Back to article

Figures and tables

Determining subduction-zone fluid composition using a tourmaline mineral probe

V.J. van Hinsberg1,

1Department of Earth and Planetary Sciences, McGill University, Montreal, Quebec, Canada

G. Franz2,

2Department of Applied Geosciences, Technical University Berlin, Berlin, Germany

B.J. Wood3

3Department of Earth Sciences, University of Oxford, United Kingdom

Affiliations  |  Corresponding Author  |  Cite as  |  Funding information

van Hinsberg, V.J., Franz, G., Wood, B.J. (2017) Determining subduction-zone fluid composition using a tourmaline mineral probe. Geochem. Persp. Let. 3, 160–169.

EU-MC-FP7

Geochemical Perspectives Letters v3, n1  |  doi: 10.7185/geochemlet.1719
Received 2 September 2016  |  Accepted 6 March 2017  |  Published 28 March 2017
Copyright © 2017 European Association of Geochemistry




Figure 1 Reading the Tauern tourmaline record. (a) Tauern Eclogite Zone P–T path (Zimmermann et al., 1994

Zimmermann, R., Hammerschmidt, K., Franz, G. (1994) Eocene high pressure metamorphism in the Penninic units of the Tauern Window (Eastern Alps): evidence from 40Ar-39Ar dating and petrological investigations. Contributions to Mineralogy and Petrology 117, 175–186.

) with conditions recorded by tourmaline growth zones shown. Inset cartoons locate these in a schematic subduction-zone cross-section. (b) PPL image of the tourmaline transect. (c) XMg sharply defines the tourmaline growth zones. (d) F-content in tourmaline, >0.5 apfu F in the outer core and mantle. (e) XCa of tourmaline, which tracks the XCa in the coexisting fluid (von Goerne et al., 2011

von Goerne, G., Franz, G. van Hinsberg, V.J. (2011) Experimental determination of Na–Ca distribution between tourmaline and fluid in the system CaO–Na2O–MgO–Al2O3–SiO2–B2O3–H2O. Canadian Mineralogist 49, 137–152.

). (f) Temperature transect across the tourmaline grain calculated from inter-sector and inter-polar thermometry. (g) Tourmaline K-content, which acts as a qualitative barometer. Uncertainties shown are the typical 1s analytical precision.
Back to article | Download in Powerpoint


Figure 2 (a) Element concen­trations in fluids reconstructed from the Tauern tourmaline outer core and mantle and its mineral inclusions (blue symbols), and eclogite minerals from other subduction-zone terrains show a consistent pattern. (b) MORB-normalised concentrations for average arc magma parallel those of subduction-zone fluids supporting a genetic link. See the Supplementary Information for data sources and calculation method. Uncertainties are given as the Interquartile Range (IQR) and maximum estimates marked with an arrow.
Back to article | Download in Powerpoint


Figure 3 Fluid compositions reconstructed from tourmaline (circles) and its mineral inclusions (triangles), show a consistent, high Ba/Th at low La/Sm sig­nature. Addition of this fluid to the arc-magmatism source region can explain the trend to ele­vated Ba/Th compared to mid-ocean ridge basalt (MORB) and primitive mantle (PM).
Back to article | Download in Powerpoint


Back to article

Supplementary Figures and Tables


Table S-1 WDS electron-microprobe data for tourmaline growth zones, mineral inclusions in tourmaline, and matrix minerals. The values shown are the average for multiple analyses and their associated 1 standard deviation spread, except for omphacite where only one analysis was available. Normalisation factors used are the sum of cations for omphacite, titanite, plagioclase, amphibole and calcite, or the oxygen sum for phengite and biotite. Tourmaline was normalised to the sum of the elements residing at the Y, Z and T sites = 15, or Si = 6 cpfu, if the former resulted in Si > 6 cpfu.


Tourmaline growth zones

inner core1smid core1souter core 11souter core 21smantle1sinner rim1smid rim1souter rim1s
SiO237.90.438.10.438.30.238.20.738.70.536.00.235.90.336.20.3
TiO20.440.050.390.030.420.100.250.030.190.020.180.050.360.090.350.08
Al2O329.10.329.30.229.70.730.80.131.80.328.20.328.70.228.40.4
MgO10.30.410.00.210.20.210.10.112.80.110.70.49.50.39.50.3
MnO0.0040.0080.010.010.0040.0060.010.020.0070.0090.0080.0070.0020.0040.0070.008
FeO9.670.8410.300.439.061.018.080.053.040.102.730.313.790.484.430.82
CaO1.20.10.930.061.10.10.640.051.60.21.30.30.560.080.90.2
Na2O2.480.052.590.012.540.102.830.022.320.102.10.22.490.092.10.2
K2O0.090.020.0950.0020.090.010.0920.0060.060.010.0100.0030.0130.0020.0120.002
F0.60.30.600.010.90.10.820.021.20.10.540.060.350.080.510.10

norm. factorY + Z + T = 15
Y + Z + T = 15
Y + Z + T = 15
Y + Z + T = 15
Y + Z + T = 15
Si = 6
Si = 6
Si = 6

Si5.920.025.930.025.950.035.940.085.930.046-6-6-
Ti0.0520.0060.0460.0030.0490.0120.0290.0040.0220.0020.0220.0060.050.010.040.01
Al5.360.025.370.015.40.15.650.045.740.055.540.085.660.065.560.04
Mg2.400.082.310.022.370.052.330.022.910.042.660.102.370.062.340.07
Mn0.0010.0010.0010.0020.0010.0010.0020.0020.0010.0010.0010.0010.0000.0010.0010.001
Fe1.30.11.340.061.20.11.050.010.390.010.380.040.530.070.60.1
Ca0.190.020.1560.0090.180.020.110.010.260.030.230.050.100.010.160.03
Na0.750.020.7810.0070.760.030.850.010.690.030.680.050.810.040.680.06
K0.0180.0030.01890.00020.0180.0030.0180.0010.0120.0030.0020.0010.00270.00050.00260.0004
F0.30.10.2950.0020.420.070.410.010.570.060.280.030.180.040.270.06

XMg0.660.030.630.010.670.030.690.000.880.000.870.020.820.020.790.04
dravite0.500.010.510.010.520.040.600.010.620.030.600.040.50.30.50.2
schorl0.270.030.290.010.260.010.270.000.080.000.090.020.120.070.130.05
uvite0.190.020.160.010.180.020.110.010.260.030.230.050.080.040.140.06
foitite0.0370.0040.0440.0020.0420.0150.0230.0180.0400.0140.090.010.30.40.30.3


Mineral inclusions in tourmaline mantle

Matrix minerals

omphacite1sphengite1stitanite1s

biotite1splagioclase
amphibole
calcite
SiO253.9-48129.60.4

37.11.759.10.251.50.80.030.02
TiO20.04-0.200.07343

0.50.4< d.l.-0.090.030.010.02
Al2O35.9-24142

14.60.921.50.2610.0010.001
MgO11.3-4.60.50.040.02

19.70.70.010.0117.70.80.470.09
MnO0.03-0.0040.0060.0090.008

0.020.01< d.l.-0.020.010.090.10
FeO5.3-2.00.30.70.3

9.70.80.140.068.30.50.330.09
CaO17.9-< d.l.-28.30.3

< d.l.-3.470.09102551
Na2O4.2-0.20.20.020.01

0.160.079.50.12.00.8< d.l.-
K2O0.002-10.80.50.0010.002

920.060.010.170.04< d.l.-
F0.01-0.60.21.10.6

1.60.40.050.080.540.090.020.03

norm. factor∑cat = 4
O = 11
∑cat = 3


O = 11
∑cat = 5
∑cat = 15
∑cat = 1

Si1.97-3.440.060.9780.008

2.830.082.7830.0017.30.10.00050.0003
Ti0.001-0.0110.0040.840.07

0.030.02--0.0100.0030.00020.0002
Al0.25-1.990.090.150.06

1.30.11.1910.0070.90.20.000020.00003
Mg0.62-0.490.060.00180.0009

2.240.090.0010.0013.70.20.0120.002
Mn0.001-0.00020.00040.00030.0002

0.0010.001--0.0030.0010.0010.001
Fe0.16-0.120.020.0190.008

0.620.050.0050.0020.990.050.0050.001
Ca0.70---1.0040.003

--0.1750.0041.50.20.9820.002
Na0.29-0.030.020.0010.001

0.020.010.860.010.60.2--
K0.0001-0.980.040.000050.00008

0.90.20.0040.0010.030.01--
F0.001-0.140.050.120.06

0.40.10.010.010.240.040.0010.002
Back to article | Download in Excel


Table S-2 Laser-ablation ICP-MS data for Ba, Th, La and Sm in tourmaline, and phengite and titanite inclusions. Most data represent the mean of multiple measurements. Uncertainties are the 1 standard deviation on the mean, or the count statistical error if only one measurement is available. Partition coefficients are those of Green and Adam (2003), combined with Dphg/tur from Klemme et al. (2011) and are used to calculate the Ba/Th and La/Sm ratios for the fluid. Analysis locations are shown in Figure 1.


Concentrations in ppm
Partition coefficients (min/fl)
Ratios
MineralBa1 stdTh1 stdLa1 stdSm1 std
BaThLaSm
Ba/Th fluidLa/Sm fluid
Tur core - 11.10.30.0090.0060.0080.0070.090.07
0.060.881.081.50
17950.12
Tur core - 13.00.70.030.020.040.020.080.10
0.060.881.081.50
17700.64
Tur core - 2a1.70.40.020.010.030.020.200.16
0.060.881.081.50
13970.18
Tur core - 2b1.91.00.020.020.0120.0060.080.09
0.060.881.081.50
19500.22
Tur core - 2c0.90.20.0070.0080.030.010.130.06
0.060.881.081.50
20880.30
Tur mantle - 3220.0110.0050.0050.0060.0280.008
0.060.881.081.50
28130.27
Tur rim - 40.810.070.0090.0010.0040.0030.040.01
0.060.881.081.50
14430.17
Tur rim - 50.80.90.020.020.0360.0030.0310.005
0.060.881.081.50
7291.64

















Phengite19081830.0120.0050.020.010.040.04
91.21.533.846.07
27820.92
Titanite55730.60.70.450.0453
0.154.418.503173.83
27100.79
Back to article | Download in Excel


Table S-3 Laser-ablation ICP-MS trace element data for tourmaline, and phengite and titanite inclusions. The tourmaline data represent the mean of multiple measurements and the uncertainties reported are 1 standard deviation on this mean. For phengite and titanite, we report the count statistical error. Measurements were corrected for differences in ablation behaviour between standards and samples by normalising to EMP Si-content.

Conc in ppmTourmalineInclusions
ElementRim of core1 stdMantle1 stdPhengite1 std cseTitanite1 std cse
Li31374102506192
V65464.10.871.60.9943
Cr209201394412222999
Mn26.60.617111.60.417.10.8
Cs0.070.040.070.0226.10.30.070.03
Sr110728137613242.40.838713
Ba0.90.60.60.21778173.80.6
Ti80511613124619419018870015341
Nb< d.l.
0.02
1.040.07772
Ta0.0070.0060.02
0.120.023.50.2
Zr< d.l.
360.070.03164
Hf< d.l.
0.10.10.0030.0050.50.1
La< d.l.
0.0010.0010.010.010.480.05
Ce< d.l.
0.0140.0020.0100.0063.40.2
Sm< d.l.
0.0240.0060.040.043.40.4
Lu< d.l.
0.0070.002< d.l.
0.560.07
Y0.030.020.020.030.060.02362
Pb1346.70.32.40.7144
Th0.010.010.010.010.0150.0090.090.03
U0.0020.0010.0080.0060.0000.0041.30.1

< d.l. Value below detection limit

Back to article | Download in Excel


Figure S-1 Tourmaline is stable for most of the P–T range typically found in subduction zones, and is an accessory phase in exposed subduction-zone lithologies for all the terrains shown. Typical subduction path (blue solid lines) from Syracuse et al. (2010) and P–T estimates for subduction-zone terrains from Zimmermann et al. (1994) and Marschall et al. (2009). Tourmaline stability fields from van Hinsberg et al. (2011b).
Back to article | Download in Powerpoint


Figure S-2 Petrography of the Tauern tourmaline grain used in this study. (a) Mineral-map showing its inclusion and matrix mineralogy, and the core, mantle and rim zones of the tourmaline grain. This map is based on back-scattered electron imaging and WDS-EMP element mapping for Ca, Mg, Ti, Fe, Si and Na. (b) Example of hourglass sector zoning in the outermost rim in PPL microscopy. (c) Back-scattered electron image of compositional heterogeneity in the tourmaline core, outlining the foliation of protolith mica grains overgrown by tourmaline. (d) Outline of an idiomorphic eclogite-stage garnet grain replaced during growth of the mantle zone in PPL microscopy. (e) Back-scattered electron image of c+-c- polar tourmaline growth.
Back to article | Download in Powerpoint


Table S-4 Partition coefficients used in this study. Tourmaline–fluid and titanite–fluid coefficients were calculated by combining the phengite–tourmaline and phengite–titanite inter-mineral partition coefficients for Syros sample SY309B (Klemme et al., 2011; Marschall, 2005) with phengite-fluid experimental partition coefficients from Green and Adam (2003).

ElementPhg/TurPhg/fluidTur/fluidTtn/fluid
Li5.32.00.40.004
B0.0040.942330.001
V1.25223179327
Cr3.028795
Mn2.8114.029
Rb2501080.40.9
Cs257.80.30.8
Sr0.201.46.913
Ba1635910.060.15
Ti0.4075187
Nb247.10.31929
Ta34511.5471
Zr2.22.91.421
Hf4.0


La3.63.81.119
Ce3.13.31.1155
Nd3.3


Sm4.06.11.5174
Eu5.4


Gd1.6


Dy1.7


Er2.8


Yb9.3


Lu4.0112.81618
Y4.0246.0
Pb4.24310116
Th1.71.50.90.9
U0.931.61.74.5
Back to article | Download in Excel

Figure S-3 Reconstructed Ba/Th vs. La/Sm ratios for subduction-zone fluids in the Tauern eclogite zone, as determined from tourmaline compositions (see Fig. 2 for locations of these points). (a) Calculated assuming constant D tur/fl, obtained by combining experimentally determined D phg/fl of Green and Adam (2003) with D phg/tur of Klemme et al. (2011). (b) Calculated from D phg/fl of Green and Adam (2003) extrapolated to relevant P–T conditions using phengite lattice systematics at constant D phg/tur (from Klemme et al., 2011). The impact of changing speciation on D tur/fl is shown by the thick arrows. Reconstructing fluid compositions with extrapolated D values results in lower Ba/Th in the fluid, with only minimal impact on La/Sm. The differences in Ba/Th between core and mantle become smaller, but the predicted effect of changes in speciation would re-introduce this difference. The gray uncertainty ellipses show the 1 standard deviation uncertainty in the ratios resulting from the spread in the trace element analyses of the minerals.
Back to article | Download in Powerpoint