Subterranean production of 39Ar and implications for Doe Canyon well gas

Ondřej Šrámek 1 and William F. McDonough 2,3

1 Department of Geophysics, Charles University, Prague, Czech Republic, [email protected] 2 Department of Geology, University of Maryland, College Park, Maryland USA, [email protected] 3 Research Center of Neutrino Sciences and the Department of Earth Sciences, Tohoku University, Japan

Presented at the Low-Radioactivity Underground Argon workshop at Pacific Northwest National Laboratory in Richland, Washington USA, on 20 March 2018 36Ar … primordial, stable 38Ar … primordial, stable Argon 40Ar … radiogenic, stable 37Ar … radioactive, t1/2 = 35 d 39Ar … radioactive, t1/2 = 269 y 42Ar … radioactive, t1/2 = 32.9 y Atmosphere Argon the third most abundant gas (~1%) 40Ar from degassing of Earth (decay of 40K) 39Ar produced cosmogenically: 40Ar(n,2n)39Ar 39Ar/Ar = 8 × 10-16 ➝ ~1 decay per sec per kg 40Ar/36Ar = 295

Underground 40Ar produced by electron capture on 40K 39Ar nucleogenic production

2 Interest in 39Ar

• Particle physics – dark matter searches … seeking low radioactivity argon as target for WIMP detection

• Hydrology – environmental radioactive tracer … timescales and pathways of ground-water transport

• Geophysics – underground 39Ar of nucleogenic (and not cosmogenic) origin carries signature of source rock composition, esp. concentration in K, Th, U

3 Geophysical interest

• O, Fe, Si, Mg make up 93% of Earth’s mass PeriodicChemistry Table of of the Elements Earth 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

1 1 Atomic # 2 2 K 1 H Symbol He Hydrogen Name Solid Metals Nonmetals Helium 1.00794 Atomic Mass C 4.002602 Alkali metals metals earth Alkaline metals Transition Poor metals nonmetals Other Noble gases 3 2 4 2 5 2 6 2 7 2 8 2 9 2 10 2 K 1 2 Hg Liquid Lanthanoids 3 4 5 6 7 8 L 2 Li Be B C N O F Ne Lithium Beryllium Gas Boron Carbon Nitrogen Oxygen Fluorine Neon ➝ 6.941 9.012182 H 10.811 12.0107 14.0067 15.9994 18.9984032 20.1797 + Al, Ca, Ni 98% of Earth’s mass 2 2 2 2 2 2 2 2 K Actinoids 11 8 12 8 13 8 14 8 15 8 16 8 17 8 18 8 L

1 2 Unknown 3 4 5 6 7 8 M

Rf • 3 Na Mg Al Si P S Cl Ar Sodium Magnesium Aluminium Silicon Phosphorus Sulfur Chlorine Argon 22.98976928 24.3050 26.9815386 28.0855 30.973762 32.065 35.453 39.948

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 K 19 8 20 8 21 8 22 8 23 8 24 8 25 8 26 8 27 8 28 8 29 8 30 8 31 8 32 8 33 8 34 8 35 8 36 8 L 8 8 9 10 11 13 13 14 15 16 18 18 18 18 18 18 18 18 M 4 K 1 Ca 2 Sc 2 Ti 2 V 2 Cr 1 Mn 2 Fe 2 Co 2 Ni 2 Cu 1 Zn 2 Ga 3 Ge 4 As 5 Se 6 Br 7 Kr 8 N Potassium Calcium Scandium Titanium Vanadium Chromium Manganese Iron Cobalt Nickel Copper Zinc Gallium Germanium Arsenic Selenium Bromine Krypton 39.0983 40.078 44.955912 47.867 50.9415 51.9961 54.938045 55.845 58.933195 58.6934 63.546 65.38 69.723 72.64 74.92160 78.96 79.904 83.798

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 K 37 8 38 8 39 8 40 8 41 8 42 8 43 8 44 8 45 8 46 8 47 8 48 8 49 8 50 8 51 8 52 8 53 8 54 8 L 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 M 8 8 9 10 12 13 14 15 16 18 18 18 18 18 18 18 18 18 N 5 Rb 1 Sr 2 Y 2 Zr 2 Nb 1 Mo 1 Tc 1 Ru 1 Rh 1 Pd 0 Ag 1 Cd 2 In 3 Sn 4 Sb 5 Te 6 I 7 Xe 8 O Rubidium Strontium Yttrium Zirconium Niobium Molybdenum Technetium Ruthenium Rhodium Palladium Silver Cadmium Indium Tin Antimony Tellurium Iodine Xenon 85.4678 87.62 88.90585 91.224 92.90638 95.96 (97.9072) 101.07 102.90550 106.42 107.8682 112.411 114.818 118.710 121.760 127.60 126.90447 131.293

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 K 55 8 56 8 72 8 73 8 74 8 75 8 76 8 77 8 78 8 79 8 80 8 81 8 82 8 83 8 84 8 85 8 86 8 L 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 M 18 18 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 N 6 Cs 8 Ba 8 57–71 Hf 10 Ta 11 W 12 Re 13 Os 14 Ir 15 Pt 17 Au 18 Hg 18 Tl 18 Pb 18 Bi 18 Po 18 At 18 Rn 18 O Caesium 1 Barium 2 Hafnium 2 Tantalum 2 Tungsten 2 Rhenium 2 Osmium 2 Iridium 2 Platinum 1 Gold 1 Mercury 2 Thallium 3 Lead 4 Bismuth 5 Polonium 6 Astatine 7 Radon 8 P −8 132.9054519 137.327 178.49 180.94788 183.84 186.207 190.23 192.217 195.084 196.966569 200.59 204.3833 207.2 208.98040 (208.9824) (209.9871) (222.0176) 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 K 87 8 88 8 104 8 105 8 106 8 107 8 108 8 109 8 110 8 111 8 112 8 113 8 114 8 115 8 116 8 117 118 8 L 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 M tens ppb of U (~10 ) 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 N 89–103 7 Fr 18 Ra 18 Rf 32 Db 32 Sg 32 Bh 32 Hs 32 Mt 32 Ds 32 Rg 32 Uub 32 Uut 32 Uuq 32 Uup 32 Uuh 32 Uus Uuo 32 O Francium 8 Radium 8 Ru herfordium 10 Dubnium 11 Seaborgium 12 Bohrium 13 Hassium 14 Meitnerium 15 Darmstadtium 17 Roentgenium 18 Ununbium 18 Ununtrium 18 Ununquadium 18 Ununpentium 18 Ununhexium 18 Ununseptium Ununoctium 18 P • (223) 1 (226) 2 (261) 2 (262) 2 (266) 2 (264) 2 (277) 2 (268) 2 (271) 1 (272) 1 (285) 2 (284) 3 (289) 4 (288) 5 (292) 6 (294) 8 Q

For elements with no stable isotopes, the mass number of the isotope with the longest half-life is in parentheses.

Periodic Table Design and Interface Copyright © 1997 Michael Dayah. http://www.ptable.com/ Last updated: May 27, 2008

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 57 8 58 8 59 8 60 8 61 8 62 8 63 8 64 8 65 8 66 8 67 8 68 8 69 8 70 8 71 8 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 19 21 22 23 24 25 25 27 28 29 30 31 32 32 La 9 Ce 9 Pr 8 Nd 8 Pm 8 Sm 8 Eu 8 Gd 9 Tb 8 Dy 8 Ho 8 Er 8 Tm 8 Yb 8 Lu 9 four times as much Th (Th/U ~ 4) Lanthanum 2 Cerium 2 Praseodymium 2 Neodymium 2 Promethium 2 Samarium 2 Europium 2 Gadolinium 2 Terbium 2 Dysprosium 2 Holmium 2 Erbium 2 Thulium 2 Ytterbium 2 Lutetium 2 138.90547 140.116 140.90765 144.242 (145) 150.36 151.964 157.25 158.92535 162.500 164.93032 167.259 168.93421 173.054 174.9668

• 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 89 8 90 8 91 8 92 8 93 8 94 8 95 8 96 8 97 8 98 8 99 8 100 8 101 8 102 8 103 8 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 Ac 18 Th 18 Pa 20 U 21 Np 22 Pu 24 Am 25 Cm 25 Bk 27 Cf 28 Es 29 Fm 30 Md 31 No 32 Lr 32 Actinium 9 Thorium 10 Protactinium 9 Uranium 9 Neptunium 9 Plutonium 8 Americium 8 Curium 9 Berkelium 8 Californium 8 Einsteinium 8 Fermium 8 Mendelevium 8 Nobelium 8 Lawrencium 9 (227) 2 232.03806 2 231.03588 2 238.02891 2 (237) 2 (244) 2 (243) 2 (247) 2 (247) 2 (251) 2 (252) 2 (257) 2 (258) 2 (259) 2 (262) 2 • a few hundred ppm K (K/U ~ 104) Michael Dayah For a fully interactive experience, visit www.ptable.com. [email protected] • Long-lived radionuclides 238U, 232Th, 40K account for >99% of radiogenic heat produced in the Earth

• A factor of 3 uncertainty in the amount of U, Th, K in the Earth… • How much power available to power plate tectonics, mantle convection?? Energy balance of the planet?? Sidenote: also GEONEUTRINOS Dynamics and thermal evolution of the Earth

4 Underground production of noble gases

radiogenic α particles

232Th radioactive 208Pb 6 4 decay Th, U 235U 207Pb ++7 4 ()e−+ ν¯e 238U 206Pb 8 α 6 4He nucleogenic neutrons 27Al 23Na 30P 26Al (α,n) 29Si 30Si + + 32S 33S 18O … α n 21Ne … 21Ne fissiogenic neutrons ~90% ~10% spontaneous fission 238U ++~2 n

nucleogenic 39Ar

(n,p) 39 + + Ar K n p 39K 39Ar 5 Calculating underground 39Ar production radiogenic α particles

232Th radioactive 208Pb 6 4 decay 235U 207Pb ++7 4 ()e−+ ν¯e • Natural α particle energy spectrum: 238U 206Pb 8 α 6 nucleogenic neutrons 27Al 23Na 30P 26Al (α,n) decay chains and spontaneous fission, 29Si 30Si + + 32S 33S 18O … α n 21Ne …

inputs from NuDat (www.nndc.bnl.gov/nudat2) fissiogenic neutrons ~90% ~10% spontaneous fission 238U ++~2 n • Slowing and stopping of α particles: nucleogenic 39Ar (n,p) + + n p stopping power from SRIM-2013.00 (www.srim.org) 39K 39Ar

• Fast neutrons from (α,n) reactions on light targets: (α,n) cross sections from TALYS version 1.6 (www.talys.eu)

• 39Ar from 39K(n,p)39Ar: using MCNP6, a general-purpose Monte Carlo N-Particle transport code (mcnp.lanl.gov)

Šrámek, Stevens, McDonough, Mukhopadhyay, Peterson 2017 Geochim.Cosmochim.Acta doi:10.1016/j.gca.2016.09.040 (arXiv:1509.07436) 6 238U 238U Decay Chain 4.468 Gy

Q [keV] 4270 Nucl β Natural 2195 Branching α 6.70 h Half-life fraction 234 0.0016 234 Th 273 234 U [keV] Pa 2195 Q α 24.10 d 1.159 m 245.5 ky 0.9984 Branching fraction 456789 Data from NNDC (http://www.nndc.bnl.gov/) 4860 production α 1 mean energy 230Th 232Th 75.38 ky 0.1 4770 238 0.01 226Ra U decay chain 1600 y 0.001 4871 222Rn 1 235U 3.824 d 0.1 5590

218 260 218 2881 218 intensity 0.01

Po At Rnα 0.00020 0.001 3.098 m 1.5 s 0.035 s 0.001 6115 6874 7263 0.99980 0.999 214Pb 1019 214Bi 3270 214Po 1 238U 0.99979 26.8 m 19.9 m 164.3 μs 0.1 5617 7833 0.00021 210Tl 5490 210Pb 63.5 210Bi 1162 210Po 0.01 ~1 ~1 1.30 m 22.20 y 5.012 d 138.4 d

3792 5036 5407 0.001 1.9e-8 1.32e-6 456789 206 206 206 Hg 1308 Tl 1532 Pb Energy in MeV 8.32 m 4.202 m STABLE

80 81 82 83 84 85 86 88 90 91 92

7 232Th 232Th Decay Chain 14.05 Gy

Q [keV] 4083 Nucl β Branching Half-life fraction 228Ra 45.9 228Ac 2127 228Th [keV] Q α 5.75 y 6.15 h 1.912 y Branching fraction Data from NNDC (http://www.nndc.bnl.gov/) 5520 224 energy spectrum Ra α 3.66 d

5789 220Rn 456789 55.6 s 6405 mean α energy 216Po 1 0.145 s 232Th 6906

212 212 212 232 mean 6.0 MeV Pb 569.9 Bi 2252 Po 0.6406 Th decay chain 10.64 h 60.55 m 299 ns

6207 8954 0.3594 0.1 208Tl 4999 208Pb 6 α particles 3.053 m STABLE max 81 82 83 84 86 88 89 90 0.01 8.78 MeV 212 235U Po 235U Decay Chain 703.8 My

Q [keV] 4679 Nucl β Branching 0.001 Half-life fraction 231Th 390 231Pa [keV] Q α 25.52 h 32.8 ky Branching fraction Data from NNDC (http://www.nndc.bnl.gov/) 5150 227Ac 44.8 227Th 0.9862 21.772 y 18.68 1 235 5042 6146 0.0138 U 223Fr 1149 223Ra 6.0 MeV 0.99994 22.00 m 11.43 d

5430 5979 0.00006 219At 1566 219Rn 0.03 56 s 3.96 s 0.1

6390 6946 0.97 215Bi 2250 215Po 715 215At 2.3e-6 7.6 m 1.781 ms 0.10 ms

7526 8178 235 ~1 U: decay chain 211Pb 1367 211Bi 574 211Po intensity 0.00276 0.01 36.1 m 2.14 m 0.516 s α 6750 7595 0.99724 207Tl 1418 207Pb 7 α particles 4.77 m STABLE 81 82 83 84 85 86 87 88 89 90 91 92 0.001

238U 238U Decay Chain 4.468 Gy 1 238 Q [keV] 4270 Nucl β Branching 2195 U 6.70 h 5.4 MeV Half-life fraction 234 0.0016 234 Th 273 234 U [keV] Pa 2195 Q α 24.10 d 1.159 m 245.5 ky 0.9984 Branching fraction Data from NNDC (http://www.nndc.bnl.gov/) 4860 230Th 75.38 ky 0.1 4770 226Ra 1600 y

4871 222Rn 3.824 d

5590 0.01

218Po 260 218At 2881 218Rn 0.00020 0.001 3.098 m 1.5 s 0.035 s

6115 6874 7263 0.99980 0.999 214Pb 1019 214Bi 3270 214Po 0.99979 26.8 m 19.9 m 164.3 μs 238 5617 7833 0.00021 U decay chain 0.001 210Tl 5490 210Pb 63.5 210Bi 1162 210Po ~1 ~1 1.30 m 22.20 y 5.012 d 138.4 d 456789 3792 5036 5407 1.9e-8 1.32e-6 8 α particles 206Hg 1308 206Tl 1532 206Pb 8.32 m 4.202 m STABLE 8 80 81 82 83 84 85 86 88 90 91 92 Energy in MeV Travel distance of α particle in rock

Stopping power from SRIM-2013.00

1600 50 Stopping power − 1 1400 Range g 2 40 1200 Upper Cont. Crust m

1000 µ 30

800 Depl. Mantle 20 600 range in α 400 10 Range for mean 200

Stopping power in MeV cm α energy ~ 25 μm 0 0 0123456789 α energy in MeV 9 Table 3: CalculatedTable production 3: Calculated rates productionS ↵ of ↵ particles, rates S ↵ asof number↵ particles, of ↵ as’s numberproduced of per↵’s second produced in 1 per kilogram second of in rock, 1 kilogram in a particular of rock, decay in a particular chain decay chain (columns 2–4)(columns and the total 2–4) (column and the 5). total (column 5). CompositionComposition 232Th 235U232Th 238235U U Total238U Total Upper ContinentalUpper Continental Crust 256 Crust 10.9256 267 10.9533 267 533 Middle ContinentalMiddle Continental Crust 158 Crust 5.24158 128 5.24292 128 292 Lower ContinentalLower Continental Crust 29.2 Crust 0.80629.2 19.8 0.80649.8 19.8 49.8 Bulk OceanicBulk Crust Oceanic Crust5.11 0.2825.11 6.91 0.28212.3 6.91 12.3 Depleted MantleDepleted Mantle0.334 0.01900.334 0.464 0.01900.817 0.464 0.817

27Al 23Na 30P 26Al (α,n) 29Si 30Si + + 32S 33S 18O … ↵ n 21Ne … Choice based on: (α,n) targets • natural abundance • (α,n) cross section Table 4: TargetTable and product 4: Target nuclides, and productQ values, nuclides, thresholdQ values, energies thresholdEth (eqn. energies8), andEth Coulomb(eqn. 8), barriers and CoulombVC (eqn. barriers7) of (V↵C, n(eqn.) reactions.7) of (↵ Also, n) reactions.Q and Also Q and Eth of neutronE inducedth of neutron reactions. induced In cases reactions. where In the cases product where nuclide the product is unstable nuclide⇤, we is show unstable the⇤, final we show stable the nuclide final as stable well. nuclide Energies as well. in MeV. Energies in MeV. Nuclide rest massesNuclide for rest calculation masses forof Q calculationand Eth taken of Q fromand ENISTth taken(2016 from). NIST (2016). Energy threshold Target ReactionTarget Reaction Product ProductQEthQEVC th VC of endothermic reactions 27Al (↵27, nAl) 30 (↵P,⇤n) 30Si30P⇤ 2.642530Si 3.03452.6425 6.8012 3.0345 6.8012 ! ! 23Na (↵23, nNa) 26 (Al↵,⇤n) 26Mg26Al⇤ 2.965926Mg 3.48232.9659 5.9577 3.4823 5.9577 ! ! 29Si (↵29, nSi) (↵, n32)S 321S.5258 1.73651.5258 7.2107 1.7365 7.2107 30Si (↵30, nSi) (↵, n33)S 333S.4933 3.95983.4933 7.1571 3.9598 7.1571 Coulomb barrier 18O(↵,18nO() ↵,21n)Ne 21Ne0.6961 0.8510.6961 4.5626 0.851 4.5626EM repulsion between α & target 26Mg (↵26, nMg) (↵, n29)Si29 0Si.0341 0. –0341 6.3298 – 6.3298Restrictive for high Z targets. 25Mg (↵25, nMg) (↵, n28)Si28 2Si.6536 2. –6536 6.3838 – 6.3838 19F(↵,19nF() 22Na↵,⇤n) 22Ne22Na⇤ 1.951322Ne 2.36241.9513 5.0754 2.3624 5.0754 ! ! 17O(↵,17nO() ↵,20n)Ne20 0Ne.5867 0. –5867 4.6168 – 4.6168 56Fe (↵56, nFe) 59 (Ni↵,⇤n) 59Co59Ni⇤ 5.096159Co 5.46075.0961 11.5272 5.4607 11.5272 ! ! 41K(↵,41nK() 44Sc↵,⇤n) 44Ca44Sc⇤ 3.389444Ca 3.72063.3894 9.0555 3.7206 9.0555 ! ! 48Ti (↵48, nTi) 51 (↵Cr, ⇤n) 51V51Cr⇤ 2.685351V 2.90942.6853 10.1118 2.9094 10.1118 ! ! ⇒ Strong energy dependence 13C(↵,13nC() ↵, n16)O216O2.2156. –2156 3.656 – 3.656 44Ca (↵44, nCa) (↵, 47n)Ti 472Ti.1825 2.38122.1825 9.3791 2.3812 9.3791of (α,n) cross sections 10 39K(n,39pK() n, 39p)Ar39 0Ar.2176 0. –2176 n/a – n/a 24Mg (n24, ↵Mg) (n, 21↵)Ne 21Ne2.5554 2.66282.5554 2.6628 n/a n/a

16 16 Cross section of (α,n) reaction

Calculated using TALYS version 1.6

27Al Calculated with TALYS v1.6 600 18 23Na O 29Si 26Mg 500 30Si 18O Only cross section 26Mg comparison 400 25Mg 19F Does not account for 300 17O 23Na targets' natural 56Fe abundance 41K 200 48Ti 13C 27Al Cross section in mbarn 100 44Ca

0 123456789 α energy in MeV 11 (α,n) neutron production

Thick target neutron production function Probability that α particle of an initial energy participates in (α,n)

1.5e−06 27Al Uses Upper Continental Crust rock composition Atomic density of nuclide in rock

1e−06 Accounts for rock 23Na composition ,n) reaction α

5e−07 29Si 30Si

Probability of ( 18O 26Mg 0 456789 Initial α energy in MeV 12 Neutron yield and 0123456 8e−07 232 27Al Th 23 neutron energy 6e−07 Na 29Si 30 Following α-decays 4e−07 Si down the decay chain 18O 26 − 1 2e−07 Mg Neutron yield Neutrons per decay of 1 atom 0 of parent radionuclide 8e−07 235U 6e−07

4e−07 branching α intensity 2e−07

Differential neutron 0 production function 8e−07 238U

Differential yield spectrum in MeV 6e−07

4e−07 spontaneous fission

2e−07

Neutron yield spectrum 0 i 0123456 dY↵,n = ... Neutron energy in MeV dEn 13 39K(n,p)39Ar

exothermic reaction Q = 217.6 keV Neutron yield spectrum + rock composition 39K is 93.3% of natural potassium, 0123456

8e−07 232 there is ~2 wt% of K in shallow 27Al Th 23 6e−07 Na continental crust (global average) 29Si 30 4e−07 Si 18O 26 − 1 2e−07 Mg

0

8e−07 235U 6e−07 39Ar yield per decay 4e−07

2e−07 of 1 atom of 232 235 238 0 Th / U / U 8e−07 238U

Differential yield spectrum in MeV 6e−07

4e−07 spontaneous fission

2e−07

0 0123456 Neutron energy in MeV 14 Results Calculated with rock composition of Upper Continental Crust Table 5 S O. 378 Calculated results for neutron yield Y, neutron production rate S , 39 Ar production rate S , and 2139Ne production rate by n; a reaction Sn;a . 21Ne production rate by a; n reaction is equal to Neutron yield per Neutronn production rate39Ar Ar productionð rateÞ per21Ne ð Þ neutron production rate with 18O target. Detailed results—yields or rates from each neutron production channel in each decay chain—are reported for Upper CC rock composition. Only selected output is listed fordecay other compositions. of 1 atom Neutron yields areper given kg per rock decay ofper 1 atom year of long-lived radionuclide.kg rock Production per year rates given per year per kilogram of rock. Neutron yield (Y) Neutron production rate (S ) 39Ar production rate (S ) 21Ne prod. rate by n; a (Sn;a ) n 39Ar ð Þ 21Ne Target 232Th 235U 238U 232Th 235U 238U Sum 232Th 235U 238U Sum 232Th 235U 238U Sum Upper Continental Crust 27Al 1.69e 6 1.48e 6 1.05e 6 2265.0 72.8 1107.0 3445.0 5.15 0.12 2.00 7.27 0.0016 0.0000 0.0000 0.0016 À À À 23Na 1.15e 6 1.07e 6 7.66e 7 1547.0 52.5 805.6 2405.0 2.99 0.07 1.22 4.27 0.0003 0.0000 0.0000 0.0003 À À À 29Si 4.74e 7 4.32e 7 3.13e 7 636.9 21.2 328.7 986.9 2.42 0.08 1.26 3.77 0.0089 0.0000 0.0017 0.0107 À À À 30Si 4.09e 7 3.50e 7 2.53e 7 549.2 17.2 266.0 832.4 1.04 0.03 0.49 1.55 0.0001 0.0000 0.0000 0.0001 À À À 18 ˇ

O 3.28e 7 3.51e 7 2.80e 7 441.4 17.2 294.2 752.8 2.30 0.09 1.41 3.80 0.0082 0.0001 0.0018 0.0100 ra

À À À ´ 26Mg 2.01e 7 1.99e 7 1.43e 7 270.0 9.8 150.1 429.8 1.43 0.05 0.76 2.24 0.0131 0.0003 0.0051 0.0185 370–387 (2017) 196 Acta Cosmochimica et Geochimica / al. et mek À À À 25Mg 1.18e 7 1.19e 7 8.53e 8 158.1 5.8 89.8 253.7 1.00 0.04 0.59 1.64 0.0544 0.0021 0.0304 0.0869 À À À 19F 6.95e 8 7.18e 8 5.36e 8 93.4 3.5 56.4 153.3 0.24 0.01 0.11 0.36 0.0001 0.0000 0.0000 0.0001 À À À 17O 3.56e 8 3.77e 8 3.04e 8 47.9 1.8 31.9 81.6 0.27 0.01 0.17 0.46 0.0028 0.0001 0.0011 0.0039 À À À 56Fe 3.86e 8 7.11e 9 9.45e 9 51.9 0.3 9.9 62.1 0.06 0.00 0.00 0.07 0.0000 0.0000 0.0000 0.0000 À À À 41K 1.99e 8 1.14e 8 9.81e 9 26.7 0.6 10.3 37.6 0.06 0.00 0.02 0.08 0.0000 0.0000 0.0000 0.0000 À À À 48Ti 1.30e 8 4.65e 9 4.92e 9 17.5 0.2 5.2 22.9 0.06 0.00 0.01 0.07 0.0001 0.0000 0.0000 0.0001 À À À 13C 3.85e 9 4.12e 9 3.49e 9 5.2 0.2 3.7 9.0 0.04 0.00 0.03 0.08 0.0034 0.0001 0.0023 0.0059 À À À 44Ca 5.97e 9 3.14e 9 2.82e 9 8.0 0.2 3.0 11.2 0.02 0.00 0.01 0.03 0.0001 0.0000 0.0000 0.0001 À À À SF 2.35e 11 1.30e 10 1.14e 6 0.0 0.0 1198.0 1198.0 0.00 0.00 2.97 2.97 0.0000 0.0000 0.0211 0.0211 À À À Total 6119 203 4360 10 680 17.1 0.49 11.1 28.7 0.093 0.0027 0.064 0.159

Middle Continental Crust ~12% of fast neutrons from SF, the rest from (α,n) 18O 268.2 8.1 139.0 415.4 SF~11000 0.0 neutrons 0.0 per 576.6 year per 576.7 kg rock 39 Total~30 3876Ar atoms 100 per 2137 year per 6114 kg rock 9.1 0.21 4.6 13.915 0.108 0.0025 0.054 0.165

Lower Continental Crust 18O 48.1 1.2 20.8 70.1 SF 0.0 0.0 88.7 88.7 Total 766 17 347 1129 0.52 0.010 0.21 0.75 0.074 0.0015 0.029 0.104

Bulk Oceanic Crust 18O 8.2 0.4 7.1 15.8 SF 0.0 0.0 31.1 31.1 Total 133 5.8 121 260 0.0126 0.00051 0.0104 0.0235 0.0253 0.0011 0.0189 0.0452 Depleted Upper Mantle 18O 0.53 0.03 0.47 1.03 SF 0.00 0.00 2.09 2.09 Total 11.4 0.54 10.5 22.4 0.00014 0.000007 0.00011 0.00026 0.0207 0.0010 0.0150 0.0366 4He, n, 21Ne, 39Ar production rates

Concentration of K, Th, U decreases with depth in the Earth Table 6: Summary of production rates of 4He, neutrons, 21Ne, and 39Ar in number of atoms or neutrons per year per kilogram of Weightrock. fraction Compositional estimates are described in Table 1. Production rates, 1/(yr kg) K Th U Composition 4He neutrons 21Ne 39Ar 2.32 % 10.5 ppm 2.7 ppm Upper Continental Crust 1.64 1010 10680 753 28.7 ⇥ 1.91 % 6.5 ppm 1.3 ppm Middle Continental Crust 8.98 109 6114 416 13.9 ⇥ 0.51 % 1.2 ppm 0.2 ppm Lower Continental Crust 1.53 109 1129 70.2 0.749 ⇥ 650 ppm 210 ppb 70 ppb Bulk Oceanic Crust 3.79 108 260 15.8 0.0235 ⇥ 60 ppm 13.7 ppb 4.7 ppb Depleted Upper Mantle 2.51 107 22.4 1.06 0.000257 ⇥

Upper Continental Crust mantle upper 20–80 km Middle Continental Crust ~35 km ave. lower mantle Lower Continental Crust

Upper Mantle

Table 7: Coecients for evaluation of neutron production (columns 3–5) and 39Ar production (columns 6–8) for arbitrary compo-

sition.core Second column and third row indicate the elements whose abundances (A as weight fractions) cross-multiply a particular Rudnick & Gao 2014 doi:10.1016/B978-0-08-095975-7.00301-6 coecient. Neutron production coecients in units of “neutrons per year per kg-rock per wt-frac-target-elem per wt-frac-chain- White & Klein 2014 doi:10.1016/B978-0-08-095975-7.00315-639 parent-elem”. Argon-39 productionSalters coe cients& Stracke in units 2004 of doi: “ Ar10.1029/2003GC000597 atoms per year per kg-rock per wt-frac-target-elem per wt-frac- chain-parent-elem per K-wt-frac”. See section 6.3 for examples. 16 Neutron production 39Ar production Chain 232Th 235U 238U 232Th 235U 238U (↵, n) target A Th U U Th U U 27Al Al 2.65e+9 3.31e+8 5.03e+9 2.59e+8 2.36e+7 3.92e+8 23Na Na 6.07e+9 8.02e+8 1.23e+10 5.04e+8 4.46e+7 8.02e+8 29Si Si 1.95e+8 2.53e+7 3.91e+8 3.19e+7 4.05e+6 6.46e+7 30Si Si 1.68e+8 2.05e+7 3.17e+8 1.37e+7 1.35e+6 2.51e+7 18O O 8.77e+7 1.33e+7 2.27e+8 1.97e+7 2.89e+6 4.68e+7 26Mg Mg 1.72e+9 2.41e+8 3.72e+9 3.92e+8 5.43e+7 8.13e+8 25Mg Mg 1.01e+9 1.44e+8 2.22e+9 2.75e+8 4.04e+7 6.33e+8 19F F 1.60e+10 2.34e+9 3.75e+10 1.78e+9 2.06e+8 3.17e+9 17O O 9.50e+6 1.43e+6 2.47e+7 2.34e+6 3.39e+5 5.70e+6 56Fe Fe 1.28e+8 3.35e+6 9.55e+7 6.81e+6 2.53e+4 1.62e+6 41K K 1.10e+8 8.92e+6 1.64e+8 1.06e+7 4.72e+5 1.22e+7 48Ti Ti 4.35e+8 2.20e+7 5.00e+8 5.88e+7 2.29e+6 5.89e+7 13C C 3.61e+8 5.48e+7 9.96e+8 1.31e+8 2.10e+7 4.02e+8 44Ca Ca 2.98e+7 2.22e+6 4.29e+7 3.96e+6 2.13e+5 4.80e+6 SF 1 3.01e+3 2.37e+3 4.44e+8 2.88e+2 3.09e+2 4.73e+7

20 Attempt at quantifying uncertainty Assuming rock composition is known precisely (false…), what is uncertainty in the calculation? • Half-lives, branching ratios, α intensities … relatively small uncertainty <1% • Stopping power for α particles … ~3.5% • Cross sections of (α,n), (n,p) … challenging to estimate Cross sections from TALYS compared to experimental data from EXFOR database. Difference in integrated cross sections used as an estimate of uncertainty. … not satisfying but perhaps best approach given the lack of consistent uncertainty estimates

27Al(α,n)30P

relative difference in cross sections integrated up to 6 MeV = 36%

17 Table 8: EstimateAttempt of calculation uncertainty at and its variousquantifying contributions, assuming chemical uncertainty composition is known precisely. Contribution of the specific (↵, n) channel and spontaneous fission to neutron and 39Ar production is shown in columns 3 and 4. Calculated with Upper CC composition. † No experimental data available, uncertainty was arbitrarily set at 10 %. Uncert. est. Neutron 39Ar % % contrib. % contrib. Decay data, ↵ production < 1 Stopping power 3.5 27Al(↵, n) cross section 36 32 25 23Na(↵, n) cross section 7.7 23 15 29Si(↵, n) cross section 7.3 9.2 13 30Si(↵, n) cross section 20 7.8 5.4 18O(↵, n) cross section 17 7.0 13 26 Mg(↵, n) cross section 10† 4.0 7.8 25 Mg(↵, n) cross section 10† 2.4 5.7 Spontaneous fission 1 11 10 Overall (↵, n), neutron production 12 Overall (↵, n), 39Ar production 10 39K(n, p) cross section 28 Neutron production calculation 13 21Ne production calculation 17 % 39Ar production calculation 30

18

Table 9: Calculated production rates of 21Ne by (↵, n) and (n, ↵) and 21Ne/4He ratio. Compositional estimates are described in Table 1. Composition 21Ne prod. in atoms/kg-yr % contrib. (↵, n)(n, ↵) Total (↵, n)(n, ↵) 21Ne/4He Upper Continental Crust 753 0.159 753 99.98 0.02 4.59 10 8 ⇥ Middle Continental Crust 415 0.165 416 99.96 0.04 4.63 10 8 ⇥ Lower Continental Crust 70.1 0.104 70.2 99.85 0.15 4.58 10 8 ⇥ Bulk Oceanic Crust 15.8 0.0452 15.8 99.71 0.29 4.17 10 8 ⇥ Depleted Upper Mantle 1.03 0.0366 1.06 96.55 3.45 4.23 10 8 ⇥

19 Comparison to other studies

This study:

Mei et al. 2009 10.1016/j.nima.2009.04.032 Granitic rock composition: ▸ neutron production rate 5500 5400 ± 700

Mei et al. 2010 10.1103/PhysRevC.81.055802 ▸ 39Ar production rate 7 16 ± 5 disagreement Yokochi et al. 2012 10.1016/j.gca.2012.04.034 Cont. crust composition: ▸ 39Ar production rate 24 13 ± 3

Yokochi et al. 2014 10.1016/j.chemgeo.2014.02.004 "Lava Creek Tuff" rock ▸ 39Ar production rate 120 ± 60 140 ± 40 35% SLOWN2 code calculation, assumes mono-energetic 2 MeV neutrons decrease 91 ± 26 Units: neutrons/(kg yr) atoms/(kg yr) 19 Plug-in formulae Example: α particles from 238U decay, neutrons by 18O(α,n) 380 O. Sˇra´mek et al. / Geochimica et Cosmochimica Acta 196 (2017) 370–387

Table 7 Coefficients for evaluation of neutron production (columns 3–5) and 39Ar production (columns 6–8) for arbitrary composition. Second column and third row indicate the elements whose abundances (A as weight fractions) cross-multiply a particular coefficient. Neutron production coefficients in units of ‘‘neutrons per year per kg-rock per wt-frac-target-elem per wt-frac–chain-parent-elem”. Argon-39 production coefficients in units of ‘‘39Ar atoms per year per kg-rock per wt-frac-target-elem per wt-frac–chain-parent-elem per K-wt-frac”. See Section 6.3 for examples. Table of coefficients χn and χ39Ar Neutron production 39Ar production Chain 232Th 235U 238U 232Th 235U 238U a; n target A Th U U Th U U ð Þ 27Al Al 2.65e+9 3.31e+8 5.03e+9 2.59e+8 2.36e+7 3.92e+8 23Na Na 6.07e+9 8.02e+8 1.23e+10 5.04e+8 4.46e+7 8.02e+8 29Si Si 1.95e+8 2.53e+7 3.91e+8 3.19e+7 4.05e+6 6.46e+7 30Si Si 1.68e+8 2.05e+7 3.17e+8 1.37e+7 1.35e+6 2.51e+7 18O O 8.77e+7 1.33e+7 2.27e+8 1.97e+7 2.89e+6 4.68e+7 26Mg Mg 1.72e+9 2.41e+8 3.72e+9 3.92e+8 5.43e+7 8.13e+8 25Mg Mg 1.01e+9 1.44e+8 2.22e+9 2.75e+8 4.04e+7 6.33e+8 19F F 1.60e+10 2.34e+9 3.75e+10 1.78e+9 2.06e+8 3.17e+9 17O O 9.50e+6 1.43e+6 2.47e+7 2.34e+6 3.39e+5 5.70e+6 56Fe Fe 1.28e+8 3.35e+6 9.55e+7 6.81e+6 2.53e+4 1.62e+6 41K K 1.10e+8 8.92e+6 1.64e+8 1.06e+7 4.72e+5 1.22e+7 48Ti Ti 4.35e+8 2.20e+7 5.00e+8 5.88e+7 2.29e+6 5.89e+7 13C C 3.61e+8 5.48e+7 9.96e+8 1.31e+8 2.10e+7 4.02e+8 44Ca Ca 2.98e+7 2.22e+6 4.29e+7 3.96e+6 2.13e+5 4.80e+6 SF 1 3.01e+3 2.37e+3 4.44e+8 2.88e+2 3.09e+2 4.73e+7

☞ Python script at http://tinyurl.com/argon39 20 Ziegler et al. (2010) report the accuracy of SRIM stop- uncertainty. Koning and Rochman (2012) describe the ping power calculations when compared to experimental method of rigorous uncertainty estimation in TALYS data, which is 3.5% for stopping of a particles. Since these nuclear data evaluations which consists of a Monte Carlo stopping powers are vital to implementation technologies in approach to propagation of uncertainty in various nuclear the semiconductor industry, these values and uncertainty model parameters. Unfortunately, the available versions of can be used with confidence. TENDL (TALYS-based evaluated nuclear data library) do Estimation of a; n and n; p cross section uncertainties not include the uncertainty in a; n cross section and angu- ð Þ ð Þ ð Þ are challenging. Neutron reactions on common elements in lar distribution. the few MeV energy range are important to many applica- In the case of a; n reactions, we integrate TALYS- ð Þ tions for fission reactors. Therefore one might expect data, calculated cross sections, used in this study, and the avail- models, and codes to have quite small uncertainties. Vari- able experimental dataset (using linear interpolation ous nuclear data libraries exist, such as the ENDF/B-VII between the data points) up to 6 MeV (mean a energy in library (www.nndc.bnl.gov/endf) used by MCNP6, which the 232Th chain). The relative difference between the inte- provide nuclear cross section data. The ‘‘evaluated nuclear grals is taken as 1r uncertainty estimate. In the case of data” come from assessment of experimental data com- the 39K n; p reaction, the cross section from ENDF/B- ð Þ bined with nuclear theory modeling. The evaluated library VII.0 library used by MCNP6 and an experimental dataset datasets come with no uncertainty estimate, however. Often are integrated up to 3 MeV (above which the neutron large differences exist between cross sections from various energy spectrum is negligible). Again, the resulting relative libraries and without provided uncertainties, any rigorous difference is adopted as 1r uncertainty. comparison is impossible. Various experimental datasets Table 8 lists the various contributors to the calculation are available, some of which include uncertainty estimate. uncertainty. We considered the seven most important However, some experimental data are at odds with cross a; n target nuclides, together with spontaneous fission ð Þ sections from nuclear data libraries. accounting for >96% of neutron production. Experimental Given this unfortunate situation, we adopt the following data from Flynn et al. (1978) (EXFOR entires A0509005, strategy for estimating cross section uncertainty. Cross sec- A0509007, A0509008) were used for 27Al, 29Si and 30Si; tions used in this study are compared to experimental data (Norman et al. (1982)) (C0731001) data for 23Na; combined available via the EXFOR database at www-nds.iaea.org/ (Bair and Willard (1962)) (P0120002) and (Hansen et al. exfor. We integrate each cross section over energy up to a (1967)) (P0116003) data for 18O. In the case of 25Mg and certain upper energy bound and calculate the relative differ- 26Mg where no experimental data are available in the ence between the integrated data sets. We use this difference EXFOR database, we arbitrarily set the uncertainty at as a measure of the cross section uncertainty (1r). 10%. The experimental 39K n; p cross section dataset is ð Þ Clearly, this method of uncertainty estimation is not sat- constructed from Johnson et al. (1967) and Bass et al. isfying. However, given the lack of rigorous uncertainty (1964) experimental data. Nolte et al. (2006) report a estimates provided by existing data libraries, the variability non-zero 39K n; p 39Ar cross section in the epithermal in the cross section data gives us at least some estimate of ð Þ energy range. However, the cross section is smaller by at Noble gases in natural CO2 reservoirs from SW USA 1175

1. INTRODUCTION Whilst primordial noble gas isotopes have been studied in CO2-rich well gases since 1961 (e.g. Zartman et al., Carbon dioxide (CO2) has been identified as the most 1961; Phinney et al., 1978; Caffee et al., 1999; Ballentine important compound currently affecting the stability of et al., 2005; Holland and Ballentine, 2006), no systematic the Earth’s climate (Intergovernmental Panel on Climate study using a full noble gas data set to investigate CO2 ori- Change, 1996, 2001, 2007). Geologic storage of anthropo- gin and quantify its interaction with the groundwater sys- genic CO2 in depleted oil and gas reservoirs is one of the tem subsurface has yet been undertaken. Here, we present key options for short-term control of CO2 emissions. How- a He, Ne, Ar, Kr and Xe isotopic study of the CO2 from ever, CO2 is a reactive gas that has significantly more influ- five separate producing gas fields located throughout the ence upon the host rocks and their formation waters than Colorado Plateau and Rocky Mountain provinces. In this 3 1182 S.M.V. Gilfillan et al. / Geochimica et Cosmochimica Acta 72 (2008) 1174–1198 petroleum fluids (Baines and Worden, 2004). In order for paper we show how measured CO2/ He ratios enable us this technology to be safely implemented the long term con- to determine the CO source in each system. We then use 2 1 sequence of injecting CO into the subsurface must be the noble gases to identify the subsurface processes acting 2 St Johns Dome quantified. on the CO2 and develop a quantitative model of gas and McElmo Dome Natural subsurface CO2 accumulations have existed on groundwater interaction. Doe Canyon geological timescales and provide key analogues that in- form us about the feasibility of long term storage of anthro- 2. THE COLORADO PLATEAU AND ROCKY ) -3 0.1 pogenic CO2 (e.g. Baines and Worden, 2004; Haszeldine MOUNTAIN NATURAL CO2 RESERVOIRS et al., 2005). Despite the amount of information available from these sites, in many natural CO2 reservoirs the source The Colorado Plateau is a massive, high-standing tec- STPcm 3 of the CO2 and basin scale processes that act on them are tonic block located in the south western US, centred on poorly understood. This is partially due to the multiple ori- the Four Corners of the states of Colorado, New Mexico, (cm 2 0.01

Origin of ColoradoN Plateau CO2 gins of CO2 in natural gases. These include methanogenesis, Utah and Arizona (Fig. 1). It is abruptly flanked to the east oil field biodegradation, kerogen decarboxyilation, hydro- by the majestic Rocky Mountains, the result of at least 2 carbon oxidation, decarbonation of marine carbonates km of• upliftmethanogenesis during the Laramide Orogeny and later Ceno- Largely magmatic origin and degassing of magmatic bodies (Jenden et al., 1993; zoic uplifts (Baars, 2000). The bulk of magmatic activity on Wycherley et al., 1999). 13C(CO ) can be used to distin- the Coloradooil Plateaufield biodegradation occurred in the Late Cenozoic, along d 2 • 0.001 ⟹ Mantle signature in guish between some of these different sources. However, the transitionalkerogene south western decarboxylation margin (primary between 5 0.0001 0.001 0.01 0.1 13 • the d C(CO ) of natural gas fields containing high CO and 15 Ma) and the northwest southeast trending Rio accompanying4 3 gases-3 2 2 hydrocarbon oxidation He (cm STPcm ) concentrations (>70% CO2) almost always lies in the over- Grande• Rift (0–5 Ma) (Fitton et al., 1991). This has coin- 4 lapping range between magmatic degassing and carbonate cided with the most-recent and best constrainedFig. uplift 2. Plot event of N2 against He concentration for the McElmo Dome, Doe Canyon and St. John’s Dome fields, showing a positive correlation • decarbonation of marine carbonates4 3 4 breakdown, and these different sources cannot be readily at the Plateau’s southwest margin betweenbetween 6 and N2 1and Ma He. This relationship is attributed to thelow gas phase CO (CO2/2)He stripping old groundwater containing accumulated He and N2 degassing of magmatic(Ballentine bodies and Sherwood Lollar, 2002). distinguished. In addition, because of the high CO2 solubil- • ity in water and CO2 reactivity, the extent of interaction of the CO phase with the groundwater is a critical parameter 2 1.00e+12 in quantifying CO2 sinks. The groundwater systems associ- McCallum Dome ated with CO2 gas deposits are often similar to those of oil Sheep Mountain and gas field brines and there are few techniques available Bravo Dome 1.00e+11 Crustal CO2 ONLY to quantify regional groundwater movement through age/ McElmo Dome Doe Canyon residence time determination or the groundwater volumes St. Johns Dome that have interacted with the trapped reservoir phase. 1.00e+10 Noble gas isotopic and abundance measurements can be Magmatic He ratio CO2 Dilution used to constrain CO2 origins and subsurface interaction 3 / CO2 Range 2 with the groundwater system. This is because noble gases 1.00e+9 from the terrestrial atmosphere, dissolved in groundwater, CO Doe Canyon?? can be distinguished isotopically from those from the man- tle that contain a primordial signature, and those produced 1.00e+8 by the radiogenic decay of U, Th and K within the crust. CO Loss and Dilution CO Loss When combined with the distinct elemental abundance pat- 2 2 terns of the different sources, it is possible to resolve the rel- 1.00e+7 ative noble gas input from each source to any crustal fluid. 0.7 0.8 0.9 1.0 This allows the extent of crustal, mantle and atmospheric 3 -3 contributions to the fluid to be constrained (e.g. Ballentine CO2 Volume (cm STPcm ) et al., 2002). Additionally, noble gases can be used to con- 3 Fig. 3. Plot of CO2/ He against CO2 concentration for all of the reservoirs studied. All error bars are smaller than printed symbols. The strain natural gas associated water residence times (Zhou 3 shaded region highlights theGilfillan range ofet CO al.2/ 2008He values 10.1016/j.gca.2007.10.009 measured in pure magmatic samples (MORB). All of the samples from the gas et al., 2006) and quantify the degree of interaction a crustal reservoirs plot within or below this range, indicating the presence of a significant quantity of magmatic 3He. This implies21 a mantle origin for fluid has had with the groundwater system (Ballentine et al., the majority of the CO2 contained within these fields (Sherwood Lollar et al., 1999; Ballentine et al., 2001). 1991; Pinti and Marty, 1995; Ballentine et al., 1996; Torger- Fig. 1. Map of the Colorado Plateau illustrating the sites of major sen and Kennedy, 1999; Ballentine and Sherwood Lollar, Cenozoic igneous provinces, location of the natural CO2 reservoirs 7 7 20 36 values of 2.89 · 10 to 5.08 · 10 within all of the fields are John’s Dome. As with Ne, Ar increases with proximity sampled and other CO2 reservoirs within the region. 2002; Zhou et al., 2005). higher than both the measured average crust value of to the gas/groundwater contact. 1.71 · 107 and theoretical estimates of between 2.02 and Measured 40Ar/36Ar ratios from all of the reservoirs are 2.64 · 107 (Ballentine and Burnard, 2002). This is consistent considerably above the air value of 295.5, as a result of a with release from a shallow, low temperature regime (Ball- resolvable excess of 40Ar (40Ar*) (Fig. 4). Helium isotopes entine et al., 1994; Ballentine and Sherwood Lollar, 2002). indicate both mantle and crustal contributions to the noble gases, the 40Ar* is therefore a mixture of mantle and crus- 4.3.3. Argon tal-derived 40Ar. In the case of Bravo Dome, 40Ar*/4He In all of the reservoirs except Sheep Mountain 40Ar cor- correlates with 3He/4He isotope variation with a correlation relates directly with 4He and 20Ne. A similar relationship coefficient of 0.993 (Ballentine et al., 2005). Extrapolating 36 20 3 4 8 exists between Ar and Ne within all the reservoirs and to He/ Hecrust =1· 10À (Ballentine and Burnard, 2002) 4 40 is particularly strong within Bravo, McCallum and St. resolves He/ Arcrust = 22.0. This is significantly above Doe Canyon CO2 well gas

• Gas from deep CO2 wells, spec. Doe Canyon, shows low level of 39Ar • 39Ar activity a factor of 1400 ± 200 below atmospheric DarkSide-50: 2016 10.1103/PhysRevD.93.081101 • Underground nucleogenic production in rock with ? sufficiently low K, Th, U concentration…? • Gas sequestered underground (negligible cosmogenic D.-M. MEI et al. replenishment) with limited exchange…?PHYSICAL REVIEW C 81,055802(2010)

-10 - 10 µ Capture on 39K Constrained by µ- capture 1 10-11 µon-induced neutrons -12 Constrained by µ-induced neutrons )) 10

-1 -1 -13 y 10 10 Constrained by (α,n) neutrons

-1 Natural radioactivity induced neutrons -14 At depths >2000 m.w.e.

-2 Ar 10 Atmospheric level

10 40 10-15 nucleogenic 39Ar production -16 10-3 10 Ar to Ar -17 39 10 dominates 10-4 10-18 10-19 10-5 10-20 -21 10-6 The raio of 10 Ar Production (atoms (g -22

39 10 10-7 10-23 10-24 -8 10 from Mei-25 et al. 2010 10.1103/PhysRevC.81.055802 1234 56 10 12345622 Depth (km.w.e.) Depth (km.w.e.)

39 FIG. 4. (Color online) The Ar production as a function of depth. FIG. 5. (Color online) The ratio of the 39Ar to 40Ar calculated The eight common elements found in Earth’s rocks are used in the using formulas and methods described in the text. The eight common calculation. elements found in Earth’s rocks are used in the calculation. Shown is aratioassumingthattheproducedargonismixedata10%levelby 39 The depth dependence of the Ar production is shown underground gas. in Fig. 4.Notethattherockchemicalcompositionvaries from location to location. However, the dependence of the understanding of 39Kcontentandradioactivitylevelintherock 39 production of Ar from the change in the rock chemical surrounding the gas field or water reservoir where the argon composition is much weaker than that of the change in depth. gas is extracted from is essential. 39 40 If one assumes that Ar and Ar have the same possibility The underground production of 40Ar is primarily from 40K to diffuse out of the grains of the rock and to be mixed with via electron capture decays. Therefore, an accurate knowledge 39 40 existing gas in the pore volume, then the Ar/ Ar ratio in the of the potassium content is critical. The underground produc- pore volume can be expressed as tion of 42Ar is negligible.

N39 p R Ar × , (7) = N 0 N d p Ar + Ar × III. EXPERIMENTAL RESULTS OF POTASSIUM where p is the possibility of 39Ar and 40Ar being mixed in the CONTENT 0 40 pore volume, NAr is the original number of Ar contained in a certain amount of gas from the 40Ar concentration in ambient A. Underground water site d air, which is related to the pore volume, NAr is the number The community of Wall, SD pumps water using seven of 40Ar atoms produced underground through 40K EC decays. wells that extract water from the Fall River aquifer in an Using Eq. (7), the ratio of the 39Ar to 40Ar can be calculated old rock formation 3200 feet below the surface. The average with known natural abundance of 39K in the rock. Given a water consumption is 200 gal/min. The water temperature rock density of 2.7 g/cm3,arockporosityof20%,andan assumption that the pore volume is filled with gas with the 10-10 - -11 Constrained by µ capture ambient air concentrations (78%/21%/1% of N2/O2/Ar), the 10 ratio of 39Ar to 40Ar can be determined. The results are shown 10-12 Constrained by µ-induced neutrons 10-13 Constrained by (α,n) neutrons in Fig. 5. -14

Ar 10 Atmospheric level Equation (7)canbeappliedtoaundergroundwaterreser- 40 10-15 39 40 voir, where p is the possibility of Ar and Ar being dissolved 10-16 3 Ar to -17 in water. For a rock density of 2.7 g/cm ,arockporosityof 39 10 20%, and an assumption that the pore water contains gas with 10-18 10-19 the ambient air concentrations (78%/21%/1% of N2/O2/Ar) -20 39 40 10 -21 at a solubility of 61 mg/Lforargon,theratioof Ar to Ar The raio of 10 as a function of depth is shown in Fig. 6. 10-22 As can be seen in Fig. 4, the larger the depth, the smaller the 10-23 39 10-24 production of Ar. On the other hand, at the large depth, the -25 production of 39Ar is dominated by stopping negative muon 10 123456 Depth (km.w.e.) capture and the neutrons from natural radioactivity. Because 40 39 Ca(n,2p) Ar reaction requires neutron energy to be greater FIG. 6. (Color online) The ratio of the 39Ar to 40Ar calculated than 8.3 MeV and the majority of the (α,n)neutronsinducedby using formulas and methods described in the text. The eight common radioactivity is less than this reaction threshold, the production elements found in Earth’s rocks are used in the calculation. Shown is of 39Ar at large depth underground is dominated by stopping aratioassumingthattheproducedargonisdissolvedata10%level negative muon capture on 39Kand39K(n ,p )39Ar. Hence, the by underground water.

055802-4 Doe Canyon CO2 well gas

Message from the mantle? Or story of the crust?

Accumulation of gas in Gas from low K, Th, U source rock? isolated reservoir in the crust? 23 Predicting 39Ar/40Ar produced underground

Accumulation of 39Ar and 40Ar over time from initially degassed rock

Atmosphere Ar 40 Ar/ 39

DarkSide-50 measurement No rock composition explains the measurement

Age in Myr 24 How to explain anomalously low 39Ar?

Argon sequestered at depth with no 39Ar production nor influx ➝ 39Ar activity lowered by a factor of 1400 in 2800 years (~10.5 half-lives with t1/2 = 269 yr)

Atmospheric component

Reservoir Crustal component

Mantle component

Rates of inflow (increase in 39Ar) vs. residence time (decay of 39Ar) 25 Noble gases in natural CO2 reservoirs from SW USA 1183

30000

25000

20000 Ar 36 15000 Ar/ 40 McCallum Dome 10000 Sheep Mountain Bravo Dome McElmo Dome 5000 Doe Canyon St. Johns Dome 0 01234 3 4 Noble gasHe/ He (R/Ra ) isotopic constraints 3 4 40 36 3 4 40 36 Fig. 4. Plot of He/ He (R/Ra) against Ar/ Ar for all samples. The clear correlation between He/ He and Ar/ Ar within Bravo Dome contrasts with the lack of variation measured in both Sheep Mountain and McElmo Dome and the anti correlation observed in McCallum. Noble gases in natural CO2 reservoirs from SW USA 1183 Gilfillan et al. 2008 10.1016/j.gca.2007.10.009 13 30000 MORB McCallum Dome Sheep Mountain Bravo Dome 25000 12 McElmo Dome Mantle - Crust mixing line Doe Canyon St. Johns Dome 20000 11 Ne Ar 22 36 15000 Ar/

Ne/ Air - Mantle mixing line 40

20 10 Air McCallum Dome 10000 Sheep Mountain Air - Crust mixing line Bravo Dome 9 McElmo Dome 5000 Doe Canyon St. Johns Dome Doe Canyon?? Crust 8 0 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 01234 21 22 3 4 Ne/ Ne He/ He (R/Ra ) This Doe Canyon gas may not be 20 22 21 22 3 4 40 36 3 4 40 36 Fig. 5. Plot of Ne/ Ne ratio against Ne/ Ne for all samples. Bravo Dome exhibits mixingFig. 4. between Plotrepresentative of aHe/ pre-mixedHe (R/R crust/aira) against of DarkSide componentAr/ Ar and for gas! all samples. The clear correlation between He/ He and Ar/ Ar within Bravo Dome the mantle. St. John’s, Sheep Mountain and McCallum20Ne, highlight 21Ne, mixing 22Ne between nucleogenic a pre-mixedcontrasts crust/mantle with the lack component of variation and measured air. The distinctin both Sheep Mountain and McElmo Dome and the anti correlation observed in McCallum. values measured from McElmo Dome show that this field contains the highest proportional contribution from crustal/radiogenic sources. 3He, 36Ar primordial Well defined end-member ratios 4 40 4 40 13 He, Ar radiogenic average crustal production of He/ Arcrust = 5 and is con- Dome samples are not reported here, a recent analysis of MORB McCallum Dome 4 21 for air, crust, mantle sistent with the resolved He/ Necrust discussed above, a second suite of samples are similarly indistinguishable Sheep Mountain Bravo Dome indicating a thermal regime controlling the release of crus- from the air value suggesting that at the level12 of analytical McElmo Dome Mantle - Crust mixing line tal radiogenic light noble gases in preference to heavier no- precisionNoble no gas kinetic isotopes fractionation can of Ar provide isotopes due some to clues. Doe Canyon ble gases (Ballentine et al., 1994; Ballentine and Sherwood commercial production (e.g. Zhou et al., 2005) or signifi- St. Johns Dome 4 40 36 38 Lollar, 2002). The mantle derived He/ Ar relationships cantNeed radiogenic measurementsAr or Ar contributions of actual11 are resolvable. DarkSide gas. Ne at Bravo Dome are discussed elsewhere (Ballentine et al., 22 26 2005; Holland and Ballentine, 2006). Within the other 4.3.4. Krypton, Xenon and groundwater noble gas 3 4 Ne/ Air - Mantle mixing line reservoirs, without significant He/ He variation, contribu- components 20 10 tions to 40Ar* from both the mantle and crust cannot 84Kr in all fields correlate with 36Ar and also with 20AirNe be simply resolved and the 40Ar* must be considered as within Bravo, McCallum and St. John’s Domes. In both St. Air - Crust mixing line a mix of the crustal and mantle components John’s and Bravo Domes there is a clear and9 coherent in- 40 84 ( Arcrust + mantle). crease in Kr on moving towards the gas/groundwater 38Ar/36Ar values in all measured samples are indistin- contact confirming that the majority of the 84Kr is derived Crust 8 guishable from the atmospheric ratio. While the Bravo from the dissolved air component within the0.00 groundwater. 0.02 0.04 0.06 0.08 0.10 0.12 0.14 21Ne/22Ne

Fig. 5. Plot of 20Ne/22Ne ratio against 21Ne/22Ne for all samples. Bravo Dome exhibits mixing between a pre-mixed crust/air component and the mantle. St. John’s, Sheep Mountain and McCallum highlight mixing between a pre-mixed crust/mantle component and air. The distinct values measured from McElmo Dome show that this field contains the highest proportional contribution from crustal/radiogenic sources.

4 40 average crustal production of He/ Arcrust = 5 and is con- Dome samples are not reported here, a recent analysis of 4 21 sistent with the resolved He/ Necrust discussed above, a second suite of samples are similarly indistinguishable indicating a thermal regime controlling the release of crus- from the air value suggesting that at the level of analytical tal radiogenic light noble gases in preference to heavier no- precision no kinetic fractionation of Ar isotopes due to ble gases (Ballentine et al., 1994; Ballentine and Sherwood commercial production (e.g. Zhou et al., 2005) or signifi- Lollar, 2002). The mantle derived 4He/40Ar relationships cant radiogenic 36Ar or 38Ar contributions are resolvable. at Bravo Dome are discussed elsewhere (Ballentine et al., 2005; Holland and Ballentine, 2006). Within the other 4.3.4. Krypton, Xenon and groundwater noble gas reservoirs, without significant 3He/4He variation, contribu- components tions to 40Ar* from both the mantle and crust cannot 84Kr in all fields correlate with 36Ar and also with 20Ne be simply resolved and the 40Ar* must be considered as within Bravo, McCallum and St. John’s Domes. In both St. a mix of the crustal and mantle components John’s and Bravo Domes there is a clear and coherent in- 40 84 ( Arcrust + mantle). crease in Kr on moving towards the gas/groundwater 38Ar/36Ar values in all measured samples are indistin- contact confirming that the majority of the 84Kr is derived guishable from the atmospheric ratio. While the Bravo from the dissolved air component within the groundwater. Summary

• Calculation of underground nucleogenic production of 39Ar: α particles from natural radioactivity, fast neutrons from (α,n) on light targets and from spontaneous fission, 39Ar from 39K(n,p)

• Evaluation for several representative rock compositions, plug-in formulae to evaluate for an arbitrary rock-like composition.

• Estimate of calculation uncertainty (30% for 39Ar production rate) and comparison to previous results.

• Low 39Ar level in Doe Canyon gas is puzzling – isotopic composition suggests both atmospheric and crustal contribution to the magmatic origin gas, low 39Ar requires ≫3000 year isolation.

• Possible scenario: Gas of magmatic origin with mantle signature, mixes with crustal and atmospheric components, sequestered in an underground reservoir for sufficient time.

• Measurement of noble gas isotopic composition of Doe Canyon well gas is needed. Thank you. 27