Evaluation of Pb-Pb and U-Pb Laser Ablation ICP-MS Zircon

GEOSTANDARDSNEWSLETTERThe Journal of Geostandards and Geoanalysis

Evaluation of Pb-Pb and U-Pb Laser Ablation ICP-MS Zircon Dating using Matrix-Matched Calibration Sampleswith a Frequency Quadrupled (266 nm) Nd-YAG Laser

Vol. 25 — N° 2-3 p . 3 6 1 - 3 7 3

This paper reports the successful application of laser ablation (LA) ICP-MS to the in situ analysis of207Pb/206Pb and 206Pb/238U isotopic ratios on zircon crystals using matrix-matched calibrationsamples as external calibrators. For 207Pb/206Pbanalyses, LA-ICP-MS results on reference materials(UQ-Z1 and G91500) indicated individual precisionsin the range 1-10% (2s), most analyses being betterthan 6%. The resulting weighted means were associated with errors typically better than 1% withages of 1148 ± 5 Ma (2s) and 1069 ± 9 Ma (2s) respectively. Analyses of well-dated late Archaeangranitic rocks from the western margin of the YilgarnCraton (Australia) are presented and show a close agreement with the reference values. An orthogneissdated at 2662 ± 4 Ma (2s) by ion microprobe(SHRIMP) gave a 207Pb/206Pb age of 2657 ± 6 Ma(2s). A more complex zircon population from a syenite emplaced at 2654 ± 5 Ma containing a ≤ 3.25 Ga inherited component has been investigated using a spot size of approximately 45µm. LA-ICP-MS provides a 207Pb/206Pb age of 2653± 6 Ma with older grains yielding ages of up to3.23 Ga. Dating of younger rocks (< 1 Ga), however,was limited by poor precision in the measurement ofthe 207Pb/206Pb isotopic ratios and by inter-elementfractionation between Pb and U during the ablationprocesses. Using a high power density, variations ofthe 206Pb/238U ratios during one spot analysis appeared to correlate positively with time over the first minute of ablation. A linear fit of the dataacquired during this period allowed a 206Pb/238Uratio to be calculated, thus reducing the magnitudeof the fractionation and improving precision toaround 5% (2s). Results on the G91500 zircon

Cet article présente l’application de la techniqued’ablation laser par ICP-MS à l’analyse in situ desrapports isotopiques 207Pb/206Pb et 206Pb/238U surcristaux de zircons en utilisant un matériel de référence de matrice identique. Pour les analyses207Pb/206Pb, les résultats des expériences d’ablationlaser sur zircons gemmes (UQ-Z1 et G91500) fournissent des précisions analytiques de l’ordre de 1 à 10% (2s) , la plupart des analyses étantmeilleures que 6% (2s). Les moyennes pondéréesfournissent des erreurs typiques de l’ordre de 1%avec des âges de 1148 ± 5 Ma (2s) et 1069 ± 9 Ma(2s) respectivement. Les résultats obtenus sur desgranitoïdes Archéens d’âge connu affleurant sur la bordure Ouest du Craton du Yilgarn (AustralieOccidentale) sont présentés et sont en bon accordavec les valeurs de référence. Un orthogneiss, datéà 2662 ± 4 Ma (2s) par sonde ionique (SHRIMP),fournit un âge 207Pb/206Pb de 2657 ± 6 Ma (2s).Une syénite, mise en place à 2654 ± 5 Ma (2s) etprésentant une population de zircon plus complexeavec un héritage ancien pouvant atteindre 3.25 Gaa été datée par ablation laser à 2653 ± 6 Ma (2s).Des âges plus anciens, pouvant atteindre 3.23 Ga, ont également été reconnus sur des cœurs hérités.L’application à des roches plus récentes (< 1 Ga) estcependant limitée par la faible précision dans lamesure des rapports isotopiques 207Pb/206Pb et parun fractionnement inter élémentaire important entrele Pb et l’U durant le processus d’ablation. En utilisant une puissance importante, la variation desrapports 206Pb/238U présente une corrélation linéaireavec le temps durant la première minute d’analyse.Une régression des données obtenues durant cetintervalle permet de calculer le rapport 206Pb/238U

3 6 1

1201

Olivier Bruguier (1)*, Philippe Télouk (2), Alain Cocherie (3), Anne-Marie Fouillac (3) and Francis Albarède (2)

(1) ISTEEM-CNRS, Service ICP-MS, cc 049, Université de Montpellier II, Place E. Bataillon, 34095 Montpellier, France. * Corresponding author, e-mail: [email protected](2) Laboratoire des Sciences de la Terre, Ecole Normale Supérieure de Lyon, 46 Allée d’Italie, 69364 Lyon Cedex 7, France(3) BRGM, 3 Avenue C. Guillemin, BP 6009, 45060 Orléans, France

Received 10 Nov 00 — Accepted 04 Jun 01

U-Pb zircon geochronology is probably one of themost widely used and reliable dating techniques forthe determination of (re)crystallisation age of rocks ina wide variety of environments. This is partly due tothe ubiquitous presence of zircon and to the robust-ness of the U-Th-Pb systems in this mineral, whichmake it prone to survive anatexis and high-grademetamorphic conditions, as well as to pass throughthe sedimentary cycle (e.g. Compston and Pidgeon1986). To date, the different competing techniquescapable of dating zircon with reasonable precisionare limited to the somewhat expensive SIMS andID-TIMS analyses. Laser ablation (LA) ICP-MS hasbeen shown recently to be a potentially valuablealternative to Pb-Pb dating (Feng et al. 1993, Fryere t al . 1993, Ludden et al . 1995, Machado andGauthier 1996, Machado et al . 1996, Scott andGauthier 1996). Although less precise, this techniquehas been shown to be cheaper and faster thanthe other two and still has the advantage of in situanalysis at the scale of a few tenths of a micrometresquare, thus allowing the determination of parts ofgrains identified within one crystal. The first work onLA-ICP-MS used a Nd:YAG laser operated at 1064nm (e.g. Machado and Gauthier 1996), but recent

progress in laser technology make it possible to qua-druple and even to quintuple (Jeffries et al. 1998) thefundamental wavelength and to use the fourth (266nm) and f i f th (213 nm) harmonic , so improv ingabsorption of the laser beam by the target materialand thus increasing laser efficiency as well as reducingthe spot size. The large inter-element fractionationobserved during ablation processes (Longerich et al.1996) , however, has hampered the wide use ofLA - ICP-MS in U -Pb g e o c h r o n o l o g y . O n l y f e wa t t e m p t s h a v e b e e n s u c ce s s f u l i n ob ta i n i ng206Pb/238U ratios and these studies have so far dealtwith the evaluation of the method using either inter-national NIST certified reference materials, which arevery different from natural zircon crystals (Hirata andNesbitt 1995, Hirata 1997) or l iquid calibrations(Horn et al. 2000). In this study, we chose to investi-gate the capabil i t ies of a frequency quadrupledNd:YAG laser in U-Pb zircon geochronology usingmatr ix-matched reference mater ials (UQ-Z1 andG91500 zircon crystals). The aim of these preliminaryexperiments was to establish whether an analyticalprotocol could be proposed that would providereproducible 238U-206Pb ratios at the precision levelexpected for a quadrupole ICP-MS.

3 6 2


reference sample yielded a 206Pb/238U age of 1057± 14 Ma, in good agreement with the publishedreference value (1062.4 Ma). Late Hercynian zirconsfrom a Corsican alkali granite dated at ca. 285 Maby TIMS and SHRIMP yielded a younger but consistent age of 277 ± 11 Ma. These results showthat using a somewhat simple apparatus (quadrupole ICP-MS and 266 nm Nd:YAG laser),the technique has the capability of producing precise and reliable 207Pb/206Pb and 206Pb/238Uages with a minimal sample preparation and a highthroughput. The present limitations are associatedwith the high density power used in this study, asanalyses must be conducted on grains larger than80 µm that are free of inclusions and fractures,which often result in “catastrophic” ablation. Shorterwavelength lasers, which yield a better laser-targetcoupling and which produce smaller ablated particles, should help to reduce these drawbacks.

Keywords: laser ablation, ICP-MS, U-Pb geochronology,zircon, inter-element fractionation.

avec une meilleure précision (ca. 5%), tout en réduisant l’amplitude du fractionnement. Les résultats obtenus sur le cristal gemme G91500 fournissent un âge 206Pb/238U de 1057 ± 14 Ma en bon accord avec la valeur publiée de 1062.4Ma. Un granite alcalin hercynien daté à 285 Mapar TIMS et par SHRIMP, fournit un âge par ablation laser de 277 ± 11 Ma. Ces résultats montrent que grâce à un appareillage relativementsimple (ICP-MS quadrupôlaire et laser Nd:YAG à266 nm) la technique d’ablation laser permet d’obtenir rapidement des âges 207Pb/206Pb et206Pb/238U fiables et précis avec un minimum depréparation. Une limite à cette technique résidedans le fait que la puissance utilisée nécessite descristaux de taille supérieure à 80 µm, dépourvusd’inclusions et de fractures qui, souvent produisentune ablation catastrophique. L’utilisation de lasers àlongueur d’onde plus courte, qui présentent unemeilleure absorption du faisceau et qui produisentdes particules de taille plus petite, devrait permettrede contourner ces problèmes.

Mots-clés : ablation laser, ICP-MS, géochronologie U-Pb, zircon, fractionnement inter élémentaire.

Experimental

Apparatus

Ablation experiments were carried out using aSpectraphysic GCR-130 Nd-YAG pulsed laser, opera-ting in the ultraviolet region at 266 nm. The laser wasoperated in the Q-switched mode at a repetition rateof 10 Hz and a pulse duration of 5 ns. Under theseconditions, the maximum laser energy output was 50mJ per shot at 266 nm. A plano-convex lens was usedto focus the laser beam onto the sample surface andthe ablation pit diameter for zircon was kept constantat ca. 40-50 µm throughout this study. The ablationcell was made in-house of Teflon with a silica windowon top with a total volume of ca. 5 cm3. Samples wereplaced in the cell and flushed with an argon flow forabout fifteen minutes before starting measurements.Sample preparat ion requirements were minimal .Zircons grains, selected using a binocular microscopef rom non-magnet ic concent ra tes , were mountedtogether with chips of reference samples in epoxy resinand slightly polished to expose the top of the grainsand to give a reasonably flat surface. The mounts werethen carefully washed with tri-distilled water, soap andalcohol and flushed with nitrogen before their intro-duction into the ablation cell. The laser apparatus wascoupled to a VG Plasmaquad II ICP-MS modified byaddition of a rotary pump, allowing a two fold decreaseof the vacuum in the expansion chamber and, thus,enhancing sensitivity. Before each laser session, initialset-up of the instrument was done using a 5 ng g-1

solution of In, Pb and U, which typically yielded a sen-sitivity of 100 x 106 cps per µg g-1 of In. The machinewas then quickly set to the laser mode by diverting theargon nebu l i se r gas f low to the ab la t ion ce l l .Conditions were subsequently refined for dry plasmaanalysis by ablating a NIST SRM 610 silicate glassCRM containing ca. 500 µg g-1 of Pb and other traceelements. The torch box position, lens setting, gas flowsand resolution were optimized and the machine wastuned on 208Pb in order to minimise background andHg interference. This generally resulted in a loss of sen-sitivity, but in a more favourable signal/backgroundratio. Under these conditions, the ICP-MS achieved atypical sensitivity of up to 2000 cps per µg g-1 Pb.Laser and ICP-MS operating parameters are summari-zed in Table 1.

All laser ablation experiments were conducted inthe peak jumping acquisition mode using three pointsper peak. The dwell time for all peaks was set to

10.24 ms except for 202Hg, 204Pb and 232Th, whichwere counted in 5.12 ms and 207Pb, which was mea-sured for a longer time (20.48 ms). A typical acquisi-tion consisted of five repeats of 10 s each. 238U wasalso used to screen the data and repeats with lessthan 10 000 cps on this isotope were rejected fromthe calculation in a similar way to that described byMachado and Gauthier (1996). The samples werepre-ablated for 10 s before measurement to avoid theinitial signal pulses and to achieve enhanced preci-sion. The typical procedural runs included one gasblank measurement fol lowed by three calibrationsamples, one gas blank, six unknowns, one gas blank,six unknowns and finally three calibration samples.Blanks were averaged and the values were subtractedfrom calibration samples and unknowns.

Pb and U-Pb fractionation

Early studies (e.g. Feng et al. 1993, Hirata andNesbitt 1995) have shown that the main problem inPb-Pb and U-Pb laser ablation isotopic analyses arerelated to mass bias, inter-elemental fractionation andinterferences that can combine to restrict seriously pre-cision and accuracy of the results. Mass bias is mainlythe result of space charge effects taking place in both


3 6 3

Table 1.Laser and ICP-MS operating parameters

LaserLaser type Quadrupled Nd-YAGWavelength 266 nmLaser mode Q-switchedRepetition rate 10 HzPrimary output power 50 mJ per pulse at 266 nmPulse width 5 nsAblation cell Teflon made, 5 cm3 internal volumeTransportation system Tygon tube, ca. 3 m total length

ICP-MSModel VG PQII+ (with one additional rotary pump)Forward power 1350 WReflected power < 5 WCool gas 14-15 l min-1

Auxilliary Gas 1-1.1 l min-1

Carrier gas 1.1 l min-1

Acquisition ParametersDetector mode Pulse countingMeasured isotopes 202Hg, 204Pb, 206Pb, 207Pb, 208Pb, 232Th, 238UDwell time per isotope 10.24 ms (20.48 on 207Pb, 5.12 on 202Hg,

204Pb and 232Th)Quad settle time 10 msPoints/peak 3Pre-ablation time 3-10 sAcquisition time 10 sNo. of repeats 5

the plasma and the sampler-skimmer cone interfaceregion (Hirata 1996). This fractionation favours theheavy isotopes, which are preferentially transmitted,while lighter isotopes are more easily dispersed awayfrom their trajectory. In this work, the magnitude ofmass bias was evaluated using external calibrationprocedures either on the silicate glass NIST SRM 610or on the natural zircon crystals UQZ1 and G91500.Results of these ablation experiments are reported in

Figure 1, which shows a series of 207Pb/206Pb ratiomeasurements performed within a period of a singleday (Figure 1a and 1b) or over a few weeks (Figure 1c).For the NIST SRM 610 glass, the data yielded a weigh-ted mean of 0.9111 ± 0.0052 (2s), which gives a massbias value of 0.16% when compared to the referencevalue of Walder at al. (1993). Measurements achievedon the G91500 gem crystal (Wiedenbeck et al. 1995)provided a weighted mean of 0.07503 ± 0.00032

3 6 4


Figure 1. Mass bias diagrams calculated from a set of 207Pb/206Pb measurements performed on (a) the silicate

glass CRM NIST SRM 610 (where the mass bias = 0.16%; reference value = 0.9096 ± 0.0008 (Walder et al. 1993),

mean value = 0.9111 ± 0.0052 mean square of the weighted deviates MSWD = 0.22) and natural gem zircon crystals,

(b) the zircon reference sample G91500 (mass bias = 0.19%; reference value = 0.07488 ± 0.00002, age = 1065 ± 1

Ma (Wiedenbeck et al. 1995); mean value = 0.07503 ± 0.00032, age = 1069 ± 9 Ma MSWD = 0.49) and (c) the zircon

reference sample UQ-Z1 (mass bias = 0.29%; reference value = 0.07784 ± 0.00004, age = 1143 ± 1 Ma (Machado

and Gauthier 1996); mean value = 0.07807 ± 0.00019, age = 1148 ± 5 Ma MSWD = 2.8). Error bars represent 2s in

all cases. In (b) the determination shown by the filled symbol was discarded from the age calculation as it includes a

significant proportion of non-radiogenic Pb. In all of these diagrams the dashed line corresponds to the reference value.

Number of measurements

Number of measurements Number of measurements

20

7Pb

/20

6Pb

isot

opic

ra

tio2

07Pb

/20

6Pb

isot

opic

ra

tio

20

7Pb

/20

6Pb

isot

opic

ra

tio

0.067

0.069

0.071

0.073

0.075

0.077

0.079

0.081

0.083

0.085

0 10 20

G91500 zircon reference sample

b

0.70

0.74

0.78

0.82

0.86

0.90

0.94

0.98

1.02

1.06

0 10

NIST SRM 610 Glass CRM a

5 150.068

0.070

0.072

0.074

0.076

0.078

0.080

0.082

0.084

0 10 20 30 40 50

UQ-Z1 zircon reference material

c

(2s) corresponding to an apparent age of 1069 ± 9Ma. The calculated mass bias is almost identical tothat obtained on the NIST SRM 610 glass referencematerial, with a value of 0.19%. The zircon referencesample UQZ1 (Machado and Gauthier 1996) wasalso measured over a period of several months andthe results of more than fifty spots gave a weightedmean 207Pb/206Pb ratio of 0.07807 ± 0.00019 (2s),which results in an apparent age of 1148 ± 5 Ma anda mean calculated mass bias of about 0.3%. Thisslightly higher value is thought to reflect variations inmass bias that are also dependent on the daily opera-ting conditions. In spite of very different optical, thermal,chemical and mineralogical characteristics, the NISTSRM 610 synthetic silicate glass (rhyolitic glass) andnatural zircon crystals (light pink zirconium silicates)yielded similar mass bias values, which suggests that,for the nanosecond UV laser, and at the precision levelachieved with our apparatus, this parameter is notcontrolled by properties of the target material. Therequirement for matrix matched calibration samples,which is so important for quantitative analyses (Jarvisand Williams 1993), is thus not necessary in Pb-Pbisotopic ratio determination. Based on this observation,the NIST SRM silicate glass or the UQ-Z1 zircon wereboth used to correct for mass bias, although, takinginto account the requirements of sample preparation, itwas found easier to include chips of the UQ-Z1 zircontogether with “unknown” zircon crystals in the epoxyresin. During a set of measurements (see sect ion

above), the mean of the first three calibration sampleswas used to correct the first six unknowns. The last sixunknowns were corrected using the mean of the lastthree calibration samples. Each ratio was corrected formass bias following the relationship below:

Rcorr = Rmeas. • (1 + C)δm (1)

where δm is the mass difference and C the mass biascorrecting factor determined from measurements onreference samples.

In contrast to Pb/Pb ratios, for which the mass biascorrection is straightforward and significantly lowerthan the individual analytical precision of each spotanalysis (between 0.5 and 5% at the 1s level), ele-mental fractionation between U and Pb is a moreserious problem. As a consequence of the very diffe-rent behaviour of U relative to Pb during ablation, U-Pb ratios evolve during each individual spot analysis.Figure 2a shows the evolution of the 206Pb/238U ratioagainst time for laser experiments performed with a ca.30 µm spot on the NIST SRM 610, where each repeathad a 2 s duration. Excluding the 10 s initial pulse, thefirst part of the diagram (i.e. from 10 to 30 s), yields aregular inc rease in the 206Pb/238U ra t io , whichappears to be a linear function of time (see inset inFigure 2a). Using a linear least squares fit, the “initial”206Pb/238U ratio can be calculated with a precision ofaround 5% (2s), which compares favourably with the


3 6 5

0.15

0.19

0.23

0.27

0.31

0.35

0.39

0.43

0.47

0.51

0.55

0.59

0 10 20 30 40 50

UQ-Z1 true value��(0.1933)

UQ-Z1��F = 45 µm

b

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 10 20 30 40 50 60

10 15 25 30

0.25

0.2

0.3

0.35

0.4

y = 0.0048x + 0.2298

R2 = 0.9938

20 35

NIST SRM 610

NIST true ratio: 0.2249

a

F = 30 µm

0.15

0.19

0.23

0.27

0.31

0.35

0.39

0.43

10 20 30 40 50

G91500 true value��(0.1791)

G91500��

c0

F = 45 µm

Figure 2. (a) Evolution of the 206Pb/238U ratio

against time for a 30 µm diameter spot (F)

drilled in the NIST SRM 610 glass CRM. After the

first 10 s, which corresponds to the initial pulse,

the 206Pb/238U ratio yielded a linear correlation

with time (inset). Defocusing of the laser beam

and loss of energy coupled to the sample is

responsible for the large increase of the U-Pb

ratio observed (main figure). (b) Plot of 206Pb/238U

ratio against time for the zircon reference sample

UQ-Z1 (Machado and Gauthier 1996) for a spot

diameter (F) of ca. 45 µm; slope of the best-fit

line = 0.0318 ± 0.0060, intercept = 0.3915 ±

0.019, MSWD = 0.29, probability = 0.83. (c) Plot

of 206Pb/238U ratio against time for the zircon

reference sample G91500 (Wiedenbeck et al.

1995) for a spot diameter (F) of ca. 45 µm; slope

of best-fit line = 0.0223 ± 0.0051, intercept

= 0.3429 ± 0.016, MSWD = 0.70, probability

= 0.55. All error bars represent 2s.

20

6Pb

/23

8U

ra

tio

Time (s)

10-15% measurement error. In addition, the magnitudeof the inter-element fractionation correction is signifi-cantly reduced, thus improving accuracy in the deter-mination of the 206Pb/238U ratio. The calculated ratiois then used to define the magnitude of the elementalfractionation, and is corrected by comparison withvalues obtained from a set of measurements performedon a reference material (Figures 2b and c). In the lastpart of the diagram, the 206Pb/238U ratio increasessignificantly and can even reach four times the valueof the true ratio. Absorption of the optical energyinvolves a series of mechanisms that can cause melting,vaporisation, ionisation, ejection of particles (clustersand solid fragments), plasma initiation and expansion.At the laser power density used in this study, we observedthat the true isotopic ratio (0.2249 for NIST SRM 610)was approached in the first measurements but ratioswere always higher, suggesting that laser ablation isdominated by thermal vaporisation. If ionisation werethe main operating mechanism, the 206Pb/238U ratiowould be lower than the true value, since U is easierto ionise (6.194 eV) than Pb (7.417 eV). The regularincrease of the 206Pb/238U ratio is thought to reflectprogressive defocusing of the laser beam as it drillsthrough the sample and an associated loss of energyimparted to the sample. This results in a preferentialenrichment of Pb in the vapour phase due to analmost two times lower latent heat of vaporisation ofPb than for U. Absorption or scattering of a proportionof the laser energy by the optically dense vapourplume above the ablation crater can further cause adecrease in the amount of energy transferred into thesample and thus contribute to the selective enrichmentof Pb in the vapour phase. A plot of the 206Pb/238Uratio against laser flashlamp voltage (Figure 3) showsthat the Pb/U rat ios are more reproducible anddecrease with increasing laser energy. Decreasing theflashlamp voltage causes a decrease in the laser inci-dent power, which results in lower signal intensity andpoorer reproducibili ty. The Pb/U ratio, however, issignificantly different, indicating preferential Pb enrich-ment and, thus, substantiating the suggestion that athermal mechanism appears to govern the laser abla-tion process and could describe the ablation beha-viour and inter-element fractionation at low laserpower density. Horn et al. (2000) observed a similarpositive correlation between Pb/U fractionation anddepth with a 15 ns pulse duration excimer laser (193nm) and proposed that, at this wavelength, a mecha-nism such as selective condensation of refractory ele-ments (U) on the crater walls successfully describesinter-element fractionation. A less efficient ejection of

the particles as the crater deepens can also be invokedat 266 nm and could explain the large increase of theU-Pb ratio in the last part of Figure 2a. From this pointof view, it is interesting to note that the relationshipbetween crater aspect ratio (diameter against depth)and fractionation observed by Eggins et al. (1998) isan important parameter and that the best results areobtained with a ratio close to 1. Although the drillingrate should be different in the case of the NIST SRM610 and natural zircons, it can be seen that the biggerthe spot diameter, the longer the linear correlation.

Interference and common lead correction

Using a dry plasma, interferences are generally nota problem except for isobaric interferences of 204Hgon 204Pb. Hirata and Nesbitt (1995) suggested thatmost of the Hg comes from the argon itself and that fil-tering the gas using charcoal reduced this interferenceby 30 to 90%. In the course of this study, this interfe-rence was corrected by monitoring 202Hg and assu-ming a 204Hg/202Hg ratio of 0.2293. The remaining204Pb was ascribed to common Pb. The difficulty ofmeasuring precisely the small 204Pb isotope generallymakes any common lead correction inaccurate and

3 6 6


0.6

1.1

1.6

2.1

2.6

0 2 4 6 8 10

NIST SRM 610 Glass Flash Lamp: 80%Mean: 2.02 ± 0.28

Flash Lamp: 90%Mean: 1.10 ± 0.07

Figure 3. Measured/true ratios diagram for the U-Pb ratios of

NIST SRM 610 glass CRM at different laser flashlamp voltages.

Adjusting laser flashlamp voltage from 80 to 90% of the

maximum resulted in a higher ablation rate and consequently

a better precision and reproducibility. The 206Pb/238U ratio

tends to be lower at higher flashlamp voltage. Since almost

twice as much energy is needed to vaporise U than Pb, the

amount of vaporised U increases faster with increasing energy,

which can be predicted from a thermal vaporisation model.

Error bars represent 2s.

Measurements

Pb/U

ra

tios

(mea

sure

d/t

rue)

we used therefore a correction threshold based on thelimit of detection (LOD), defined as three times thestandard deviation of the intensity in cps of 204Pb inthe blank. Most analyses (unknowns and calibrationsamples) yielded a 204Pb intensity lower than the LODand, thus, were not corrected for common Pb. For theother analyses, a 204Pb correction was applied, basedon an assumed common Pb composition modelled ascontemporaneous Pb (Stacey and Kramers 1975).Corrected isotopic ratios and weighted averages werethen calculated according to Ludwig (1987). Analyticaluncertainties are listed as 1s and uncertainties in agesas 95% confidence levels. Decay constants are thoserecommended by the IUGS subcommission on geo-chronology (Steiger and Jäger 1977).

Results

The zircon populations selected for this study havewell known ages as they have already been subjectedto U-Pb analyses, either by the conventional ID-TIMSmethod, or by using the SHRIMP, or in some cases byboth. The samples come from a wide variety of rocksand cover an age span ranging from late Archaean toLate Palaeozoic . The two f i rs t examples are f romArchaean rocks of the Jimperding Metamorphic Beltlocated on the western margin of the Yilgarn Craton(Western Australia), which have been investigatedonly for 207Pb/206Pb systematics. U-Pb analyses arepresented for younger material, i.e. the ca. 1 Ga oldG91500 zircon reference sample and Late Palaeozoiczircons from an alkali granite from Corsica (France). Inall cases, the UQ-Z1 zircon crystal (Machado andGauthier 1996) was used as an external calibrationsample for evaluating the magnitude of the mass biasand inter-element fractionation.

207Pb/206Pb isotopic measurements

Samples W400 and W395 (Jimperding Meta-morphic Belt, Western Australia): U-Pb ages fromthese two samples have been obtained previously byID-TIMS and SHRIMP analyses (see details in Bosch etal. 1996 and Pidgeon et al. 1996). W400 is an ortho-gneiss for which crystallisation and emplacement hasbeen dated at 2662 ± 4 Ma by SHRIMP (Figure 4a).The zircon population shows a simple behaviour withall points being concordant to slightly discordant. SEMobservation of SHRIMP spots showed the occurrenceof unzoned recrystallised domains that have statis-tically younger ages (black symbols in Figure 4a).Recrystallisation is attributed to amphibolite facies

regional metamorphism which peaked at ca. 2650Ma in this area (Nemchin et al. 1994). Laser ablationanalyses were performed on non-magnetic zircon crys-tals selected to be flawless, with no visible fractures orinclusions. The 207Pb/206Pb ratios for thirteen spots onthirteen grains are reported in Table 2. All analysesshowed a small amount of 204Pb, but below the LOD(55 cps), except analysis #3, which yielded an unu-sually high 204Pb intensity (246 cps). The appliedcorrection gave a much younger age than the remainder


3 6 7

Figure 4. Isotopic results for the W400 orthogneiss

(Jimperding Metamorphic Belt, Western Australia).

(a) Concordia diagram for SHRIMP analyses. Polygons

represent 1s uncertainties. Unfilled polygons: main euhedral

population (2664 ± 4 Ma); black polygons: unzoned

domains (2640 ± 16 Ma); grey polygon: core (2684 ± 14

Ma). (b) Laser ablation ICP-MS 207Pb/206Pb diagram,

showing a mean value of 0.18050 ± 0.00064 (MSWD

= 0.5), giving an age of 2657 ± 6 Ma. Error bars are 2s.

Number of measurements

207Pb/235U

20

6Pb

/23

8U

20

7Pb

/20

6Pb

isot

opic

ra

tio

a0.480

0.500

0.520

0.540

0.560

11 12 13 14 15

2550

2600

2650

2700

2750

2800

W400 OrthogneissSHRIMP analyses

0.170

0.172

0.174

0.176

0.178

0.180

0.182

0.184

0.186

0.188

0 5 10

Age: 2657 ± 6 Ma

W400 OrthogneissLA-ICP-MS analyses b

3 6 8


Table 2.Laser ablation ICP-MS isotopic results

Measured Corrected atomic ratios3 Apparent age (Ma)

Sample 204Pb1 206Pb2 207Pb/ ±(%) 208Pb/ 232Th/ 206Pb/ ±(%) 207Pb/ 206Pb/ Disc(cps) (cps) 206Pb (1s) 206Pb 238U 238U (1s) 206Pb 238U (%)

CINQUE FRATI- Alkali granite- Corsica- 286 ± 2 Ma (ref. in Cocherie et al. 1999)Properties: large grains (> 200 µm), euhedral, translucentLOD (204Pb): 46 cps5FRATI-1 1 92647 0.05242 0.6 0.2879 0.344 0.0441 7.9 304.0 278.1 9.35FRATI-2 1 106867 0.05157 0.7 0.2811 0.309 0.0454 8.2 266.4 286.2 -6.95FRATI-3 1 100047 0.05208 0.5 0.2793 0.305 0.0452 8.0 289.0 284.7 1.55FRATI-4 20 93551 0.05191 1.1 0.2818 0.331 0.0497 17.8 281.6 312.8 -10.05FRATI-5 22 82964 0.05208 1.0 0.2950 0.331 0.0517 16.8 288.9 325.2 -11.25FRATI-6 1 38821 0.05202 0.6 0.2568 0.291 0.0414 9.0 286.2 261.6 9.45FRATI-7 3 43933 0.05098 2.3 0.2430 0.242 0.0472 7.6 239.9 297.5 -19.45FRATI-8 21 39309 0.05198 1.6 0.2679 0.294 0.0419 7.2 284.6 264.3 7.75FRATI-9 1 47937 0.05303 1.0 0.2924 0.268 0.0446 7.7 330.0 281.2 17.35FRATI-10 1 19464 0.05083 3.1 0.2443 0.273 0.0415 8.7 233.1 262.0 -11.05FRATI-11 13 22476 0.05530 2.0 0.2482 0.217 0.0423 6.8 424.2 267.3 58.75FRATI-12 16 22517 0.05193 1.8 0.2434 0.265 0.0421 10.1 282.5 265.6 6.35FRATI-13 7 25856 0.05385 1.7 0.2853 0.298 0.0426 8.4 364.7 268.8 35.75FRATI-14 6 36470 0.05194 0.9 0.2731 0.300 0.0421 8.5 282.9 265.9 6.45FRATI-15 7 31222 0.05156 2.2 0.2500 0.251 0.0426 7.7 265.8 268.9 -1.25FRATI-16 19 17985 0.05503 2.3 0.2929 0.290 0.0433 12.9 413.3 273.1 51.45FRATI-17 5 11796 0.05380 2.3 0.1676 0.186 0.0400 8.4 362.9 253.0 43.45FRATI-18 11 30939 0.05149 1.3 0.2459 0.244 0.0476 9.9 263.1 299.6 -12.25FRATI-19 8 57897 0.05095 0.4 0.2280 0.306 0.049 6.9 238.6 311.4 -23.4Mean: - - 0.05236 - 0.2613 0.281 0.0445 - 300.1 280.4 -SD (1s): - - 0.00128 - 0.0308 0.040 0.0033 - 55.0 20.2 -

G91500- Zircon standard- Canada-1065 ± 1 Ma (ref. in Wiedenbeck et al. 1995)Properties: large fragments (mm), light pinkLOD (204Pb): 25 cpsG91500-1 4 31336 0.07500 1.2 0.0989 0.213 0.1818 2.3 1068.4 1077.0 -0.8G91500-2 1 31208 0.07509 1.0 0.0983 0.222 0.1906 2.4 1070.9 1124.4 -4.8G91500-3 1 28836 0.07458 0.6 0.0987 0.228 0.1817 2.3 1057.1 1076.1 -1.8G91500-4 17 29209 0.07542 1.7 0.0997 0.219 0.1707 8.9 1079.8 1015.8 6.3G91500-5 1 28324 0.07475 0.8 0.0987 0.221 0.1744 10.4 1061.9 1036.0 2.5G91500-6 10 27761 0.07553 1.0 0.0977 0.233 0.1760 2.4 1082.7 1045.2 3.6G91500-7 1 28608 0.07374 0.9 0.0954 0.224 0.1714 2.4 1034.4 1019.6 1.5G91500-8 4 29148 0.07453 1.0 0.0963 0.227 0.1814 6.9 1055.8 1074.6 -1.7G91500-9 1 28878 0.07518 0.9 0.0977 0.224 0.1796 2.3 1073.4 1064.8 0.8G91500-10 1 29056 0.07323 1.2 0.0957 0.229 0.1753 2.4 1020.2 1041.3 -2.0G91500-11 1 30151 0.07523 0.6 0.0964 0.234 0.1757 2.4 1074.6 1043.5 3.0G91500-12 1 23074 0.07308 2.2 0.0939 0.233 0.1778 12.1 1016.3 1055.1 -3.7G91500-13 14 25982 0.07405 1.6 0.0933 0.228 0.1696 3.4 1042.9 1010.0 3.3G91500-14 1 24916 0.07478 1.1 0.0912 0.227 0.1761 7.9 1062.6 1045.5 1.6G91500-15 13 7719 0.07500 3.5 0.1125 0.253 0.1782 2.5 1068.5 1057.3 1.1G91500-16 17 11218 0.07525 0.7 0.1142 0.233 0.1929 2.3 1075.1 1137.0 -5.4Mean: - - 0.07465 - 0.0987 0.228 0.1783 - 1059.0 1057.7 -SD (1s): - - 0.00075 - 0.0062 0.009 0.0064 - 20.4 35.1 -

W395- Syenite- Western Australia- 2654 ± 5 Ma (ref. in Pidgeon et al. 1996)Properties: small grains (+105-135 µm), euhedral, pink, non magneticLOD (204Pb): 51 cpsW395-1-1 21 224411 0.19282 3.6 0.1135 0.311 - - 2766.4 - -W395-1-2 57 226556 0.21508 0.2 0.0938 0.270 - - 2944.3 - -W395-3-1 67 443762 0.24614 1.0 0.4484 1.619 - - 3160.1 - -W395-4-1 27 304101 0.25109 2.1 0.0788 0.241 - - 3191.7 - -W395-5-1 18 495455 0.18066 0.2 0.1323 0.440 - - 2658.9 - -W395-6-1 3 305002 0.17980 0.2 0.0683 0.263 - - 2651.0 - -W395-6-2 3 150164 0.20281 0.9 0.0909 0.326 - - 2848.9 - -

and this point was, therefore, discarded from the agecalculation. The twelve remaining analyses were com-b ined to g i ve a we igh ted mean o f 0 .18050 ±0.00064 (2s) corresponding to an age of 2657 ± 6Ma (Figure 4b). The laser ablation value is very closeto the 2662 Ma SHRIMP reference value and shows

comparable error margins. Unlike the SHRIMP analyses,LA-ICP-MS analyses were not capable of detecting theyounger recrystallised domains, although some youngages would suggest that such domains could havebeen sampled by the laser beam. The short time spanbetween igneous crystallisation and recrystallisation,


3 6 9

Table 2 (continued).Laser ablation ICP-MS isotopic results

Measured Corrected atomic ratios3 Apparent age (Ma)

Sample 204Pb1 206Pb2 207Pb/ ±(%) 208Pb/ 232Th/ 206Pb/ ±(%) 207Pb/ 206Pb/ Disc(cps) (cps) 206Pb (1s) 206Pb 238U 238U (1s) 206Pb 238U (%)

W395-7-1 1 68423 0.17953 1.1 0.0803 0.285 - - 2648.5 - -W395-7-2 23 363556 0.18070 0.6 0.0793 0.208 - - 2659.3 - -W395-8-1 1 15288 0.25785 0.8 0.1778 0.479 - - 3233.6 - -W395-9-1 1 321503 0.18307 0.4 0.0882 0.241 - - 2680.9 - -W395-9-2 3 273131 0.18086 0.6 0.1071 0.246 - - 2660.7 - -W395-9-3 22 538830 0.18143 0.8 0.0871 0.205 - - 2666.0 - -W395-10-1 1 86592 0.18084 0.6 0.1235 0.461 - - 2660.6 - -W395-11-1 38 418562 0.23796 3.4 0.0753 0.219 - - 3106.4 - -W395-12-1 3 52307 0.17792 1.0 0.1579 0.908 - - 2633.6 - -W395-12-2 9 161414 0.17622 0.8 0.1132 0.261 - - 2617.6 - -W395-12-3 1 148068 0.17890 0.9 0.1250 0.306 - - 2642.7 - -W395-13-1 1 83300 0.17654 1.7 0.0566 0.168 - - 2620.7 - -W395-14-1 31 161352 0.17652 0.8 0.1297 0.348 - - 2620.5 - -W395-15-1 20 80342 0.17608 1.2 0.1150 0.310 - - 2616.3 - -W395-15-2 1 27094 0.17751 1.0 0.1027 0.264 - - 2629.8 - -W395-16-1 18 54051 0.24584 0.3 0.1084 0.399 - - 3158.2 - -W395-16-2 25 41186 0.24422 1.0 0.1101 0.317 - - 3147.7 - -W395-17-1 1 31956 0.18083 0.8 0.0851 0.215 - - 2660.5 - -W395-17-2 17 73407 0.18011 0.7 0.1471 0.367 - - 2653.9 - -Mean (young): - - 0.17927 - 0.1058 0.323 - - 2646.0 - -SD (1s): - - 0.00211 - 0.0283 0.171 - - 19.5 - -

W400- Orthogneiss- Western Australia- 2662 ± 4 Ma (ref. in Bosch et al. 1996)Properties: small grains (+135-180 µm), euhedral, light brown, non magneticLOD (204Pb): 48 cpsW400-1 14 271020 0.18011 0.8 0.2315 0.821 - - 2653.9 - -W400-2 23 242838 0.18030 0.4 0.0972 0.300 - - 2655.7 - -W400-3 246 233821 0.17709 1.8 0.1469 0.368 - - 2625.8 - -W400-4 23 233087 0.18062 1.2 0.1347 0.426 - - 2658.5 - -W400-5 1 93040 0.17995 0.3 0.1150 0.385 - - 2652.4 - -W400-6 40 163381 0.18217 0.5 0.1572 0.541 - - 2672.7 - -W400-7 9 76292 0.18066 0.8 0.1088 0.330 - - 2659.0 - -W400-8 1 66784 0.18112 1.0 0.1657 0.514 - - 2663.2 - -W400-9 14 73459 0.18278 1.3 0.0940 0.308 - - 2678.3 - -W400-10 7 106168 0.18055 1.1 0.1050 0.272 - - 2658.0 - -W400-11 23 141211 0.18000 0.8 0.1926 0.624 - - 2652.9 - -W400-12 51 210877 0.18163 1.1 0.1517 0.360 - - 2667.9 - -W400-13 23 94853 0.17969 1.7 0.1732 0.487 - - 2650.0 - -Mean: - - 0.18080 - 0.1439 0.447 - - 2660.2 - -SD (1s): - - 0.00096 - 0.0426 0.161 - - 8.7 - -

All zircon grains have been taken from the least magnetic bulk fractions.1 corrected for background and Hg isobaric interference based on a 204Hg/202Hg = 0.2293.2 corrected for background.3 lead isotopic ratios have been corrected for background, mass bias by reference to UQ-Z1 zircon reference sample (Machado andGauthier 1996) and initial common Pb (after Stacey and Kramers 1975) for measurements where 204Pb was found to be greater than the determined LOD. U-Pb ratios have been calibrated against a 206Pb/238U ratio of 0.1933 calculated by averaging the most concordant analyses given in Machado and Gauthier (1996). The 207Pb/235U ratios were calculated using the corrected 207Pb/206Pb ratio and according to a 238U/235U ratio of 137.88. Errors are 1s and refer to last digits.The right hand column (“disc”) is the percentage discordance assuming recent lead losses.

and the low precision of the LA-ICP-MS analyses makethe age difference unresolvable.

The syenite W395 outcrops in the same area andwas analysed because of a more complex zirconpopulation with evidence of an inherited component inthe source region of the magma (Figure 5a). SHRIMPanalyses yielded an age of 2654 ± 5 Ma, interpretedas dating emplacement and crystallisation of the syenite.In addition, SHRIMP spot analyses identified an inheri-ted component with ages reaching values of about3.25 Ga and very discordant analyses between 2660and 3250 Ma. Seventeen grains were analysed byLA-ICP-MS and were found to contain a small amountof 204Pb, below the LOD, except for two spots (#1-2

and 3-1) for which the beam may have struck a crackin the grain or an inclusion. The 207Pb/206Pb agespectrum, ranges from 2616 ± 88 Ma (2s) to 3233 ±56 Ma (2s), and confirms the age pattern observed bySHRIMP analyses. A histogram showing the distributionof the 207Pb/206Pb ratios (Figure 5b) indicates thatmost analyses cluster close to 2650 Ma with a weigh-ted mean of 0.18002 ± 0.00063 (2s), correspondingto an age of 2653 ± 6 Ma. This age is in close agree-ment with the SHRIMP value of 2654 ± 5 Ma. Anotherinteresting result is the identification of the inheritedcomponent, with the oldest LA-ICP-MS value (3233Ma) being also very close to the oldest SHRIMP value(3250 Ma). Results from grain #6, which was probedin two places, also demonstrate that the spatial resolu-tion available with our laser system has the capabilityto analyse distinct domains present within one crystal(see Table 2), although the intermediate value of 2848Ma for the core (#6-2) may indicate that the beamstraddled the magmatic growth zone.

206Pb/238U isotopic measurements

Sample G91500 (Ontario, Canada): Fragmentsof this 238 g zircon crystal are used as a referencesample for the calibration of the U-Pb ratios measuredby SIMS on the CAMECA IMS1270 commissioned atthe CRPG Nancy (France) and at Stockholm (Sweden).The crystal has been calibrated by different laborato-ries (see Wiedenbeck et al. 1995) and is known to behomogeneous and concordant at about 1065 Mawith a 207Pb/206Pb of 0.07488 ± 0.00002 (2s) and a206Pb/238U ratio of 0.1792 ± 0.0002 (2s). Sixteenspots were measured on several fragments of thissample (see Table 2) and were all found to be lessthan 10% discordant. The measured 204Pb counts werevery low (less than 20 cps) in good agreement withthe low common lead content of this crystal (seeWiedenbeck et al. 1995). The corrected 206Pb/238Uages range from 1010 ± 30 Ma to 1137 ± 23 Ma.Reported on the concordia diagram (Figure 6), ana-lyses define a discordia line which has an upperintercept age of 1060 ± 36 Ma and a lower interceptnot significantly different from zero (300 ± 960 Ma),which is the result of a long extrapolation. All pointscan be combined to give a 206Pb/238U weightedaverage of 0.1781 ± 0.0026 (MSWD = 1.13) corres-ponding to a mean age of 1057 ± 14 Ma (2s). ThisLA-ICP-MS age is well within the error of the TIMSvalue of 1062.4 ± 0.4 Ma (Wiedenbeck et al. 1995).Two analy ses (# 2 and 16) have s l igh t l y o lder206Pb/238U ages, but can be considered as reversely

3 7 0


Figure 5. Isotopic results for the W395 syenite (Jimperding

Metamorphic Belt, Western Australia). (a) Concordia diagram

for SHRIMP analyses. Polygons represent 1s uncertainties.

(b) Histogram showing the distribution of the 207Pb/206Pb ages

measured by laser ablation ICP-MS.

Age (Ma)

207Pb/235U

20

6Pb

/23

8U

Freq

uenc

y

0.45

0.55

0.65

0.75

10 12 14 16 18 20 22 24 26 28

2600

2800

3000

3200

3400

0.46

0.48

0.50

0.51

0.53

12

13

2550

2600

26502700

2654 ± 5 Ma

Weighted Av.

207Pb / 206Pb

a W395 SyeniteSHRIMP analyses

2650 2750 2850 2950 3050 3150 3250

2

0

4

6

8 W395 SyeniteLA-ICP-MS analyses

b

Age: 2653 ± 6 Man = 16

Mean: 0.18002 ± 0.00063(MSWD = 1.9)

discordant ( see Table 2) . Th is feature , which i sfrequently observed during in situ analyses (Williamset al. 1984), can also be ascribed to a quick drift inthe U-Pb fractionation.

Sample Cinque Frat i : Th is sample i s a La tePalaeozoic alkali granite from Corsica which has beenpreviously dated by ID-TIMS at 286.4 ± 1.8 Ma andby SHRIMP at 284.8 ± 2.3 Ma (Cocherie et al. 1999).The TIMS value corresponds to a 206Pb/238U ratio of0.0454 ± 0.0003 (2s). Nineteen spots were measuredby laser ablation on large, translucent crystals (> 200µm) devoid of fractures or inclusions. All analysesyielded very low 204Pb counts (< 25 cps) below theLOD and no common lead correction was applied tothese measurements. The nineteen analyses plot on orclose to the concordia curve (Figure 7) with 206Pb/238Uages ranging from 253 ± 21 Ma to 325 ± 53 Ma (seeTable 2). Because of their tight clustering, no discordiacan be calculated. A weighted mean for all spots givesa 206Pb/238U ratio of 0.0439 ± 0.0017 (MSWD =0.53) corresponding to an age of 277 ± 11 Ma (2s).Although slightly younger, the mean LA-ICP-MS ageoverlaps the more precise TIMS and SHRIMP values.

Discussion and conclusions

The results presented in this study demonstrate thatcoupling a 266 nm UV laser to a quadrupole ICP-MScan be succes s fu l l y u sed to de te rm ine in s i t u207Pb/206Pb and 206Pb/238U ratios for geochronologi-cal purposes. The external calibration can be achievedusing either a naturally occurring zircon crystal or amore widely available synthetic glass reference material.Both materials yielded comparable mass bias valuesand could thus be used to correct the 207Pb/206Pbratios. Measurements could be routinely achieved onsingle zircon grains with a spatial resolution of ca.40-50 µm without the need for matrix matched refe-rence materials. As a single measurement requiredonly 1 minute of data acquisition, a comprehensivestudy of a zircon population could be achieved withina few hours. This is the case even for heterogeneouszircon populations, such as those from granitic rockswith multiple inherited components and/or sedimentaryrocks, which often result from erosion of a wide varietyof source rocks, both in terms of ages and composi-tions. Analysis of the 207Pb/206Pb ratios in these zirconpopulations can thus provide a wealth of informationto pinpoint the deep-seated source regions of magmasor to fingerprint the sources of clastic sediments, bothof which may have important geodynamic implications.The high throughput rate of the technique is, in theseapplications, a major advantage over TIMS or SIMStechniques. Open system behaviour, which is frequentlyobserved on zircon but also on monazite, is a seriousdrawback but can be minimised by selecting the least


3 7 1

Figure 6. Concordia diagram for LA-ICP-MS analyses of fragments

of the G91500 zircon crystal. All analyses are represented by

ellipses sized according to their 2s uncertainties. The weighted

mean 207Pb/206Pb ratio = 0.07479 ± 0.00035 (MSWD = 0.7),

representing an age of 1063 ± 9 Ma. The 206Pb/238U

weighted mean age = 1057 ± 14 Ma.

207Pb/235U

20

6Pb

/23

8U

Figure 7. Concordia diagram for LA-ICP-MS analyses of zircons

from the Late Palaeozoic Cinque Frati alkali granite (Corsica,

France). All analyses are represented by ellipses sized according

to their 2s uncertainties. The weighted mean 207Pb/206Pb ratio

= 0.05178 ± 0.00036 (MSWD = 2.8), representing an age of

276 ± 16 Ma. The 206Pb/238U weighted mean age = 277 ± 11 Ma.

207Pb/235U

20

6Pb

/23

8U

1140

1060

980

900

0.145

0.155

0.165

0.175

0.185

0.195

0.205

0.8 1.2 1.6 2.0 2.4 2.8

Upper Intercepts at:1060 ± 36 Ma(MSWD = 0.1)

To 309 ± 960 Ma

G91500 (Canada)Zircon

206Pb/238U AgeWeighted Mean: 1057 ± 14 Ma

400

500

300

200

100

0.00

0.02

0.04

0.06

0.08

0.0 0.2 0.4 0.6

Cinque Fratti (Corsica)Zircon

206Pb/238U Age

Weighted Mean: 277 ± 11 Ma

magnetic grains. Potentially discordant analyses canbe detected and discarded from age calculation, asthey result in a spread of ages towards lower valuesand skewed 207Pb/206Pb histograms. The 204Pb isotopecannot be measured accurately by ICP-MS because ofboth very low counts on this isotope and a significantHg interference. We are, therefore, generally unable tocorrect precisely for the common lead contributionusing the 206Pb/204Pb ratio (see Table 2). This howeverdoes not seem to be a problem for laser ablation ana-lyses and generally we observed that the measured207Pb/206Pb values were very close to the expectedvalue. This is partly due to the high sensitivity of theICP-MS, which results in a high count rate on the206Pb isotope ( > 20 000 cps and up to 400 000cps) and to the fact that pristine zircons are usuallydevoid of common lead. For Archaean and Proterozoiczircons, the common lead contribution is generallyundetectable.

The external calibration technique used in thisstudy allowed us to control the high inter-element frac-tionation observed between U and Pb and to correctthis fract ionation to achieve a precision down toaround 5%. This is due to the linear correlation obser-ved between the variation of the 206Pb/238U ratio andtime during ablation. The calculated ratio showed amuch better precision than the measured ratio. Thissimple technique allowed the age determination ofyoung material for which the 207Pb/206Pb ratio couldnot be used. Due to the high sensitivity of the ICP sour-ce, this opens up new possibilities in the dating of veryyoung rocks with a precision that should exceed thecapabilities of SIMS analyses. In terms of precision andaccuracy, analyses performed during the course of thisstudy using matrix matched calibration samples sho-wed comparable or slightly worse results than thoseobtained by external calibration using a NIST glassCRM (Hirata and Nesbitt 1995) or a liquid calibration(Horn et al. 2000) respectively. Thus, the capability oflaser ablation ICP-MS analyses to determine U-Pbages without the need to use well calibrated naturallyoccurring gem-quality reference samples constitutes amajor advantage over matrix-dependent SIMS ana-lyses, particularly when investigating other accessoryminerals for which phase mineral relationships are betterdocumented than for zircon (e.g. sphene, allanite) or evenon high U-Pb ratio calcretes or biogenic phosphates.

An important problem inherent to the methodis related to the quality of the grains selected foranalyses. Several experiments on zircon populations

containing fractures or inclusions revealed that catas-trophic ablation often occurred, during which grainsshattered, even when tightly embedded in epoxy.Ejection of the large fragments led to spiky signals thatproduced high relative standard deviations. In suchcases, the ablation behaviour did not match that of thecalibration sample and this precluded any acceptablecalibration from being made. Grains must, therefore,be carefully selected from the non-magnetic concen-trate, in a similar manner to the conventional isotopedilution method, and the laser beam must be focusedon homogeneous grain domains. There is, therefore, arisk of analyses biased towards the highest qualitycrystals from the zircon population yielding an unre-presentative age spectrum. Shorter wavelength lasersoperating in the far UV region have the smoothestablation behaviour and are expected to reduce thisdrawback to a significant extent.

The spatial resolution used in this study was abouttwo times worse than that normally achieved by SIMS(20-30 µm), which can be a disadvantage for highspatial resolution studies (overgrowth, recrystalliseddomains). However, previous studies have shown that,providing sensitivity can be enhanced, comparable(Horn et al. 2000) or even smaller (Hirata and Nesbitt1995) spot sizes can be achieved. Our results, obtai-ned on Palaeozoic and Precambrian zircons, show thathigh precision age measurements can be obtained byLA-ICP-MS instrumentation that is more widely avai-lable than high resolution secondary ion microprobes.

Acknowledgements

This work was supported by a BRGM post-doctoralgrant to the first author. Thanks are due to NunoMachado (UQAM, Montréal, Canada) and EtienneDeloule (CRPG, Nancy, France) for providing chips ofthe zircon standards UQ-Z1 and G91500 respectively.Constructive reviews by two anonymous reviewershelped improve this manuscript.

References

Bosch D., Bruguier O. and Pidgeon R.T. (1996)The evolution of an Archaean Metamorphic Belt: Aconventional and SHRIMP U-Pb study of accessory minerals from the Jimperding Metamorphic Belt, YilgarnCraton, Western Australia. Journal of Geology, 104,695-711.

Compston W. and Pidgeon R.T. (1986)Jack Hills, evidence of more very old zircons in WesternAustralia. Nature, 321, 766-769.

3 7 2


references

Cocherie A., Rossi P., Fanning M. and Guerrot C. (1999)Comparative use of TIMS and SHRIMP for zircon dating:A reappraisal of the estimation of age uncertainties.Some examples from Corsica, France. In: Barbarin ?(ed.), 4th Hutton Symposium on the Origin of Granitesand Related Rocks, Clermont-Ferrand, France, BRGMedition, 290, 206.

Eggins S.M., Kinsley L.P.J. and Shelley J.M.M. (1998)Deposition and element fractionation processes duringatmospheric pressure laser sampling for analysis by ICP-MS.Applied Surface Science, 127, 278-286.

Feng R., Machado N. and Ludden J.N. (1993)Lead geochronology of zircon by laser probe-inductivelycoupled plasma-mass spectrometry (LP-ICP-MS).Geochimica et Cosmochimica Acta, 57, 3479-3486.

Fryer B.J., Jackson S.E. and Longerich H.P. (1993)The application of laser ablation microprobe-inductivelycoupled-mass spectrometry (LAM-ICP-MS) to in situ (U)-Pbgeochronology. Chemical Geology, 109, 1-8.

Hirata T. and Nesbitt R.W. (1995)U-Pb isotope geochronology of zircon: Evaluation of thelaser probe-inductively coupled plasma spectrometrytechnique. Geochimica et Cosmochimica Acta, 59,2491-2500.

Hirata T. (1996)Lead isotopic analyses of NIST standard reference materials using multiple collector inductively coupledplasma-mass spectrometry coupled with a modifiedexternal correction method for mass discrimination effect.Analyst, 121, 1407-1411.

Hirata T. (1997)Soft ablation technique for laser ablation-inductively coupled plasma-mass spectrometry. Journal of AnalyticalAtomic Spectrometry, 12, 1337-1342.

Horn I., Rudnick R.L. and McDonough W. (2000)Precise elemental and isotope ratio determination bysimultaneous solution nebulization and laser ablation-ICP-MS: Application to U-Pb geochronology. ChemicalGeology, 167, 405-425.

Jarvis K.E. and Williams J.G. (1993)Laser ablation inductively coupled plasma-mass spectrometry (LA-ICP-MS): A rapid technique for the direct,quantitative determination of major, trace and rare-earthelements in geological samples. Chemical Geology,106, 251-262.

Jeffries T.E., Jackson S.E. and Longerich H.P. (1998)Application of a frequency quintupled Nd:YAG source (λ = 213 nm) for laser ablation ICP-MS analysis of minerals.Journal of Analytical Atomic Spectroscopy, 13, 935-940.

Longerich H.P., Günther D. and Jackson S.E. (1996)Elemental fractionation in laser ablation inductively coupled plasma-mass spectrometry. Fresenius’ Journal ofAnalytical Chemistry, 355, 538-542.

Ludden J.N., Feng R., Gauthier G., Stix J., Shi L.,Francis D., Machado N. and Wu G. (1995)Applications of LAM-ICP-MS analysis to minerals.Canadian Mineralogist, 33, 419-434.

Ludwig K.R. (1987)Isoplot200, a plotting and regression program for isotope geochemists, for use with HP series 200 computers. USGS Open-file report 85-513.

Machado N. and Gauthier G. (1996)Determination of 207Pb/206Pb ages on zircon and monazite by laser ablation ICP-MS and application to astudy of sedimentary provenance and metamorphism insoutheastern Brazil. Geochimica et Cosmochimica Acta,60, 5063-5073.

Machado N., Schrank A., Noce C.M. and Gauthier G. (1996)Ages of detrital zircon from Archaean-Paleoproterozoicsequences: Implications for Greenstone Belt setting andevolution of a Transamazonian foreland basin inQuadrilatero Ferrifero southeast Brazil. Earth andPlanetary Science Letters, 141, 259-276.

Nemchin A.A., Pidgeon R.T. and Wilde S.A. (1994)Timing of Late Archaean granulite facies metamorphismin the southwestern Yilgarn Craton of Western Australia:Evidence from U-Pb ages of zircons from mafic granulites.Precambrian Research, 68, 307-321.

Pidgeon R.T., Bosch D. and Bruguier O. (1996)Petrogenetic implications of inherited zircon and titanite inthe Archaean Katrine syenite, Southwestern YilgarnCraton, Western Australia. Earth and Planetary ScienceLetters, 141, 187-198.

Scott D.J. and Gauthier G. (1996)Comparison of TIMS (U-Pb) and laser ablation microprobeICP-MS (Pb) techniques for age determination of detritalzircons from Paleoproterozoic metasedimentary rocksfrom northeastern Laurentia, Canada, with tectonic implications. Chemical Geology, 131, 127-142.

Stacey J.S. and Kramers J.D. (1975)Approximation of terrestrial lead isotope evolution by atwo stage model. Earth and Planetary Science Letters, 6,15-25.

Steiger R.H. and Jäger E. (1977)Subcommission on geochronology: Convention on theuse of decay constants in geo- and cosmochronology.Earth and Planetary Science Letters, 36, 359-362.

Walder A.J., Abell I.D., Platzner I. and Freedman P.A. (1993)Lead ratio measurement of NIST 610 glass by laser ablation inductively coupled plasma-mass spectrometry.Spectrochimica Acta, 48, 397-402.

Wiedenbeck M., Allé P., Corfu F., Griffin W.L., Meier M., Oberli F., Von Quadt A., Roddick J.C.and Spiegel W. (1995)Three natural zircon standards for U-Th-Pb, Lu-Hf, traceelement and REE analyses. Geostandards Newsletter, 19, 1-23.

Williams I.S., Compston W., Black L.P., Ireland T.R.and Foster J.J. (1984)Unsupported radiogenic Pb in zircon: A cause of anomalously high Pb-Pb, U-Pb and Th-Pb ages.Contributions to Mineralogy and Petrology, 88, 322-327.


3 7 3

Documents

Evaluation of Pb-Pb and U-Pb Laser Ablation ICP-MS Zircon