas samples in 2018 [13].
• Abiotic methane (Fischer–Tropsch-type, FTT). Methane generated by catalytic CO₂/CO hydrogenation ($\mathrm{CO_2 + 4H_2 \rightarrow CH_4 + 2H_2O}$) on Fe/Ni/Cr mineral surfaces during serpentinization is isotopically distinctive but genuinely difficult to separate cleanly from thermogenic gas by δ¹³C alone: reported abiotic δ¹³C-CH₄ spans from as depleted as −47 ‰ to strongly enriched positive values, and can display the classic isotope-reversal pattern (δ¹³C₁ > δ¹³C₂ > δ¹³C₃, opposite the normal thermogenic ordering) — though this reversal pattern’s reliability as a discriminator is itself disputed in the literature, with at least one source arguing the same ordering instead reflects high-thermal-maturity biotic gas (unverified) [20][21]. The Lost City Hydrothermal Field (Mid-Atlantic Ridge) is the type example where bulk stable-isotope methods alone cannot cleanly resolve the microbial-vs-abiotic contribution to a mantle-carbon-sourced methane pool, motivating the clumped-isotope method below [20].
• Clumped isotopologues of methane (Δ¹³CH₃D, Δ¹²CH₂D₂). Rather than measuring bulk δ¹³C and δD separately, clumped-isotope analysis measures the excess abundance (relative to a random statistical distribution) of doubly substituted methane isotopologues. Because this “clumping” is governed purely by intramolecular thermodynamic equilibrium once methane has finished forming, it functions as an internally consistent geothermometer independent of the isotopic composition of the parent source material, a substantial advance over δ¹³C/δD alone. This is the seminal application demonstrated by Stolper et al. (2014), who used Δ¹³CH₃D to directly reconstruct formation temperatures of thermogenic and biogenic methane [15][17]. Modern calibration work has extended the experimental/theoretical reference frame for both isotopologues from 1–500 °C [16][18], and a 2025 advance demonstrated rapid, high-sensitivity clumped-isotope analysis by mid-infrared laser absorption spectroscopy (at 1076.97 cm⁻¹ / 9.29 µm for ¹²CH₂D₂ and 1163.47 cm⁻¹ / 8.60 µm for ¹³CH₃D) as a faster, lower-cost alternative to high-resolution IRMS [17][19].
• δ¹³C-CO₂. Co-measured wherever CO₂ is present; used both as an independent genetic tracer (biogenic CO₂ from organic decay is typically isotopically light; deep/mantle or carbonate-buffered CO₂ is comparatively heavy) and, via the δ¹³C-C₁ vs. δ¹³C-CO₂ cross-plot, as one leg of the revised three-diagram genetic classification scheme [13].
• ³He/⁴He (R/Ra). The primary discriminator of mantle-vs-crustal fluid plumbing: typical crustal radiogenic helium sits at R/Ra ≈ 0.01–0.05, atmospheric helium is R/Ra = 1 by definition, and typical upper-mantle helium is R/Ra ≈ 8 [25][27]. Field examples show the full continuum in practice: Sacramento Basin natural gas wells range from R/Ra 0.11 to 2.75, implying 1–34% mantle-derived helium in different wells of the same basin [26]. Full treatment in Helium & Co-Tracers.
• ⁴He/²⁰Ne (air-correction ratio). Atmosphere and air-saturated water are enriched in ²⁰Ne relative to both mantle and crustal reservoirs (essentially degassed of primordial/atmospheric neon), so ⁴He/²⁰Ne is the standard tool for detecting and correcting for atmospheric contamination of a helium sample before its ³He/⁴He ratio is interpreted: air itself has ⁴He/²⁰Ne ≈ 0.318–0.319, and any sample with a substantially higher ratio carries a real crustal or mantle helium excess above the unavoidable atmospheric/groundwater-recharge background [28].
• ⁴⁰Ar/³⁶Ar and ²¹Ne (secondary corroborating radiogenic/nucleogenic tracers). Atmospheric ⁴⁰Ar/³⁶Ar = 298.6 is the reference baseline; radiogenic ⁴⁰Ar (from ⁴⁰K β-decay) accumulates in situ in crustal rock and fluid, so ratios elevated above the atmospheric value indicate a genuine crustal (or, at higher enrichment, mantle) argon component, directly analogous to the He system [29]. ²¹Ne is produced nucleogenically via (α,n) and (n,α) reactions on O, Mg, and F nuclei in host rock; because ⁴⁰Ar, ⁴He, and ²¹Ne are all radiogenic/nucleogenic products whose production rates scale with source-rock K, U, Th content, the ratios among them constrain the K/U ratio of the source material and provide a cross-check on whether a fluid’s noble-gas budget is consistent with a purely crustal origin [29][28].
• ³H/³He (tritium-helium) age-dating. A specialized noble-gas method for dating young (post-1965, thermonuclear-bomb-pulse-labeled) groundwater by the in-growth of “tritiogenic” ³He from ³H decay; not a source-origin fingerprint per se, but directly relevant to environmental covariance work because it independently ages the water a candidate H₂ show may have equilibrated with, which bears on the H₂–H₂O geothermometer’s interpretation [30; full treatment cross-referenced to Mass Spectrometry §2].

Performance envelope

Two tables cover this: what a given signature means, and what it costs in precision, sample, time, and money to obtain it.

Table 1 — Origin-discrimination: signature → value range → origin implication → method.

Signature	Value range	Origin implication	Method	Source
δD-H₂	−831 to −555 ‰	Green H₂: electrolytic or biological (nitrogenase/hydrogenase)	GC-IRMS / TC/EA-IRMS	[1]
δD-H₂	−377 to +196 ‰	Grey/turquoise H₂: fossil-fuel-derived (SMR, pyrolysis)	GC-IRMS	[1]
δD-H₂	−379 to 0 ‰	Brown H₂: biomass/wood combustion	GC-IRMS	[1]
δD-H₂	−996 to −49 ‰	White/geologic H₂: overlapping crustal + mantle + re-equilibrated sources	GC-IRMS	[1]
δD-H₂	< −650 ‰	Broadly “crustal” H₂ (serpentinization, radiolysis, low-T water–rock)	GC-IRMS + geologic context	[2][12]
δD-H₂	≈ −650 to −100 ‰	Mantle / primordial H₂, or crustal H₂ re-equilibrated at higher T	GC-IRMS + He/CH₄ corroboration	[2][12]
δD-H₂	−740 to −836 ‰ (field example)	Serpentinization at ~10 °C near-ambient equilibration (Kansas; resembles Oman ophiolite gases)	GC-IRMS + H₂–H₂O geothermometer	[5]
δD-H₂	≈ +130 ‰ (atmosphere)	Tropospheric background — NOT a geologic source; contamination indicator	GC-IRMS, atmospheric flask network	[9][10][11]
δD-CH₄	−250 to −170 ‰	Microbial, CO₂-reduction pathway	GC-IRMS on co-produced CH₄	[14]
δD-CH₄	−400 to −250 ‰	Microbial, acetate-fermentation pathway	GC-IRMS	[14]
δD-CH₄	−275 to −100 ‰	Thermogenic (kerogen/oil cracking)	GC-IRMS	general natural-gas isotope literature
δ¹³C-CH₄	< −60 ‰	Microbial/biogenic methane	GC-IRMS	[13]
δ¹³C-CH₄	> −50 ‰	Thermogenic methane	GC-IRMS	[13]
δ¹³C-CH₄	−47 to strongly positive	Abiotic (Fischer–Tropsch-type) methane — overlaps thermogenic; needs isotope-reversal or clumped-isotope check	GC-IRMS + clumped isotopes	[20][21]
δ¹³C₁ vs. δ¹³C₂ vs. δ¹³C₃ ordering	Normal: C₁ < C₂ < C₃ (heavier with chain length)	Thermogenic	GC-IRMS (multi-alkane)	[21]
δ¹³C₁ vs. δ¹³C₂ vs. δ¹³C₃ ordering	Reversed: C₁ > C₂ > C₃	Abiotic (FTT) methane — diagnostic value disputed (unverified)	GC-IRMS (multi-alkane)	[21]
Δ¹³CH₃D / Δ¹²CH₂D₂	Calibrated 1–500 °C	Direct methane formation-temperature geothermometer, source-composition-independent	HR-IRMS or mid-IR laser spectroscopy	[15][16][17][18]
³He/⁴He (R/Ra)	≈ 8 Ra	Mantle-derived helium/H₂ plumbing	Noble-gas static MS	[25][27]
³He/⁴He (R/Ra)	≈ 0.01–0.05 Ra	Crustal radiogenic helium	Noble-gas static MS	[25][27]
³He/⁴He (R/Ra)	1 Ra (by definition)	Atmospheric helium / air contamination	Noble-gas static MS	[25]
³He/⁴He (R/Ra)	0.11–2.75 Ra (field example)	Sacramento Basin: 1–34% mantle-derived He fraction	Noble-gas static MS	[26]
⁴He/²⁰Ne	≈ 0.318–0.319 (air)	Baseline for air-contamination correction	Noble-gas static MS	[28]
⁴He/²⁰Ne	≫ air value	Genuine crustal or mantle He excess, not atmospheric	Noble-gas static MS	[28]
⁴⁰Ar/³⁶Ar	298.6 (air)	Baseline atmospheric argon	Noble-gas static MS	[29]
⁴⁰Ar/³⁶Ar	≫ 298.6	Radiogenic (⁴⁰K-derived) crustal argon component	Noble-gas static MS	[29]
²¹Ne (nucleogenic excess)	Above air-saturated-water background	(α,n)/(n,α) production in situ; corroborates crustal radiogenic budget	Noble-gas static MS	[29]
Radiolytic H₂ yield	0.40 ± 0.04 molecules/100 eV	Quantitative radiolysis production rate, dose 900–3,500 Gy	Experimental γ-irradiation + gas analysis	[6]
H₂/CH₄ molecular ratio	> ~40 (proposed)	Proposed abiotic indicator, shown unreliable alone	GC	[45]

Table 2 — Measurement chain: precision, sample requirements, turnaround, cost class.

Method	Precision (1σ)	Sample requirement	Turnaround	Cost class	Source
Offline combustion + dual-inlet IRMS (“high-precision”)	δ¹³C ±0.1 ‰; δD ±3.5 ‰	Gas in IsoTube®/IsoBag®/steel cylinder	Days–weeks (batch)	Premium tier	[33]
Online GC-IRMS (continuous-flow, “standard-precision”)	δ¹³C ±0.3 ‰; δD ±5 ‰	Same containers	Faster than offline; days	Standard tier	[33]
TC/EA (or Cr-EA) reduction-furnace IRMS, generic academic practice	δD ±2 ‰ (2σ) typical target; reported field range ±0.5–2.0 ‰ for organics/waters, up to ±5 ‰ for gas-phase H₂ given H₃⁺ chemistry	≥20 mg solid-equivalent for <3 ‰ reproducibility; <20 mg gives 6–15 ‰ variance	Batch, hours of instrument time per run + days of queue	Lab-infrastructure capex	[41]; cf. Mass Spectrometry
Clumped-isotope HR-IRMS (Δ¹³CH₃D/Δ¹²CH₂D₂)	Sub-‰ on clumped Δ values	Larger CH₄ sample volumes than bulk δ¹³C/δD	Slow — specialized, low-throughput academic method	High-capex, few labs worldwide	[15][16][18]
Mid-IR laser clumped-isotope spectroscopy	Comparable precision to HR-IRMS at higher throughput (claimed)	Smaller sample than HR-IRMS (claimed)	Faster than HR-IRMS	Emerging, lower capex than magnetic-sector HR-IRMS	[17][19]
Noble-gas static-vacuum MS (He/Ne/Ar isotopes)	Field/instrument-specific; not separately quantified in sources reviewed	Copper tube (~1 cm OD, ~19 cm³ water) or evacuated flask; no elastomer contact	Days–weeks, specialist labs	High-capex, few labs worldwide	[31][32]; cf. Mass Spectrometry
Generic gas-sample compositional + isotope package (“Full GC”)	He/H₂ detection 25 ppm–100%	IsoTube/IsoBag/steel cylinder	Quote-dependent	Quote-only	[33]
USGS Reston Stable Isotope Lab (context, water-focused)	Not H₂-gas-specific	Water/solid	2–4 weeks routine; 24–48 hr priority (water H/O only)	Government lab rate; contact for quote	[36]

In summary, realistic exploration-grade δD-H₂ precision sits at roughly ±2 to ±5 ‰ (2σ) for routine GC-IRMS work, tightening to ~±3.5 ‰ (1σ) for premium offline dual-inlet methods [33], fine enough to separate the ~150–800 ‰-wide origin bins in Table 1, but not fine enough, on its own, to resolve fine structure within the wide, overlapping “white” H₂ envelope (−996 to −49 ‰) without companion He/CH₄ data.

Environmental covariance & cross-sensitivity

Unlike the electrochemical, MOX, and catalytic sensor families treated elsewhere in this wiki, IRMS-based isotope measurement does not suffer classical chemical cross-sensitivity — CO, H₂S, and humidity do not directly bias a properly chromatographically separated δD-H₂ measurement. The vulnerabilities here are about sample integrity and interpretation rather than instrument chemistry, and every one below biases a result in a predictable direction.

Atmospheric contamination during sampling. Because atmospheric H₂ is isotopically heavy (δD ≈ +130 ‰) while virtually every geologic/biotic H₂ source of interest is deuterium-depleted, any air leak into a sample pulls the measured δD-H₂ upward, a directionally predictable artifact but one that is easy to induce with imperfect field technique (an under-purged sample line, a slow valve leak, an incompletely evacuated container) [9][10][11]. This compounds with, and can be mistaken for, a genuine mixed-source signal.

Post-emplacement microbial consumption. Per the Rayleigh mechanism above, active H₂-oxidizing microbial metabolism preferentially consumes the light isotopologue, so a natural H₂ show that has been sitting in contact with an active near-surface microbial sink for any length of time will read isotopically heavier (less negative δD) than its true source signature, and could in principle be mistaken for a shallower or more mantle-like source than the one that actually generated it [43].

This is mechanistically the same soil microbial sink responsible for consuming the majority of the global atmospheric H₂ budget (see Geologic Context) — it operates on any H₂ pool, geologic or otherwise, that transits the shallow biosphere, and simultaneous equilibrium exchange with local water during active metabolism further complicates a clean read [43].

Partial re-equilibration during ascent. D–H exchange kinetics between H₂ and H₂O become fast above roughly 300 °C, meaning gas that formed hot at depth and cooled during ascent can retain a δD signature partway between its true formation temperature and the (cooler) temperature of its last significant water contact — biasing the geothermometer toward underestimating true subsurface formation temperature [7][8]. This is a systematic, one-directional bias, not random noise, and should be flagged explicitly whenever δD-H₂-derived temperatures are reported to decision-makers.

Non-linear isotope mixing. When two H₂ sources of substantially different isotopic composition mix (e.g., a deep geologic component and a shallow microbial/atmospheric component), the mixture’s bulk δD is not a simple linear (mole-fraction-weighted) average of the two end-member δD values once the isotope ratios differ by more than a few tens of per mil — correct mass-balance requires mixing in concentration/mole-fraction space (using $R = D/H$ or isotope amounts) rather than directly averaging δ values, a well-documented general pitfall in stable-isotope mixing-model literature that applies with force here given how wide the H₂ δD spread is (nearly 1000 ‰ peak to peak).

A naive linear-δ mixing calculation can therefore misestimate the mixing fraction of a deep vs. shallow H₂ component by a material amount, particularly when the deep end-member is near −900 ‰ and the shallow/atmospheric end-member is near +100 to +130 ‰.

Diffusive/kinetic fractionation during migration and sample containment. Graham’s-law diffusion (flux ratio $\propto \sqrt{m_{light}/m_{heavy}}$) means H₂ (light isotopologue) diffuses measurably faster than HD through any porous medium, membrane, or micro-leak; this is the same physical property — H₂’s exceptionally high diffusivity relative to CH₄/CO₂ — already documented as a first-order confound for concentration measurements in Geologic Context, and it applies with equal force to isotope ratios: a slowly leaking sample container should be expected to lose light H₂ preferentially, driving the residual gas’s measured δD progressively less negative (heavier) the longer it sits or the more it leaks.

This is a physically well-founded inference rather than an independently confirmed empirical result, and a lab QA/QC program (see Calibration & QA/QC) should watch for it.

H₂’s notoriously high permeability through many polymer materials reinforces why gas-tight, low-permeability containment — steel cylinders, IsoTubes, or, for the noble-gas companion measurements, copper tubes with metal-to-metal UHV seals and zero elastomer contact — is standard practice, not an over-engineered precaution [44][31][32].

H₃⁺ interference and instrument-side artifacts. As detailed in Mass Spectrometry §1.3, the ion-molecule reaction $\mathrm{H_2^+ + H_2 \rightarrow H_3^+ + H}$ inside any EI ion source creates a source-pressure-dependent artifact at m/z = 3 that directly contaminates the HD⁺ signal used to compute δD.

Production IRMS software runs an automated “H₃⁺ factor” correction that must be characterized before each analytical session and has been shown stable to within ~7 ppm/nA over several weeks on a well-maintained instrument — but this is an active calibration burden, not a “set and forget” correction, and a drifting or poorly characterized H₃⁺ factor is a silent source of systematic δD bias [41][42].

Sample-size floor. Reduction-furnace (Cr-EA/TC-EA) methods show a sharp precision cliff below roughly 20 mg solid-equivalent sample (or the gas-phase equivalent): sub-threshold samples show 6–15 ‰ variance versus better than 3 ‰ above threshold [41] — directly relevant to planning how much gas volume a field sample must contain, especially for low-flux seeps or small soil-gas draws.

The bias asymmetry

Every artifact above — atmospheric contamination, microbial alteration, incomplete ascent re-equilibration, container leakage — biases δD-H₂ in the same direction: toward less negative, more “shallow/heavy”-looking values. The reverse (a genuinely shallow artifact masquerading as deep) is much less physically plausible. In practice, a strongly, unambiguously deep-looking δD-H₂ (very negative, e.g., < −650 ‰) corroborated by a clearly non-atmospheric ³He/⁴He is hard to produce through sampling error alone. A marginal or ambiguous result cannot be resolved without repeat sampling and companion-isotope confirmation.

Co-evolved-gas cross-checks, not cross-sensitivities. The main compositional cross-check is co-measuring companion gases (He, CH₄, CO₂) from the same sample, both because a single-species measurement cannot access the multi-proxy discrimination this technique requires, and because inconsistent behavior across species (e.g., a δD-H₂ that looks deep-crustal paired with a ³He/⁴He that looks purely atmospheric) is itself one of the most useful artifact detectors available.

Use for geologic H₂

Field summary

Isotopic and noble-gas fingerprinting is the only measurement suite in this wiki that can make an affirmative, physically defensible case that a specific H₂ show is deep-sourced rather than a drilling artifact, a microbial signal, or an industrial leak. It is essential for confirming a high-value target and unsuitable as a real-time or field-portable tool.

Treat it as a confirmatory laboratory bundle (δD-H₂ + δD/δ¹³C-CH₄ + ³He/⁴He with ⁴He/²⁰Ne air correction), budgeted at days-to-weeks turnaround and specialist-lab cost, applied to a short list of candidate shows that have already cleared a cheaper, faster screening stage (soil-gas survey, mud-gas logging, concentration sensing). A single δD-H₂ number should not be presented as conclusive proof of deep origin.

The case for isotopes. No field-portable concentration sensor treated elsewhere in this wiki — electrochemical, MOX, catalytic-bead, thermal-conductivity, or even portable Raman/IR — can distinguish a genuinely deep geologic H₂ system from a drilling artifact, a microbial fermentation signal, or a leaking industrial source. Mud Gas Logging flags this as an open gap: isotopic discrimination as an artifact-vs-natural diagnostic is conceptually attractive but was not found documented in the mud-gas-logging literature reviewed.

The gap is real but closeable: δD-H₂, δD/δ¹³C-CH₄, and ³He/⁴He together are the only measurement suite that can, with defensible physical-chemical reasoning, make the affirmative case that a specific show is deep-sourced.

Practical limitations

Practical limitations.

1. Turnaround is days to weeks. Every sourced turnaround figure in Table 2 is batch-lab-scale [33][36]. Isotope data cannot inform a real-time drilling decision the way a wellsite mud-gas concentration reading can; it is a confirmatory tool, not a real-time one, and program timelines must budget the lag explicitly.
2. Cost and specialist-lab scarcity make “sample everything” infeasible. High-precision hydrogen-isotope work, and especially noble-gas static-MS work, is concentrated in a small number of specialist commercial and academic labs. Notably, UC Davis’s Stable Isotope Facility — a major light-element IRMS provider referenced repeatedly in the broader isotope-geochemistry literature — has announced it will sunset its current operations effective July 2026 [35], a sign that the commercial and academic lab landscape for this measurement class is neither large nor fully stable. Qualify at least two independent labs rather than depend on a single provider.
3. The white-H₂ δD envelope overlaps every other category. At −996 to −49 ‰, “white/geologic” spans the entire range occupied separately by green, grey, and brown H₂ [1]. A single δD-H₂ measurement is necessary but not sufficient evidence of deep origin; report it to an investor or partner together with the companion He/CH₄ data, not as conclusive on its own, a point the peer-reviewed literature makes as well [2][3][4].
4. Sample-integrity discipline is demanding. Air contamination, microbial post-emplacement alteration, partial re-equilibration during ascent, and slow container leaks all bias δD-H₂ in the same direction (see the bias-asymmetry note above); a poorly handled sample from a genuinely deep system can masquerade as a shallow artifact, though the reverse is much less physically plausible.
5. No downhole isotope-measurement capability exists, and none should be expected. Consistent with Downhole Deployment and Mass Spectrometry’s “hard no” on downhole MS, every isotope method here is a surface/laboratory technique fed by a physical sample (copper tube, IsoTube, steel cylinder) collected downhole or at surface and shipped or hand-carried to a lab. There is no fiber-optic, electrochemical, or in-situ analog for isotope ratio determination at exploration-relevant precision, now or on any near-term roadmap identified.

Recommended workflow

Recommended workflow.

1. Treat δD-H₂ + δ¹³C/δD-CH₄ + ³He/⁴He (with ⁴He/²⁰Ne air correction) as a minimum co-sampled bundle for every candidate show worth a second look.
2. Collect gas samples in IsoTubes/steel cylinders (bulk composition and δD-H₂/δ¹³C-CH₄) and, wherever a dissolved-gas or free-gas sample of sufficient volume exists (see Dissolved Gas & Groundwater), a copper-tube noble-gas sample in parallel, using the same field visit — remobilizing to a site for a second sampling round is a real cost a Calibration & QA/QC-style program should design out from the start.
3. Budget multi-week lab turnaround into program timelines rather than treating isotope results as available before the next field decision.
4. Qualify at least two labs given the concentrated, and currently contracting, commercial landscape.
5. Report δD-H₂ together with the companion He data (and, where methane is present, clumped-isotope or isotope-reversal data), rather than presenting a single number as conclusive of deep origin.

Instruments, labs & standards

Vendor / lab	Product / service	Category	Key specs (as sourced)	Notes	Source
Isotech Laboratories (a Stratum Reservoir company)	Compositional + stable-isotope gas analysis (“Basic GC,” “Full GC,” high-precision offline, standard online GC-IRMS)	Commercial gas-isotope lab	Full GC: He/H₂ 25 ppm–100%; offline IRMS δ¹³C ±0.1‰ / δD ±3.5‰; online GC-IRMS δ¹³C ±0.3‰ / δD ±5‰	Self-described largest commercial stable-isotope lab; C1–C5 hydrocarbon isotope capability	[33]
Stratum Reservoir	IsoTube® / IsoBag® gas sampling system	Field sample containment	3+ year claimed chemical/isotopic storage stability if leak-free; pre-evacuated, faster sampling, no condensation/air-flush issue	Industry-standard mud-gas/wellhead sampling container	[34]
UC Davis Stable Isotope Facility (SIF)	Continuous-flow IRMS services, H/C/N/O/S	Academic core lab	14 IRMS instruments + cavity ring-down spectrometer; >190,000 samples/yr historically	Announced sunset effective July 26, 2026; do not treat as a durable long-term vendor without confirming successor arrangements	[35]
USGS Reston Stable Isotope Laboratory	Analytical services, various isotope systems	Government reference lab	2–4 week routine turnaround; 24–48 hr priority (water H/O only, not gas)	Contact required for current pricing	[36]
Thermo Fisher Scientific	253 Plus IRMS	Gas-source magnetic-sector IRMS	Hydrogen continuous-flow capability; all-metal 10 kV EI source; dynamic range >10⁷; variable inlet-side conductance (VISC)	Marketed successor to the long-running MAT 253 line	[37]
Thermo Fisher Scientific	MAT 253	Gas-source magnetic-sector IRMS	Long-standing “gold standard” reference instrument for gas IRMS	Widely deployed legacy instrument base across academic/commercial labs	[38]
Thermo Fisher Scientific	253 Ultra	High-resolution IRMS	Purpose-built for clumped-isotope work (new collector design)	Relevant to Δ¹³CH₃D/Δ¹²CH₂D₂ methane-formation-temperature work	[39]
Elementar (Isoprime)	isoprime precisION	Gas-source IRMS	Flexible multi-inlet architecture (centrION continuous-flow interface); combinable inlet configurations	Alternative platform to Thermo line for δD-H₂/δ¹³C-CH₄ work	[40]
(Static-vacuum noble-gas MS family, generically referenced)	Magnetic-sector noble-gas MS with Faraday cup + electron-multiplier detection	Noble-gas isotope MS	Faraday cups for abundant isotopes (⁴He, ²⁰Ne, ⁴⁰Ar); electron multiplier for trace isotopes (³He, Kr, Xe)	Full vendor/model detail developed in Mass Spectrometry	cross-ref Mass Spectrometry
USGS Groundwater Dating Laboratory	³H/³He sampling protocol	Sampling standard/procedure	Copper-tube collection (~1 cm OD, ~19 cm³ water), refrigeration-type pinch clamps, no O-ring/elastomer seal at the UHV interface	Directly reusable protocol for any noble-gas-in-water sampling program	[31]
Eawag (Swiss Federal Institute of Aquatic Science and Technology)	Copper-tube-fed static-vacuum MS extraction/purification line	Noble-gas MS instrumentation	UHV-tight copper-tube-to-cutting-edge seal, no elastomer connections	Same “no elastomer” design philosophy applies broadly	[32]
VSMOW / VSMOW2, SLAP2 (IAEA reference materials)	δD/δ¹⁸O primary and secondary reference standards	Isotope reference standard	International Atomic Energy Agency-distributed water standards defining the VSMOW δD = 0 anchor and the SLAP2 low end of the scale	Every δD-H₂ value on this page is ultimately traceable to this standard; see Standards & Regulations	general isotope-geochemistry practice
Thermo Isodat software	H₃⁺-factor correction subroutine	Data-reduction software	Automated H₃⁺ interference correction; factor stability ~6.8 ± 0.07 ppm/nA over weeks on a maintained instrument	Not a hardware product but a load-bearing part of the measurement chain	[41][42]

Open questions

• Extract the polynomial coefficients ($A/T^2+B/T+C$-type fits) from the H₂O(l)–H₂(g) fractionation-factor papers [7][8] so a formation/re-equilibration temperature can be computed directly from a measured δD-H₂, rather than relying on qualitative direction-of-effect statements.
• Confirm current, quotable pricing and turnaround for a full δD-H₂ + δ¹³C/δD-CH₄ + noble-gas-isotope bundle from at least two labs — every cost figure here beyond Isotech’s stated precision numbers is unconfirmed or “quote-only.”
• Identify a successor lab or confirm continuity plans following UC Davis SIF’s announced July 2026 sunset [35] before any sampling program is designed around that facility specifically.
• Directly source the leak/diffusive-fractionation artifact behavior described above — the Graham’s-law prediction that leaking H₂ samples drift isotopically heavier is physically well-motivated but lacks a dedicated empirical case study in the literature reviewed.
• Obtain the exact revised genetic-diagram boundaries from the 2018 >20,000-sample recalibration [13], rather than the older Bernard/Whiticar numbers used as first-order anchors in Table 1 — the original paper was accessible only via abstract this session.
• Scope mid-infrared laser clumped-isotope instrumentation [17][19] as a lower-cost, faster alternative to HR-IRMS for methane-formation-temperature work, given how thin the HR-IRMS lab landscape is worldwide.
• Confirm with a wellsite mud-gas logging provider whether a companion noble-gas (copper-tube) sample can be added to an existing IsoTube workflow without an additional rig-site visit.
• Verify the white- and green-hydrogen δD ranges directly against the full underlying dataset [1] to confirm how many measurements fall in each bin and whether the reported ranges are driven by outliers.

Related pages

• Helium & Co-Tracers — the noble-gas half of this fingerprinting bundle, treated in its own depth
• Geologic Context — the field-practice anchor this page’s interpretive framework serves
• Mass Spectrometry — the instrumentation deep-dive (IRMS, quadrupole, H₃⁺ chemistry) underlying every measurement on this page
• Mud Gas Logging — the wellsite concentration workflow this page’s isotope bundle is meant to confirm, not replace
• Dissolved Gas & Groundwater — sampling context for water-hosted H₂/He where noble-gas copper-tube sampling applies
• Soil Gas Surveys — the near-surface screening stage that should precede committing to an isotope bundle
• Downhole Deployment — why isotope methods have no downhole analog
• Calibration & QA/QC — sample-integrity discipline (leak testing, container selection) this page depends on
• Environmental Covariance — the broader cross-sensitivity framework this page’s artifact discussion sits within
• Why H₂ Is Hard — why concentration alone never resolves origin
• Standards & Regulations — VSMOW and related reference-standard traceability
• Glossary — δD, R/Ra, Rayleigh distillation, and other terms used throughout this page
• Vendors — consolidated vendor/lab directory

References

Full reference list (0 entries)

[1] Gibson, J.J., Eby, P., Jaggi, A. (2024). “Natural isotope fingerprinting of produced hydrogen and its potential applications to the hydrogen economy.” International Journal of Hydrogen Energy 66, 468–478. https://www.sciencedirect.com/science/article/pii/S0360319924013338; PDF: https://web.uvic.ca/~jjgibson/mypdfs/hydrogen_fingerprint.pdf

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

[3] “Comment on ‘Molecular hydrogen in surface and subsurface natural gases: Abundance, origins and ideas for deliberate exploration’ by Alexei V. Milkov [Earth-Science Reviews, 230, 2022: 104063].” Earth-Science Reviews (2023). https://www.sciencedirect.com/science/article/abs/pii/S0012825223000272

[4] “Natural H2 exploration: tools and workflows to characterize a play.” Science and Technology for Energy Transition (STET) 78, 27 (2023). https://www.stet-review.org/articles/stet/full_html/2023/01/stet20230116/stet20230116.html

[5] Coveney, R.M. et al. (1987). “Serpentinization and the origin of hydrogen gas in Kansas.” AAPG Bulletin 71(1), 39–48. https://archives.datapages.com/data/bulletns/1986-87/data/pg/0071/0001/0000/0039.htm; OSTI record: https://www.osti.gov/biblio/6532504

[6] Lin, L.-H., Slater, G.F., Sherwood Lollar, B., Lacrampe-Couloume, G., Onstott, T.C. (2005). “The yield and isotopic composition of radiolytic H2, a potential energy source for the deep subsurface biosphere.” Geochimica et Cosmochimica Acta 69(4), 893–903. https://www.sciencedirect.com/science/article/abs/pii/S0016703704006271

[7] “Kinetics of D/H isotope fractionation between molecular hydrogen and water.” Geochimica et Cosmochimica Acta (2018). https://www.sciencedirect.com/science/article/abs/pii/S0016703718305325; OSTI: https://www.osti.gov/biblio/1570230

[8] “Experimental determination of hydrogen isotopic equilibrium in the system H2O(l)-H2(g) from 3 to 90 °C.” Geochimica et Cosmochimica Acta (2025). https://www.sciencedirect.com/science/article/abs/pii/S0016703725001073

[9] “Hydrogen isotope biogeochemistry.” Wikipedia, citing Hoefs, J. (2009), Stable Isotope Geochemistry. https://en.wikipedia.org/wiki/Hydrogen_isotope_biogeochemistry

[10] Rahn, T., Eiler, J.M., Boering, K.A. et al. (2003). “Extreme deuterium enrichment in stratospheric hydrogen and the global atmospheric budget of H2.” Nature 424, 918–921. https://www.nature.com/articles/nature01917

[11] “Modeling the photochemical origins of the extreme deuterium enrichment in stratospheric H2.” Journal of Geophysical Research: Atmospheres (2007). https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2006JD007403

[12] “Natural hydrogen (white hydrogen) exploration methods and identification of sources: A comprehensive overview.” ScienceDirect (2026). https://www.sciencedirect.com/science/article/pii/S1342937X26000651

[13] Milkov, A.V., Etiope, G. (2018). “Revised genetic diagrams for natural gases based on a global dataset of >20,000 samples.” Organic Geochemistry 125, 109–120. https://www.sciencedirect.com/science/article/abs/pii/S0146638018301980

[14] Whiticar, M.J. (1999). “Carbon and hydrogen isotope systematics of bacterial formation and oxidation of methane.” Chemical Geology 161, 291–314. https://www.sciencedirect.com/science/article/abs/pii/S0009254199000923

[15] Stolper, D.A. et al. (2014). “Formation temperatures of thermogenic and biogenic methane.” Science 344(6191), 1500–1503. https://www.science.org/doi/abs/10.1126/science.1254509

[16] “Comparison of Experimental vs Theoretical Abundances of 13CH3D and 12CH2D2 for Isotopically Equilibrated Systems from 1 to 500 °C.” ACS Earth and Space Chemistry (2019). https://pubs.acs.org/doi/abs/10.1021/acsearthspacechem.9b00244

[17] “Rapid High-Sensitivity Analysis of Methane Clumped Isotopes (Δ13CH3D and Δ12CH2D2) Using Mid-Infrared Laser Spectroscopy.” PMC (2025). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11755397/

[18] “Application of an absolute reference frame for methane clumped-isotope analyses.” Chemical Geology 646, 121922 (2024) — NOT Geochimica et Cosmochimica Acta as originally listed here (venue corrected after verification). https://www.sciencedirect.com/science/article/abs/pii/S0009254124000020

[19] “Correction to ‘Rapid High-Sensitivity Analysis of Methane Clumped Isotopes (Δ13CH3D and Δ12CH2D2) Using Mid-Infrared Laser Spectroscopy.’” PMC (2025). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12120809/

[20] “Multiple origins of methane at the Lost City Hydrothermal Field.” Earth and Planetary Science Letters (2010). https://www.sciencedirect.com/science/article/abs/pii/S0012821X10003638

[21] “Does This Finally Settle the Age-Old Debate over Abiotic Methane?” AAPG Explorer (2022). https://www.aapg.org/news-and-media/explorer/does-this-finally-settle-the-age-old-debate-over-abiotic-methane/ (does not support the claim that reversed δ¹³C₁>δ¹³C₂>δ¹³C₃ ordering indicates abiotic methane — the article’s central argument is the opposite, that this “rollover” pattern in the Arkoma/Appalachian/Barnett/Fayetteville gases is caused by high thermal maturity of biotic organic matter and “has nothing to do with abiogenesis”)

[22] “Reduced gas seepages in ophiolitic complexes: Evidences for multiple origins of the H2-CH4-N2 gas mixtures.” Geochimica et Cosmochimica Acta (2017). https://www.sciencedirect.com/science/article/abs/pii/S0016703717307949

[23] “Hydrogen generation during serpentinisation in ophiolite complexes: A comparison of H2-rich gases from Oman, Philippines and Turkey.” EGU General Assembly abstract (2013). https://ui.adsabs.harvard.edu/abs/2013EGUGA..15.5553B/abstract

[24] (claim supported by ref [45]; original placeholder removed)

[25] “The principles of helium exploration.” Petroleum Geoscience (2021), Lyell Collection. https://www.lyellcollection.org/doi/full/10.1144/petgeo2021-029

[26] “Mantle helium in Sacramento basin natural gas wells.” Geochimica et Cosmochimica Acta (1986). https://www.sciencedirect.com/science/article/abs/pii/0016703786902310

[27] “Optimizing Helium Exploration: Noble Gas Isotopes as Probes for Migration–Enrichment Processes.” Energies 18(16), 4218 (2025). https://www.mdpi.com/1996-1073/18/16/4218 (cross-ref: Mass Spectrometry)

[28] Wiersberg, T. et al. (2007). “A helium isotope cross-section study through the San Andreas Fault at seismogenic depths.” Geochemistry, Geophysics, Geosystems 8(6). https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2006GC001388

[29] “The Use of Noble Gases to Constrain Subsurface Fluid Dynamics” and “Using Radiogenic Noble Gases to Trace the Conditions of Crustal Fluid Migration.” Frontiers in Earth Science (2022) / Procedia Earth and Planetary Science (ScienceDirect). https://www.frontiersin.org/journals/earth-science/articles/10.3389/feart.2022.895312/pdf; https://www.sciencedirect.com/science/article/pii/S1878522017300504

[30] USGS Groundwater Dating Laboratory, “3H/3He Dating Background.” https://water.usgs.gov/lab/3h3he/background/ (cross-ref: Mass Spectrometry)

[31] USGS Groundwater Dating Laboratory, “3H/3He Noble Gas Sampling.” https://water.usgs.gov/lab/3h3he/sampling/

[32] Eawag, “A Mass Spectrometric System for the Analysis of Noble Gases and Tritium in Water.” https://www.eawag.ch/fileadmin/Domain1/Abteilungen/wut/schwerpunkte/umweltisotope/EAWAG_02814.pdf

[33] Isotech Laboratories (Stratum Reservoir), “Gas Analyses” service page. https://isotechlabs.com/services/gas

[34] Stratum Reservoir, “The ALL NEW IsoTube: The Industry Standard for Gas Sampling.” https://www.stratumreservoir.com/wp-content/uploads/2024/08/SR_IsoTube_Brochure_082224.pdf

[35] UC Davis Stable Isotope Facility, homepage and sunset notice. https://stableisotopefacility.ucdavis.edu/

[36] USGS, “Analytical Services and Prices,” Reston Stable Isotope Laboratory. https://www.usgs.gov/labs/reston-stable-isotope-laboratory/analytical-services-and-prices

[37] Thermo Fisher Scientific, “253 Plus” IRMS brochure. https://documents.thermofisher.com/TFS-Assets/CMD/brochures/BR-30333-IRMS-253-Plus-BR30333-EN.pdf

[38] Thermo Fisher Scientific, “MAT 253: The Gold Standard for Isotope Ratio MS” brochure. https://documents.thermofisher.com/TFS-Assets/CMD/brochures/BR-30029-IRMS-MAT-253-BR30029-EN.pdf

[39] Thermo Fisher Scientific, “253 Ultra High Resolution Isotope Ratio MS” brochure. https://assets.thermofisher.com/TFS-Assets/CMD/brochures/BR-30293-IRMS-253-Ultra-BR30293-EN.pdf

[40] Elementar, “isoprime precisION stable isotope mass spectrometer” product page. https://www.elementar.com/en/products/stable-isotope-analyzers/mass-spectrometers/isoprime-precision

[41] “An Improved Cr-EA-IRMS Method for the Effective Determination of the Hydrogen Isotopes in Coal Samples.” Frontiers in Chemistry (2022), PMC9021440. https://pmc.ncbi.nlm.nih.gov/articles/PMC9021440/

[42] “Correction of H3+ Contributions in Hydrogen Isotope Ratio Monitoring Mass Spectrometry.” ResearchGate record. https://www.researchgate.net/publication/242363722_Correction_of_H3_Contributions_in_Hydrogen_Isotope_Ratio_Monitoring_Mass_Spectrometry

[43] “H2 Kinetic Isotope Fractionation Superimposed by Equilibrium Isotope Fractionation During Hydrogenase Activity of D. vulgaris Strain Miyazaki.” Frontiers in Microbiology (2019), PMC6636216. https://pmc.ncbi.nlm.nih.gov/articles/PMC6636216/ (supports the kinetic+equilibrium superposition claim; does not support the claim that kinetic D/H fractionation is “roughly two orders of magnitude larger than the corresponding carbon KIE” — that comparison does not appear in this source)

[44] “Recent Advances and Prospects in Design of Hydrogen Permeation Barrier Materials for Energy Applications—A Review.” PMC (2022), PMC9570512. https://pmc.ncbi.nlm.nih.gov/articles/PMC9570512/

[45] “The H2/CH4 ratio during serpentinization cannot reliably identify biological signatures.” PMC (2016), PMC5036043. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5036043/

[46] Sherwood Lollar, B., faculty profile, University of Toronto Department of Earth Sciences. https://www.es.utoronto.ca/people/directories/all-faculty/barbara-sherwood-lollar