Siderophile-Enriched Template: Geochemical Insights

Updated 4 July 2026

The siderophile-enriched template is defined by enhanced concentrations of Fe, Ni, Cr, and Mn relative to Ca, serving as a key compositional construct in polluted white dwarf analyses and planetary geochemistry.
It is constructed using calibrated meteorite whole-rock quantiles and Bayesian inference, with discriminating element panels (e.g., Cr, Fe, Mg, Ni, Ti) that optimize detectability against natural abundance models.
In planetary geochemistry, the template elucidates processes such as core formation, late accretion, and HSE retention, acting as a chrono-geochemical operator for Earth, Moon, and Mars.

“Siderophile-enriched template” denotes a family of compositional constructs in which siderophile concentration is treated as the defining signal of a reservoir, endmember, or process history. In one precise contemporary usage, it is a fixed processed-composition endmember for polluted white dwarf abundance inference: a Ca-normalized, metal-rich, silicate-poor composition with enhanced Fe, Ni, Cr, and Mn, used only as a detectability probe against a meteorite-trained natural reference (Huang et al., 28 May 2026). In planetary geochemistry and cosmochemistry, the same expression is naturally extended to templates for metal-rich or HSE-bearing reservoirs produced by core formation, sulfide segregation, late accretion, nebular transport, or meteoritic condensation and processing (Brasser et al., 2021, Yoshizaki et al., 2018). This suggests that the term is best treated as a class of chemically and dynamically motivated endmembers rather than a single universally valid composition.

1. Fiducial processed endmember

In the polluted-white-dwarf literature, the fiducial siderophile-enriched template is defined in Ca-referenced log mass-ratio space,

$r_Z \equiv \log_{10}\!\left(\frac{m_Z}{m_{\rm Ca}}\right),$

and is described as “a metal-rich, silicate-poor component with enhanced Fe, Ni, Cr, and Mn.” The template is “best viewed as a deliberately stylized endmember used to evaluate detectability rather than as a unique prediction for technology.” Its values are constructed directly from meteorite whole-rock quantiles: enriched elements are set to the 99th percentile plus $0.6$–$1.0$ dex, depleted elements to the 5th percentile, and neutral elements to the median (Huang et al., 28 May 2026).

Element $Z$	$r_Z=\log_{10}(m_Z/m_{\rm Ca})$	Qualitative role
Al	$-0.364$	suppressed
Cr	$0.788$	enhanced
Fe	$2.359$	enhanced
Mg	$-0.401$	suppressed
Mn	$0.473$	enhanced
Na	$0.6$0	suppressed
Ni	$0.6$1	enhanced
P	$0.6$2	optional factor
S	$0.6$3	optional factor
Si	$0.6$4	reduced but not absent
Ti	$0.6$5	suppressed

For the main multivariate tier $0.6$6, the core inference elements are $0.6$7, with P and S treated separately as optional univariate factors. Mixing is performed in linear ratio space, not in log space: $0.6$8 with $0.6$9 the Ca-normalized mixing fraction. The template is therefore chemically explicit, but epistemically non-unique.

2. Bayesian inference, detectability, and discriminating element panels

The comparison baseline for this template is a multivariate natural-composition reference trained on 3,493 whole-rock meteorite analyses. That reference is modeled as a broad-group Gaussian mixture over chondrite, achondrite, and “other,” evaluated only on the observed abundance subspace for each record. Record-level support is quantified by

$1.0$0

with $1.0$1 the natural-only model and $1.0$2 the natural-plus-template mixture with $1.0$3 (Huang et al., 28 May 2026).

Dataset	Strong-support counts	$1.0$4
atm	$1.0$5 with $1.0$6; $1.0$7 with $1.0$8	$1.0$9
acc_ss	$Z$ 0 with $Z$ 1; $Z$ 2 with $Z$ 3	$Z$ 4

Support for the template is therefore uncommon. In the full photospheric compilation, 50/697 records have $Z$ 5, 8/697 have $Z$ 6, and 4/697 have $Z$ 7; in the diffusion-adjusted steady-state subset, 25/148 have $Z$ 8, 6/148 have $Z$ 9, and 1/148 has $r_Z=\log_{10}(m_Z/m_{\rm Ca})$ 0. The posterior medians of the detectable-incidence parameter are $r_Z=\log_{10}(m_Z/m_{\rm Ca})$ 1 for atm and $r_Z=\log_{10}(m_Z/m_{\rm Ca})$ 2 for acc_ss, with the paper stressing that this is a record-level detectable-incidence parameter, not a unique-star occurrence rate and not necessarily a physical processed-material incidence.

Injection–recovery calibration shows that discrimination is driven mainly by chemical information and typically requires $r_Z=\log_{10}(m_Z/m_{\rm Ca})$ 3 detected elements for decisive support. For this particular template, the strongest exact five-element panels repeatedly include Fe, Mg, Cr, and Ti together with Ni, Si, or Na. The highest-power five-element panel in both atm and acc_ss is Cr, Fe, Mg, Ni, Ti, with mean power $r_Z=\log_{10}(m_Z/m_{\rm Ca})$ 4 and $r_Z=\log_{10}(m_Z/m_{\rm Ca})$ 5, respectively, at $r_Z=\log_{10}(m_Z/m_{\rm Ca})$ 6. This establishes a concrete observational signature: the template is not merely “metal-rich,” but statistically most identifiable when metal enrichment is co-constrained against suppressed Mg and Ti and one additional discriminator.

3. Geochemical basis of siderophile enrichment

The planetary-geochemical usage of a siderophile-enriched template rests on the behavior of the highly siderophile elements

$r_Z=\log_{10}(m_Z/m_{\rm Ca})$ 7

which partition strongly into metallic cores during metal-silicate segregation. Silicate mantles should therefore be extremely depleted in HSEs after primary differentiation; yet the mantles of Earth, Moon, Mars, and Vesta contain HSEs in approximately chondritic relative proportions and in abundances too high to be explained by equilibrium residual silicate partitioning alone, making them standard tracers of late accretion and retention timing (Brasser et al., 2021).

For Earth, the choice of reference composition is itself a template problem. Measurement and modeling of refractory lithophile element ratios in enstatite chondrites identify non-carbonaceous inner-solar-system material, especially enstatite chondrites, as the closest known starting materials for the bulk of the silicate Earth and the core. In that framework, the Bulk Silicate Earth hosts Earth’s inventory of Ti, Zr, Nb, and Ta, but not the full complement of V; the BSE has $r_Z=\log_{10}(m_Z/m_{\rm Ca})$ 8, while primitive enstatite chondrites have $r_Z=\log_{10}(m_Z/m_{\rm Ca})$ 9 (Yoshizaki et al., 2018). A siderophile-enriched template for Earth therefore cannot be constructed simply by subtracting a CI-like bulk from the silicate Earth; the cosmochemical baseline must be NC-like rather than CC-like.

The same logic extends beyond the canonical HSE suite. Zn, long treated as effectively lithophile in some mass-balance arguments, is experimentally slightly siderophile under relevant core-formation conditions. Using metal-silicate partitioning calibrated at $-0.364$ 0 GPa and $-0.364$ 1– $-0.364$ 2 K, one study estimated $-0.364$ 3 ppm Zn in Earth’s core and $-0.364$ 4 ppm Zn in the bulk Earth, with a geochemical upper bound of $-0.364$ 5 S for the core under chondritic $-0.364$ 6 assumptions (Mahan et al., 2016). For W and Mo, liquid metal-liquid silicate partitioning is strongly affected by redox and metallic composition: W is $-0.364$ 7, Mo is $-0.364$ 8, and both become more siderophile with increasing C content of the metal, leading to the conclusion that successful Earth models require early accreting material to be sulfur-depleted and carbon-enriched (Jennings et al., 2022).

4. Natural cosmochemical realizations

Meteorites provide explicit natural analogues for siderophile-enriched reservoirs. In CR chondrite metal, interior, margin, and isolated grains show very similar Ni-normalized trace-element patterns. Refractory siderophile elements Os, Ir, Ru, Rh, and Pt are approximately unfractionated relative to Ni, whereas volatile siderophile/chalcophile elements Au, Cu, Ag, and S are depleted, typically by about an order of magnitude. The least processed margin and isolated grains have Ni contents of 5.36 and 5.48 wt% and Co/Ni mass ratios of $-0.364$ 9 and $0.788$0, close to the solar value $0.788$1; the expected Ni content of a solar metallic condensate is 5.7 wt% (Jacquet et al., 2015). This is a natural siderophile-enriched template in Ni-normalized, CI-relative space: a refractory plateau plus a volatility-depletion tail.

Primary sulfides in CM2 and CR2 chondrites add a complementary metal-sulfide template. Coordinated FIB, EMPA, SXRF, and TEM analyses of pyrrhotite-pentlandite intergrowths and sulfide-rimmed metal show that CM and CR PPI sulfides have similar trace-element patterns, supporting a common formation mechanism. Ni, Co, Cu, and Se are enriched in characteristic ways, Ge is depleted, and Zn is redistributed according to mixed lithophile/chalcophile behavior; SRM sulfide and associated metal have comparable trace-element patterns, consistent with a genetic relationship such as sulfidization of metal (Singerling et al., 2021). The result is not simply “metal versus sulfide,” but a phase-specific partitioning system with predictable enrichments and depletions.

Nebular transport models provide a disk-scale realization of the same idea. A one-dimensional PSN model with dust/vapor transport, sublimation, and recondensation of ferrosilite, enstatite, fayalite, forsterite, FeS, metal iron, and nickel shows that solids are concentrated at rocklines and that the vicinity of the first alloy+sulfide rockline cluster can become Fe-rich relative to the protosolar refractory baseline. At $0.788$2 yr and 0.67 AU, near the FeS rockline, the modeled inner nebula reaches 56 wt% Fe in case A and 58 wt% Fe in case B; the maximum Fe abundance near rocklines is 62 wt%, compared with a protosolar value of 47 wt% (Aguichine et al., 2020). A plausible implication is that rocklines generate localized Fe-Ni-S-enriched precursor reservoirs even without invoking later planetary differentiation.

5. Planetary differentiation, late accretion, and retention chronologies

In lunar chronology, the principal question is not whether HSEs were added, but when they became reliably retained in the mantle. Monte Carlo impact simulations coupled to updated dynamical models of leftover planetesimals and the E-belt yield a preferred lunar HSE retention age of ca. 4450 Ma, implying that modeled lunar mantle HSE abundances trace almost all of lunar late accretion. In the same framework, the most heavily cratered lunar highlands imply surface ages of at least 4370 Ma, and the best-fit leftover planetesimal mass with $0.788$3 km at 4500 Ma is approximately $0.788$4, with the E-belt fixed at $0.788$5 (Brasser et al., 2021). The lunar template is therefore a retention-age template: HSE abundance is interpreted relative to closure, not simply impact flux.

For Earth, mantle HSEs have been used as a clock for the Moon-forming impact. Using a mantle mass $0.788$6 and mantle/chondrite HSE ratios for Re, Os, Ir, Ru, Pt, and Pd, the required chondritic late-accreted mass is $0.788$7. Coupling that mass to $0.788$8-body simulations of late accretion yields a preferred Moon-formation age of $0.788$9 Myr after condensation, while a Moon-forming event at or before 40 Myr is ruled out at 99.9% confidence in the nominal case (Jacobson et al., 2015). In this usage, a siderophile-enriched template becomes a chrono-geochemical operator: present mantle HSEs map to post-last-giant-impact accreted mass, then to impact timing.

Mars adds a stochastic-impact variant. Martian mantle HSE abundances imply a late accretion supplement of about 0.8 wt.%, and stochastic Monte Carlo accretion under a shallow remnant planetesimal size-frequency distribution requires an impactor of at least 1200 km in diameter to have struck Mars before ca. 4430 Ma. The paper further argues that the hemispheric dichotomy is a plausible visible remnant of that impact and that ejected debris could have supplied the source material for Phobos and Deimos (Brasser et al., 2017). The martian template is thus impact-dominant rather than smoothly cumulative.

A distinct Earth late-veneer model pushes the same logic to the mechanical fate of impactor metal. SPH simulations of a differentiated $2.359$0, $2.359$1 km impactor striking Earth at $2.359$2 show that the statistically most likely $2.359$3 collision shock-melts and elongates the impactor core, fragments it into a metallic hail with characteristic size $2.359$4 m, and leaves about 60 wt% of the impactor’s core available to react with water. In this model, post-impact oxidation of small iron fragments retains HSEs in silicate reservoirs instead of allowing them to merge with Earth’s core (Genda et al., 2017).

6. Competing interpretations and limitations

The white-dwarf template is explicitly non-unique. It is a stylized fixed endmember, not a unique industrial product; natural planetary processes can produce fractionated, metal-rich compositions; sparse and censored abundance panels can yield suggestive but weakly diagnostic Bayes factors; and the natural meteorite-trained reference may not exhaust extrasolar rocky diversity (Huang et al., 28 May 2026). Accordingly, support for the template is a model-comparison statement, not a direct identification of technological material.

Planetary differentiation models likewise reject any single universal interpretation of siderophile enrichment. A major alternative to the classical “complete HSE stripping by metal segregation followed by late veneer” is the Hadean matte model: metal-silicate equilibration during Earth’s core formation may actually have increased mantle HSE concentrations because HSE partition coefficients were relatively low at the high pressures of core formation, while subsequent exsolution and segregation of iron sulfide liquid stripped magma oceans of HSEs before late accretion, yielding slightly suprachondritic Pd/Ir and Ru/Ir (Kelkar, 2016). In a later SPH-constrained accretion/differentiation model, successful Earth simulations reproduced BSE W and Mo only when equilibration pressure was fixed by impact-induced melt geometry and planetesimal-delivered material was reprocessed in global magma oceans; even then, HSEs remained too abundant unless FeS exsolution was invoked (Dale et al., 2023).

A further complication is that mantle HSEs may under-record late accretion if impactor metal continues to segregate after impact. An analytical treatment of metal entrainment in magma ponds and solid mantle argued that metals from impactors larger than approximately 1 km generally sink to Earth’s core, leaving no HSE signature in the mantle. In that framework, avoiding a “mass accretion catastrophe” requires either efficient disruption of impactor core material into $2.359$5 mm fragments or a substantial oxidized, carbonaceous-chondrite-like component in the late veneer (Anslow et al., 18 Mar 2026). The broader implication is narrow but consequential: a siderophile-enriched template is always conditional on retention physics. Whether in white-dwarf abundance inference or planetary differentiation, the template is informative only when the pathway from delivery to preserved signal is specified.