Abstract Fe‐ and Mn‐oxides are common secondary minerals in faults, fractures, and veins and potentially record information about the timing of fluid movement through their host rocks. These phases are difficult to date by most radioisotopic techniques, but relatively high concentrations of U and Th make the (U‐Th)/He system a promising approach. We present new petrographic, geochronologic and thermochronologic analyses of secondary oxides and associated minerals from fault zones and fractures in southeastern Arizona. We use these phases in attempt to constrain the timing of fluid flow and their relationship to magmatic, tectonic, or other regional processes. In the shallowly exhumed Galiuro Mountains, Fe‐oxide (U‐Th)/He dates correspond to host‐rock crystallization and magmatic intrusions from ca. 1.6 to 1.1 Ga. Step‐heating 4He/3He experiments and polydomain diffusion modeling of 3He release spectra on these samples are consistent with a crystallite size control on He diffusivity, and little fractional loss of radiogenic He since formation in coarse‐grained hematite, but large losses from fine‐grained Mn‐oxide. In contrast to Proterozoic dates, Fe‐ and Mn‐oxides from the Catalina‐Rincon and Pinaleño metamorphic core complexes are exclusively Cenozoic, with dates clustering at ca. 24, 15, and 9 Ma, which represent distinct cooling or fluid‐flow episodes during punctuated periods of normal faulting. Finally, a subset of Fe‐oxides yield dates of ca. 5 Ma to 6 ka and display either pseudomorphic cubic forms consistent with oxidative retrogression of original pyrite or magnetite, or fine‐grained botryoidal morphologies that we interpret to represent approximate ages of recrystallization or pseudomorphic replacement at shallow depths.

In this paper, we use several analytical methods in an effort to better understand the systematics of the (U-Th) /He chronometer in hematite and manganese oxides. He-4 diffusion data from a polycrystalline hematite sample is consistent with diffusion from hematite crystals with a range of sizes similar to those directly observed in sample material. Combined with a compilation of hematite He diffusion data from previous studies, this supports the interpretation that in general crystal size is a primary control on He retentivity in hematite. He-4/He-3 diffusion data from a single fragment of a larger hematite crystal imply the presence of multiple diffusion domains smaller than the observed size of the crystal fragment, which may be related to cracks, inclusions, or other internal features, as well as higher concentrations of 4He in smaller domains. We use kinetic values determined in this and other studies and measurements of hematite crystal size in each dated sample to estimate approximate closure temperatures for each sample, most of which range from 140 to 240 degrees C. Relationships between minor element and parent nuclide concentrations and (U-Th)/He dates measured in aliquots of some analyzed hematite samples suggest that, in some cases, U and Th are concentrated in interstitial phases other than hematite. We identify two processes responsible for much of the dispersion in hematite dates from single samples. Removal of U- and Th-rich interstitial phases from analyzed material during sample preparation or analysis leaves behind unsupported 4He implanted in hematite crystals. This removal results in apparent (U-Th)/He dates artificially older than the time of hematite He closure. A smaller source of dispersion that is likely to still be significant in some samples can be attributed to the high He cliffusivities of observed interstitial phases, which do not retain implanted He. (U-Th)/He dating of hematite from upper and lower-plate rocks in the Buckskin and Rawhide detachment fault system of western Arizona yields ages that coincide with the timing of rapid extension along the detachment fault. Comparisons of estimated hematite closure temperatures and hematite sample ages to data from other studies of the Buckskin-Rawhide detachment system lead us to conclude that hematite dates record rapid cooling that followed detachment zone mineralization. Mn oxide (U-Th)/He dates are interpreted as formation ages of minerals formed by hydrothermal fluids that circulated through upper-plate rocks after cessation of extension along the detachment fault. Overall, we find that (U-Th)/He dating of hematite and Mn oxide minerals are promising methods for obtaining temporal information about the formation and cooling of these common secondary phases.

Oxide minerals in diagenetic cements, concretions, and fracture fill reflect episodes of ancient groundwater flow that have the potential to record tectonic, geomorphic, and climatic changes through time. To better under stand the ages, conditions of formation, and potential geologic significance of these diagenetic materials, we have measured (U-Th)/He ages and element concentrations in hematite, goethite, and Mn-oxide in Mesozoic sandstones from several locations in the Colorado Plateau. Most (U-Th)/He ages are Pliocene-Pleistocene, but some samples are as old as 25 Ma, the age of previously determined Ar-40/Ar-39 ages on Mn-oxide cement. In one region, texturally diverse Mn- and Fe-oxides yield relatively reproducible ages of ca. 2-3 Ma, and in another region, vitreous Fe-oxide fracture coatings yield ages of ca. 300 ka. Elsewhere, most cements and concretions show a wide range of ages among aliquots taken over millimeter length scales. Hematite-dominated samples show a broad positive correlation between age and U-Th concentrations, whereas most goethite-dominated samples show a negative correlation. Exterior portions of spherical goethite concretion rinds have higher U-Th concentrations and younger (U-Th)/He ages than interior portions. Taken together, the age-composition relationships of these samples suggest that wide ranges of ages in some samples largely reflect relatively recent (Pliocene-Pleistocene) U-Th addition, recrystallization, or later oxide growth that affected precursor cements that may have been oxides or other minerals. In some cases, these precursors may have formed as early as 25 Ma. In at least two areas, the (U-Th)/He dates (or in cases where samples show a wide range in ages, the minimum dates) are associated with major climate or local incision rate changes. We speculate that diagenetic oxides, and possibly other types of fine-grained secondary oxide minerals, remain open to U-Th uptake, recrystallization, or continued growth through contact with groundwater long after initial formation. This may provide opportunities to understand ground water compositions and flow regimes as samples are exhumed through the shallow crust, or as surface conditions change through time in diagenetic systems in the critical zone of Earth's crust and on Mars.