Application of Inductively Coupled Plasma Quadrupole Mass Spectrometry for the Determination of Monazite Ages by Lead Isotope Ratios

A fim de avaliar a aplicabilidade da espectrometria de massa com plasma indutivamente acoplado à determinação de idades Pb/Pb, U/Pb e Th/Pb de monazitas, estudos foram realizados, inicialmente, em padrões isotópicos de referência para chumbo (NIST SRM 981 e 982). Posteriormente, a metodologia otimizada foi aplicada a areias monazíticas de três diferentes locais, Pão de Açúcar (Rio de Janeiro), Buena (Estado do Rio de Janeiro) e Praia Negra (Guarapari, Estado do Espirito Santo); as idades médias obtidas, (581 ± 21)Ma, (552 ± 32)Ma e (535 ± 3)Ma, respectivamente, estão de acordo com os valores da literatura para amostras de zircão e de monazita dos mesmos locais. Idades Th/Pb com mono-grãos foram também determinadas para a amostra de monazita da Praia Negra, nove grãos foram analisados e o valor médio, (530 ± 26)Ma, está no acordo com o valor obtido com alíquotas maiores. Finalmente, o método foi aplicado a uma amostra do torianita do Estado de Amapá e as idades obtidas Th-Pb, U-Pb e Pb-Pb foram de (2,15 ± 0,05)Ga, (2,03 ± 0,01)Ga e (2,044 ± 0,006)Ga, respectivamente, com um valor médio de (2,08 ± 0,07)Ga, coerente com a idade de (2,08 ± 0,02)Ga do complexo de Bacuri, Amapá, relatada na literatura.


Introduction
Considering the four stable isotopes of lead, only 204 Pb is non-radiogenic.Since the other three, 206 Pb, 207 Pb and 208 Pb, are end-members of the 232 Th ( 208 Pb), 235 U ( 207 Pb) and 238 U ( 206 Pb) radioactive decay chains, different minerals show different isotopic compositions according to their age and the initial Th and U content of the source rock. 1 Although less resistant than zircon, monazite offers the advantage of generally higher U and Th content and often behaves as a closed-system for U and Pb. 2 For the purpose of lead dating, thermal ionization mass spectrometry (TIMS) using a magnetic sector mass analyzer has been the conventional method for high precision isotopic analysis with relative standard deviations (RSD) lower than 0.01% when applied to reference materials, NIST-SRM-981 [3][4][5] or BCR-SRM-278 (Mussel Tissue), NIST-SRM-1577A (Bovine Liver) and BCR-SRM-422 (Cod Muscle). 68][9] Moreover, the relatively high cost of TIMS instrumentation and the extensive chemical pre-treatment required have imposed limitations on the routine use of TIMS techniques in geochemical exploration applications. 5,10,11lthough the quadrupole inductively coupled plasma mass spectrometry (ICP-MS) shows inferior precision in comparison with TIMS, it has some important advantages, such as (i) simple sample pre-treatment, (ii) high sample throughput and simple sample introduction and (iii) widespread availability. 12uadrupole based ICP-MS instruments have been used in many studies to measure Pb atom ratios.The precisions measured in reference materials are typically 0.1-0.3%RSD, [13][14][15][16] whereas for real samples, a precision in the range of 0.2-0.5% RSD is usually obtained. 8,9,17,18][9] In the present work, the application of quadrupole ICP-MS for the determination of lead atom ratios has been evaluated by aiming at its application for dating of monazites, using the 206 Pb/ 238 U, 208 Pb/ 232 Th, 207 Pb/ 206 Pb and U-Th-total Pb methods.

Experimental
Instrumentation A Perkin-Elmer SCIEX ELAN 6000 ICP-MS equipped with the original cross flow nebulizer was used for the Pb isotopic measurement.A peristaltic pump performed the solution aspiration.The ICP-MS instrumental operating conditions are summarized in Table 1.

Reagents and standard solutions
All reagents used were of analytical-reagent grade or higher purity and de-ionized water was further purified using a Millipore Milli-Q water purification system.Certified lead atom ratio standard materials (SRM 981) and (SRM 982) were purchased from the National Institute of Standards and Technology (NIST, USA).

Monazite samples Buena monazite sample
A pure monazite concentrate sample from Buena (São Francisco de Assis county), in the north of the Rio de Janeiro State (Brazil), was obtained from INB (Indústrias Nucleares Brasileiras).

Sugar Loaf Hill and Black Sand beach samples
In both locations (Urca beach, Rio de Janeiro/RJ) and Black Sand Beach (BSB) (Guarapari, ES), the area with the highest radioactivity was localized with a cintilometer and approximately 5 kg sand samples were taken.The monazite fraction was separated as described by Barling et al. 20

Sample dissolution and analysis
For each sample, a bulk sample was prepared by crushing, by hand, several monazite grains using an agate mortar.Three 25 mg aliquots of each sample, equivalent to 150 grains, were weighed, transferred to a 10 mL platinum crucible, to which 500 μL of phosphoric acid was added and then was heated to red-hot for a few minutes by applying a Meker burner.A clear solution was observed and diluted to 10 mL with 7.2 mol L -1 nitric acid.One milliliter was removed and diluted to 50 mL with 2% (v/v) nitric acid.All determinations were standardized against a reagent blank.
Using this solution, lead, uranium and thorium content were determined by ICP-OES and ICP-MS, respectively, using thallium as internal standard.Each aliquot is analyzed three times and the method standard deviation is 1% for lead and uranium and 2% for thorium.Prior to the lead isotopic ratio determinations, the mass bias correction factor for 204 Pb, 207 Pb and 208 Pb related to 206 Pb were determined using the NIST-SRM-981 (10 μg L -1 ).The obtained values were verified using the NIST-SRM-982 (10 μg L -1 ) as a sample.These factors were verified at the end of each sample batch, analyzing the NIST-SRM-981 solution as a sample.One percent relative bias was specified as a limit for recalibration and sample reanalysis and during the present work no reanalysis was necessary.
Additionally to the 206 Pb/ 238 U, 208 Pb/ 232 Th and 207 Pb/ 206 Pb ages, it was tested the U-Th-total Pb method, applying the equation described by Rhede et al. 21and supposing that all lead is radiogenic.

Atom ratio precision and accuracy
Many factors directly control the signal acquisition process on the ELAN 6000 ICP-MS.Some are automatically optimized; others can be manually controlled, among them, the dwell time, the number of sweeps/ replicate and the number of replicates/analysis.These three parameters were selected in this study in order to determine the best operation condition.Initially, the default operation conditions were applied and a high RSD was obtained.Considering the different atom ratios involved, better results were achieved by applying different dwell times for each isotope, higher for 204 Pb, the least abundant, and lower for 208 Pb, the most abundant, according Platzner et al. 14 Based on the ELAN 6000 operation manual, 22 the dwell time for each isotope was calculated in order to obtain 10 6 counts/replicate, using 200 sweeps/replicate.Also based on that, ten replicates/analysis was chosen.As the sample volume required and the time expended for each sample were too high, the same calculation was performed for 500.000counts and 6 replicates/analysis. 202 Hg was also included because of a potential isobaric interference of 204 Hg, a dwell time equal to that of 204 Pb was chosen (Table 1).
Based on the NIST SRM 981, 1.5, 7.5, 15, 30 and 45 μg Pb L -1 solutions were prepared.The 204 Pb, 206 Pb, 207 Pb and 208 Pb values obtained using the 10 μg Pb L -1 standard solution were used to calculate the mass fractionation correction factor for 204 Pb, 207 Pb and 208 Pb related to 206 Pb.Using these factors, 204 Pb/ 206 Pb, 207 Pb/ 206 Pb and 208 Pb/ 206 Pb atom ratios were calculated and results are shown in Figure 1.For lead concentrations higher than 20 μg L -1 , 208 Pb/ 206 Pb ratios lower than the certified value were obtained due to dead time effects on the 208 Pb determination.This concentration can be taken as an upper limit of lead in the sample solution.
According to Platzner et al., 14 it is possible to reduce the relative standard deviation, in isotope composition determinations using a quadrupole ICP-MS with consecutive measurements.This procedure of multiple determinations was performed employing a solution of 10 μg L -1 of SRM 981 to nine sequential determinations.The previous procedure was repeated using a 10 μg L -1 SRM 982 solution.The results are listed in Table 2 and show that precision better than 0.25% (95% confidence limits) as well as accuracy in the range of 0.1 to 2% can be achieved.In general, for 204 Pb/ 206 Pb, 207 Pb/ 206 Pb and 208 Pb/ 206 Pb atom ratios, the relative bias increases with the distance to an atom ratio equal to one.J. Braz.Chem.Soc.

Bulk samples
The lead, uranium and thorium contents of the three analyzed monazite samples and the observed lead atom ratio are shown in Table 3.For each individual aliquot, the monazite ages were calculated applying the 206 Pb/ 238 U, 208 Pb/ 232 Th, 207 Pb/ 206 Pb and U-Th-total Pb methods and the obtained values are shown in Table 4.The agreement between the results of the 206 Pb/ 238 U and 208 Pb/ 232 Th dating methods shows the proposed procedure reliability.The association of a higher thorium content with a lower uranium concentration leads to a low 207 Pb signal and to a large standard deviation related to the 207 Pb/ 206 Pb age.The calculated U-Th-total Pb ages were in the range of the 206 Pb/ 238 U, 208 Pb/ 232 Th and 207 Pb/ 206 Pb ages showing that, at least, for monazite samples this relatively simple method can provide results that could be used as a starting point for further studies applying more accurate methods based on multi-collector mass spectrometers.
The obtained Sugar Loaf Hill 238 U/ 206 Pb age is coherent with the 560 Myears value published by Silva and Ramos. 23][26] The overall compositional variations among various monazite types are governed by substitution of REE by Th and U. 27,28 Therefore, since the observed uranium Relative bias/(%) -    and thorium contents were different (Table 3), an additional aliquot of each monazite sample, was taken and uranium, thorium and the rare earth elements (REE) were determined in order to verify if this substitution has occurred.Figure 3 shows that this substitution effectively occurs with an inverse correlation between Σ(LREE: La-Sm) and (Th+U).The chondrite-normalized REE content is shown in Figure 4.All monazite samples have patterns with negative europium anomaly and with similar light REE (LREE) composition.The heavy REE are divergent, the largest difference being found for the heaviest elements.Mohanty et al., 27 analyzing monazite sand grain from the Chhatrapur beach placer deposits, have also observed a uniform LREE enrichment with a prominent Eu anomaly, which has been attributed to the preferential incorporation of the LREE during the melting.
The method was also tested on older samples.A thorianite sample from Amapá was obtained from the Brazilian Nuclear Energy Commission.Three 10 mg

Figure 4 .
Figure 4. Black Sand beach monazites: single grain Th/Pb ages obtained in the present work (open dots) in comparison with those reported by Machado and Gauthier (closed dots refence 2), each errors bar represents two standard deviations.The black and the two red reference lines are the mean value and the 95% uncertainties calculated based on the open dot values.

Table 2 .
Results for the analysis of lead atom ratio standard reference materials, errors are based on 95% confidence limits

Table 3 .
Observed lead, uranium and thorium contents (in g kg-1) and lead atom ratios of the analyzed monazite samples (errors are based on 95% confidence limits) Rare-earth elements concentration, normalized to the Chondrites, in the studied monazite samples.Chondrite values are taken from Anders and Grevesse (reference 33).La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Table 5 .
Lead and thorium contents and 208 Pb atomic percentage observed in Black sand beach, Guarapari, monazite grains, and calculated Th/Pb ages (±95% confidence level)