School of Engineering and Science - Jacobs University

More documents

Recommendations

Info

determinations in carbonate rocks regardless of the acid matrix, concentration determinations using 62 Ni provide excellent results for samples that are diluted in HNO 3 , primarily through a reduction in the 25 Mg 37 Cl interference. Elements that appear to be hosted primarily in refractory mineral phases resistant to decomposition using the HNO 3 carbonate method include Ti, Rb, Zr, Nb, Cs, Hf, and Ta, which would be expected for these incompatible elements. Two elements that may be concentrated in different mineral phases are Ba and Pb (C- and S-rich minerals?), for which the carbonate decomposition recovers approximately 90% and 80%, respectively, of the concentrations observed for the HF-HClO 4 decomposition method. As stated above, it appears that the carbonate decomposition provides satisfactory results for W. However, W is one of three elements apparently ‘lost’ during the analytical recovery test of the carbonate decomposition method, along with Ta, and to a lesser extent Nb (see Section 6, Fig. 10). The low analytical recovery for Nb, Ta, and W are attributed to the filtering step of the carbonate decomposition. For Nb and Ta, the poor analytical recoveries during the carbonate decomposition method are inconsequential, as regardless, sample dissolution in HNO 3 is not expected to provide accurate determinations of these elements. In the case of W, the observed loss during the filtering step is estimated at ~0.002 mg/kg (Fig. 10), which is also inconsequential when considering the reference concentration of W in the JDo-1 dolomite of ~0.260 mg/kg. It is therefore unlikely that any loss of W during the filtering step for the carbonate decomposition would significantly affect W determinations, assuming W is present at concentrations ≥0.050 mg/kg in the sample powder. This is particularly true when considering the uncertainty in the JUB measured W data for JDo-1 (~15% RSD). 7.6. Marine ferromanganese nodules and crusts Samples which comprise a large fraction of the geochemical research performed with the Geochemistry Lab at JUB are marine ferromanganese nodules and crusts (referred to collectively as Fe-Mn crusts). These Fe-Mn crusts are typically 13- 15% SiO 2 , 3-5% Al 2 O 3 , ~3% each CaO and MgO, 33-38% MnO, and 8-15% Fe 2 O 3 . Therefore, the high metal content and interferences from Mn and Fe are expected to provide the greatest obstacles to accurate trace metal determinations. However, unlike other metal-rich rocks such as iron-formations, Fe-Mn crusts are highly enriched in 47
many trace metals, and therefore may be diluted significantly following decomposition. For example, the combined REY concentration within the Fe-Mn crust JMn-1 is >800 mg/kg, whereas the iron-formation IF-G contains only ~22 mg/kg REY. While IF samples typically may not be diluted by a factor greater than 1000, Fe- Mn crust analyses may benefit from higher dilution factors (≥5000). Several Fe-Mn crust CRMs are available for geochemical studies at JUB, and include JMn-1 (Table 1), NOD-A-1 and NOD-P-1 (USGS), and GSPN-2 and GSPN-3 (Institute of Rock and Mineral Analysis, People’s Republic of China). For this discussion only the JMn-1 ferromanganese crust issued by the Geological Survey of Japan is considered, though additional information regarding analyses at JUB of other Fe-Mn crust CRMs is available from K. Schmidt. Figure 21 displays comparisons between measured JUB trace metal concentrations in JMn-1 with the average reference value. Measured concentrations are lower than the average reference value for all analyzed elements, with JUB data for 24 of 32 elements between 80-95% of the reference value. The precision of the JUB is generally quite good (2-3 % RSD), suggesting that the low measured values are reproducible and do not result from a single anomalous analysis. As discussed earlier, at high TDS contents samples may suffer from signal suppression effects (Section 4.2), which may result in low measured concentrations. However, this does not appear to be the case with JMn-1, as all of the JUB analyses were performed at dilution factors of 2500-5000, and the internal standard correction factors for these analyses were typically between 0.90–1.10. Rather, it is hypothesized that absorption of water vapor by JMn-1 sample powder is primarily responsible for the low measured concentrations. For this study, the JMn-1 powder was not dried prior to analysis, as for most CRMs drying the rock powder (e.g., overnight at 105-110°C) has not proven to significantly affect the measured concentrations of the elements of interest. However, Fe-Mn crusts are microcrystalline marine precipitates composed of poorly ordered mineral phases that, unlike other rock CRMs, have never undergone mineral recrystallization during diagenesis and lithification. As a result it is possible that powdered Fe-Mn crusts may more readily absorb water from ambient laboratory air (even though all CRMs are stored in dessicators until ready for use), and any increase in sample mass due to this effect would drive measured concentrations of trace metals towards lower values. This hypothesis is supported by detailed studies of Fe-Mn crusts by K. Schmidt 48
Page 1:
Iron sedimentation and the neodymiu
Page 4 and 5:
The process(es) by which IFs were d
Page 7:
This thesis represents original and
Page 10 and 11:
viii
Page 12 and 13:
goal of the research is to discern
Page 14 and 15:
Figure 1. Map of southern Africa sh
Page 16 and 17:
appropriate to briefly discuss: 1)
Page 18 and 19:
100 10 MORB 1 0.1 raw data (mg/kg)
Page 20 and 21:
10 -2 hydrothermal fluid 10 -3 10 -
Page 22 and 23:
seawater and returned to the sedime
Page 24 and 25:
100 MCO + 3 + M(CO 3 )- 2 rare eart
Page 26 and 27:
net effect of carbonate complexatio
Page 28 and 29:
feature was not recognized in early
Page 30 and 31:
sound geochemical basis for anomalo
Page 32 and 33:
deviation of the 143 Nd/ 144 Nd rat
Page 34 and 35:
available, it should be possible to
Page 36 and 37:
modern and Archean oceans, as well
Page 38 and 39:
Bau M., Möller P., and Dulski P. (
Page 40 and 41:
Elderfield H. and Greaves M.J. (198
Page 42 and 43:
Haley B.A., Klinkhammer G.P. (2003)
Page 44 and 45:
Lee J.H. and Byrne R.H. (1993) Comp
Page 46 and 47:
Piepgras D.J. and Wasserburg G.J. (
Page 48 and 49:
Tosiani T., Loubet M., Viers J., Va
Page 50 and 51:
THIS PAGE INTENTIONALLY LEFT BLANK
Page 52 and 53: Abstract The benefits of inductivel
Page 54 and 55: List of Figures Figure 1. Sample de
Page 56 and 57: 1. Introduction The advent of induc
Page 58 and 59: Figure 1. Flow-chart diagram of the
Page 60 and 61: during decomposition (SiF 4 and CO
Page 62 and 63: Fractionation between the three iso
Page 64 and 65: cps, then 3000 cps must be subtract
Page 66 and 67: variety of ways. These range from r
Page 68 and 69: Table 3. Change in HFSE concentrati
Page 70 and 71: 0.90, corresponding to an anomalous
Page 72 and 73: 10 2 JDo-1 reference values 10 1 IF
Page 74 and 75: Figure 5. Diagram depicting various
Page 76 and 77: 20 18 basalt BHVO-2 (n=5) run preci
Page 78 and 79: The method precision for elements p
Page 80 and 81: Two approaches may be utilized to d
Page 82 and 83: For the more recent, full 32 elemen
Page 84 and 85: isotopic data for geologic CRMs, av
Page 86 and 87: Table 4. Interferences observed for
Page 88 and 89: value uncertainty (as %RSD) is repr
Page 90 and 91: 1.50 FeR-4 (n=2) 1.25 JUB / referen
Page 92 and 93: 1 IF-G/shale 0.1 0.01 La Ce Pr Nd P
Page 94 and 95: consistent (App. 1), and similar to
Page 96 and 97: 1.50 SGR-1b (n=3) 1.25 JUB / refere
Page 98 and 99: 1.50 JCh-1 (n=3) 1.25 JUB / referen
Page 100 and 101: 1.50 JDo-1 (n=5) 1.25 JUB / referen
Page 104 and 105: 1.50 JMn-1 (n=4) 1.25 JUB / referen
Page 106 and 107: eference value. The result is that
Page 108 and 109: eference value and the measured JUB
Page 110 and 111: the 32 analyzed elements, with Ti b
Page 112 and 113: Govindaraju K. (1995) 1995 working
Page 114 and 115: Appendix 1. Analytical data Appendi
Page 116 and 117: Appendix 1. Literature reference va
Page 118 and 119: Appendix 1 continued. Data in mg/kg
Page 120 and 121: the concentration of the major elem
Page 122 and 123: Potential interferences on monoisot
Page 124 and 125: 1500 0.5 M HCl 4.0 interference on
Page 126 and 127: 4000 0.40 0.20 3500 0.17 mg/kg 0.18
Page 128 and 129: interference on 95 Mo in sample sol
Page 130 and 131: THIS PAGE INTENTIONALLY LEFT BLANK
Page 132 and 133: Nd isotopes in 2.9 Ga Archean surfa
Page 136 and 137: Table 2 Trace element concentration
Page 152 and 153:
Author's personal copy B.W. Alexand
Page 154 and 155:
Page 156 and 157:
Page 158 and 159:
Page 160 and 161:
Page 162 and 163:
Page 164 and 165:
Page 166 and 167:
and preservation of oxide-facies IF
Page 168 and 169:
TOP relative stratigraphic position
Page 170 and 171:
The mechanism by which the alkali m
Page 172 and 173:
6. Widdel, F. et al. Ferrous iron o
Page 174 and 175:
Page 176 and 177:
Page 178 and 179:
82 PRESERVATION OF PRIMARY REE PATT
Page 180 and 181:
Page 182 and 183:
Page 184:
ulk, open ocean seawater tended to
show all

School of Engineering and Science - Jacobs University

Create successful ePaper yourself

Delete template?

Save as template?