Zircon Raman thermochronology: Data valuation and measurement protocol

,

Multi-band Raman analysis allows to discriminate spectra exhibiting overlap by comparing damage estimates (D) from different bands (Fig. 3).

PARTIAL ANNEALING
As for other thermochronological methods, the interpretation of zircon Raman ages depends on the thermal history of the dated sample.One aspect is the distinction of partial and complete annealing, marking the difference between mixed and reset ages (Wagner, 1981).For unannealed zircons, the measured radiation damage is equivalent to the zircon's -dose (D = D α α ), but decreases with annealing.
and Mogok Belt granite (M71, magenta) zircons affected by band overlap deviate from the expected relationship (D 2 = D 3 = D ER ) towards higher D 3 and D ER values (D 3 > D ER > D 2 ).Fig. 4 exploits the difference between D 3 and D 2 by plotting the D 3 /D 2 ratio against D 2 and comparing the data with the prediction boundaries based on random measurement errors.The TDR and M71 data exhibiting band overlap lie outside the upper prediction boundary.These data are classified as broadened by overlap and may be excluded from further analysis.

Fig. 7 .
Fig. 7. Flow chart of a suggested zircon Raman dating procedure.The protocol consists of three steps: preparation, measurement, and evaluation.The overlap test helps to detect spectra containing artifacts that may be excluded from further analysis.The annealing evaluation is crucial for correctly interpreting the apparent zircon Raman age.

Fig. 7
Fig. 7 shows a flow chart of zircon Raman dating, integrating the outlined evaluation steps into the measurement procedure for (1) detecting spectra with artificially broadened bands before measuring eU and (2) distinguishing between complete and partial annealing when interpreting the apparent zircon Raman age.

Fig. 1 .
Fig. 1.CL image and Raman spectrum of a Tardree rhyolite zircon.The bands show asymmetric shapes (arrows) from band overlap, with exception of ν 2 (SiO 4 ).The white dots in the CL images mark the Raman spots, with the spectrum being from the uppermost spot.Band overlap occurs despite the spots being in an apparently homogeneous zone.Band assignment after Kolesov et al. (2001).

Fig. 3 .
Fig. 3. D vs. D plots with D calculated from the ν 2 (SiO 4 ), ν 3 (SiO 4 ), and ER bandwidths using the calibration of Härtel et al. (2021a).The insets include the region up to 80 10 16 /g in detail α .Error bars are 2 .Some data of Tardree rhyolite (TDR) and Mogok Belt granite (M71) are affected by band overlap.The σ samples from the Katha range (K86 and K92) are partially annealed.Data from Duluth Complex FC1 are shown for comparison.

Fig. 6 .
Fig. 6.CL and Raman images of a 3 x 1 mm Plešovice zircon.(a) CL image.(b) Raman map of Γ ER indicating the degree of radiation damage.(c) Raman map of the D 3 /D 2 ratio with areas showing band overlap encircled in black.(d) Raman map of the D 2 /D ER ratio.White pixels indicate measurement spots excluded due to low signal-to-noise ratio.

Fig. 5 .
Fig. 5. Plots of D 3 /D 2 vs. D 2 .The dashed line marks the expected value of 1 for unannealed or completely reset data.Prediction boundaries are included for 90 % (dark gray), and 99 % (light gray) of the data, assuming 4 % relative errors on Γ (1 ).The σ samples affected by partial annealing are the Katha range orthogneisses (K86 and K92).