JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY FEBRUARY 1993 VOL. 8 19 Isotope Ratio Measurement of Lead Neodymium and Neodymium-Samarium Mixtures Hafnium and Hafnium-Lutetium Mixtures With a Double Focusing Multiple Collector Inductively Coupled Plasma Mass Spectrometer* Andrew J. Walder and 1. Platzneri Fisons Instruments VG Elemental Ion Path Road Three Winsford Cheshire UK CW7 3BX Philip A. Freedman Fisons Instruments VG Isotech Aston Way Middlewich Cheshire UK CWlO OHT An inductively coupled plasma (ICP) source coupled to a double focusing magnetic sector mass analyser equipped with seven Faraday detectors has been used to measure the isotopic ratios of lead reference materials using a thallium correction technique. The addition of thallium to the lead standards and the subsequent simultaneous measurement of the thallium and lead isotopes allowed a correction for mass discrimination.Measurement of six samples of National Institute of Standards and Technology (NIST) (formerly National Bureau of Standards) Standard Reference Material Pb-982 revealed 208Pb:204Pb = 36.702 f 0.022 207Pb:204Pb= 17.1 41 f 0.009 and 206Pb:204Pb=36.71 1 2 0.021 compared with the NIST certified values of 36.744 k 0.050 17.1 59 f 0.025 and 36.738 f 0.037 respectively. All errors are given as 20. The measurement of each sample used approximately 200 ng of lead and took 100 s. Measurement of six neodymium samples (obtained from Scripps Institute of Oceanography La Jolla CA USA) where internal normalization to 146Nd:'45Nd is possible revealed a 143Nd:144Nd ratio of 0.51 1825 f 0.000039 compared with the accepted value of 0.51 1859.Measurement of six La Jolla neodymium samples contaminated with a similar concentration of samarium revealed a 143Nd:'44Nd ratio of 0.51 1854 rf 0.000059. The measurement of each mixed sample used approximately 100 ng of neodymium. Similarly measurement of six samples of Johnson Matthey Company (JMC) 475 Hafnium normalized to 179Hf:'77Hf revealed a 176Hf:177Hf ratio of 0.2821 94 f 0.000020 compared with the accepted value of 0.2821 95 rf 0.00001 5. Measurement of six samples of JMC 475 Hafnium contaminated with a similar concentration of lutetium revealed a 176Hf:'77Hf ratio of 0.28221 3 k 0.000023. The measurement of each mixed sample used approximately 400 ng of hafnium. Keywords Inductively coupled plasma mass spectrometry; magnetic sector mass spectrometry; lead isotopic ratio measurement; thallium based mass discrimination correction; neodymium and hafnium isotope ratio measurement The measurement of isotopic ratios using an inductively coupled plasma (ICP) source coupled to a double focusing magnetic sector mass analyser equipped with seven Faraday detectors has previously been described.' The use of several Faraday detectors allows each isotope to be measured simultaneously thus removing signal noise as a limitation on analytical precision.A magnetic sector mass analyser produces flat topped peaks which allows the accurate measurement of each isotope. Isotopic ratio measurements of uranium and lead standard reference materials (SRMs) demonstrated levels of precision comparable to that ob- tained by thermal ionization mass spectrometry (TIMS).For example a relative standard deviation (RSD) of 0.022% was obtained from the 206Pb:204Pb measurement of six samples of National Bureau of Standards (NBS) [now National Institute of Standards and Technology (NIST)] Standard Reference Material (SRM) Pb-98 1. In common with all mass spectrometry plasma ion sources this new instrument transmits the heavier isotope in preference to the lighter; this mass discrimination effect is time independent and can be experimentally established. A solution of known isotopic composition was therefore analysed prior to the analysis of a sample to determine the magnitude ofmass bias. Lead has four naturally occurring isotopes three of which *08Pb 207Pb and loaPb are radioactive decay pro- ducts the fourth 204Pb is not of radiogenic origin and is *Presented in part at the 40th American Society for Mass Spectrometry Conference on Mass Spectrometry and Allied Topics Washington DC USA May 31-June 5 1992.?Visiting scientist at VG Elemental Winsford Cheshire UK; permanent address NRCN PO Box 9001 Beer-Sheva Israel. considered a stable reference isotope. The measurements of lead isotopic ratios are therefore used in geological and environmental studies. The measurement of lead isotopic ratios by quadrupole based ICP mass spectrometry (ICP- MS) is generally inferior to TIMS. Methods and applica- tions have been described in recent publications by Date and Gray2 and Jarvis et al.3 External precisions of approxi- mately 0.5% are possible by ICP-MS compared with 0.05% by TIMS.Existing plasma technology is therefore inade- quate for most geological dating purposes. Levels of precision are acceptable in some areas of environmental studies for example in determining the environmental effects of lead additives within petrol. Ketterer et al.4 have described how the measurement of the isotopic ratio of thallium (205Tl:203Tl) can be used as a correction for the measurement of lead isotopic ratios using a quadrupole based ICP mass spectrometer. This technique originally suggested by Longerich et al.5 improved the precision of lead isotopic ratio analysis and removed the need for the analysis of calibration standards for mass bias corrections. They achieved RSDs of 0.2% a level suitable for the routine analysis of environmental samples.An objective of this study was to apply the thallium correction technique to the analysis of lead SRMs using the plasma based multi-collector mass spectrometry system. The previous study' also included isotope ratio measure- ment of NIST SRM Sr-987. Measurement of the non- radiogenic 86Sr:88Sratio allows an internal calibration for mass discrimination which was used to correct the mea- sured 87Sr:86Sr ratio. This correction procedure was applied as the measurement proceeded thus small variations in mass bias which limit the attainable precision of uranium and lead measurements were removed. Superior levels of20 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY FEBRUARY 1993 VOL. 8 precision were therefore achieved.An RSD of 0.008% was obtained from the 87Sr:86Sr measurement of ten samples of SRM Sr-987. A second objective of this study was to extend this correction procedure to the isotope ratio measurement of other elements. Neodymium and hafnium reference ma- terials were selected because of their importance to the geological community as chronometers6 and because their isotope ratios are well established by TIMS. However natural samples of neodymium and hafnium will contain samarium and lutetium respectively thus the measure- ment of their isotope ratios must be corrected for the presence of these interfering species. Correction procedures are well established by TIMS and it was a further objective of this study to apply these correction procedures to these dual element systems using the plasma source MS system.Experimental Instrumentation The mass spectrometer system is equipped with an ICP source similar to that used on the VG PlasmaTrace mass spectrometer. The sampling and skimmer cones are both held at 5700 V which provides the acceleration potential for the ions as they enter the mass spectrometer. The circular ion beam is matched to a vertical slit profile required by the mass spectrometer using two d.c. quadrupole lens doublets. The adjustable slit at this point is the entrance definingslit on the PlasmaTrace instrument. To achieve the high abundance sensitivity necessary for the analysis of minor isotopes a further stage of differential pumping is incorporated. A second (0.3 mm wide x 5 mm high) defining slit is provided the first slit being imaged onto the second using a compound electrostatic lens.This aperture forms the entrance to the double focusing mass spectrometer. The mass spectrometer achieves a 540 mm dispersion and incorporates seven adjustable Faraday collectors on an image plane normal to the optic axis. Using 1 mm defining slits at the entrance of each collector the resolution of the instrument is fixed at 400 (5% valley definition). As the image of the source defining slit is significantly narrower than the collector slits the instrument peak shape exhibits a 'top hat' shape the central section showing a high degree of flat peak. This ensures that the recorded signal amplitude is insensitive to external events and allows a high degree of accuracy to be obtained in the measurement.This is in contrast to the Gaussian peak-shape characteristic of quadrupole based mass spectrometers. By simultaneously measuring each isotope of interest the effects of plasma noise are eliminated and hence very precise isotope ratio measurements are possible. The collector system is equipped with seven Faraday detectors which are refer- enced as low 2 low 1 axial high 1 high 2 high 3 and high 4. Analytical Procedure Samples were introduced into the plasma with a peristaltic pumping system via a Meinhard nebulizer. The preampli- fiers associated with each detector were calibrated with respect to the axial preamplifier at the beginning of each days analysis. No correction for variations in detector efficiency were made.The NIST SRMs Pb-981 Pb-982 and Pb-983 were selected for analysis. Each standard was diluted to a concentration of 1 pg ml-' using de-ionized water. Each standard was doped with thallium [Johnson Matthey Specpure ICPIdirect current plasma (DCP) standard] to a 1 pg ml-L concentration. Each Faraday collector was dedi- cated to a particular isotope i.e. 203Tl-lo~ 2 204Pb-lo~ 1 20sT1-axial *06Pb-high 1 *07Pb-high 2 208Pb-high 3. Six samples of each standard were analysed. Each sample was analysed for 100 s the analysis period comprising ten measurements each of 10 s duration. Sample-usage rate was approximately 0.1 ml min-' thus the analysis of each sample used approximately 200 ng of lead. Each sample analysis was preceded by a 3 min wash-out period. The total ion current recorded at the Faraday detectors was approxi- mately 6 x 1 0-l1 A for each lead standard; a similar current was recorded for the thallium solution. A neodymium reference material (obtained from Scripps Institute of Oceanography La Jolla CA USA) was also selected for analysis and diluted to a concentration of 1 pg ml-' using de-ionized water.The neodymium was also contaminated with samarium (Johnson Matthey Specpure ICP/DCP standard) to give a 0.5 pg ml-l samarium-0.5 pg ml-' neodymium mixture. Each Faraday detector was dedictated to a particular isotope i.e. 143Nd-lo~ 1 144Nd + L44Sm-axial 145Nd-high 1 146Nd-high 2 and 147Sm- high 3. Six samples of neodymium and six samples of neodymium-samarium mixture were analysed. Each sample was analysed for 200 s the analysis period consisting of 20 measurements each of 10 s duration.The sample solution was recirculated resulting in a neodymium-usage rate of approxi- mately 0.01 ml min-'. The analysis of each neodymium sample used approximately 200 ng while the analysis of each neodymium-samarium sample used approximately 1 00 ng of neodymium. Each sample analysis waspreceded by a 3 min wash-out period. The total ion current recorded at the Faraday detector for the pure neodymium solution wasabout 2 x lo-'' A. Johnson Matthey reference material 475 Hafnium was also chosen for analysis and diluted toaconcentration of 1 pgml-' using de-ionized water. The hafnium was also contaminated with lutetium (Johnson Matthey Specpure ICP/DCP stan- dard) to give a 0.5 pg ml-I lutetium-1.0 pg ml-' hafnium mixture.Again each Faraday detector was dedicated to a particular isotope i.e. L75Lu-low 2 176Lu+ 176Hf-lo~ 1 177Hf- axial and L79Hf-high 2. Six samples ofhafnium and six samples of the hafnium-lutetium mixture were analysed. Each pure hafnium sample was analysed for 500 s i.e. 100 measure- ments each of 5 s duration. Each hafnium-lutetium mixture sample was analysed for 1000 s i.e. 100 measurements each of 10 s duration. The sample solution was recirculated resulting in a hafnium usage of 0.0 1 ml min-l. Theanalysisof each pure hafnium sample used approximately 300 ng while the analysis of each hafnium-lutetium sample used approxi- mately 400 ng of hafnium. Each sample analysis was preceded with a 3 min wash-out period. The ion current recorded at the Faraday detectors for the pure hafnium solution was about 2 x lo-" A.Results and Discussion Space charge effects within the plasma region and supersonic gaseous expansion between the cones are responsible for the preferential transmission of the heavier isotopes into the mass spectrometer. The effect is time independent and its magnitude is approximately 1.2% per atomic mass unit 144p. Eqn. 1 has been shown to predict this mass discrimination' R = R,,,( 1 + c p where R, = true value R,,,,= measured value C= mass bias factor and dm=mass difference. Comparison of the measured ratio with the true ratio can thus be used to determine the magnitude of the correction factor (1 + C). Lead Isotope Ratio Measurement Isotopic ratio measurements of six representative samples of SRMs Pb-981 Pb-982 and Pb-983 corrected for massJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY FEBRUARY 1993 VOL.8 21 Table 1 Analysis of NIST SRM Pb-98 1; values in parentheses are the %SE Sample 208pb:204pb 207pb:204pb 206pb:204pb 1 36.695 (0.017) 15.484 (0.019) 16.939 (0.013) 2 36.695 (0.008) 15.484 (0.007) 16.940 (0.009) 3 36.692 (0.026) 15.487 (0.027) 16.939 (0.024) 4 36.696 (0.013) 15.486 (0.016) 16.939 (0.012) 5 36.676 (0.009) 15.477 (0.010) 16.930 (0.007) 6 36.690 (0.0 12) 15.480 (0.0 17) 16.937 (0.0 12) Mean 36.69 1 15.483 16.937 2 0 0.0 15 0.008 0.008 %RSD* 0.02 1 0.025 0.022 NIST value 36.721 15.49 1 16.937 2 0 0.036 0.01 5 0.01 1 *%RSD calculated from la. 208Pb:206Pb 2.1663 (0.005) 2.166 1 (0.003) 2.166 1 (0.004) 2.1663 (0.004) 2.1662 (0.004) 2.1663 (0.002) 2.1662 0.0002 0.005 2.1681 0.0008 207pb:206pb 0.9 14 1 1 (0.008) 0.9 1407 (0.003) 0.9 1424 (0.003) 0.91419 (0.005) 0.91413 (0.004) 0.9 1394 (0.0 1 1) 0.9141 1 0.0002 1 0.01 1 0.9 1464 0.00033 Table 2 Analysis of NIST SRM Pb-982; values in parentheses are the %SE Sample 208pb:204pb 207pb:204Pb 206pb:204pb 1 36.705 (0.015) 17.140 (0.012) 36.714 (0.014) 2 36.705 (0.021) 17.143 (0.022) 36.715 (0.020) 3 36.685 (0.013) 17.134 (0.012) 36.694 (0.010) 4 36.694 (0.01 5) 17.138 (0.01 6) 36.705 (0.01 5) 5 36.717 (0.025) 17.147 (0.022) 36.725 (0.022) 6 36.707 (0.01 1) 17.143 (0.010) 36.714 (0.010) Mean 36.702 17.141 36.71 1 2a 0.022 0.009 0.02 1 %RSD* 0.030 0.027 0.029 NIST value 36.744 17.159 36.738 20 0.050 0.025 0.037 *%RSD calculated from la.208Pb:206Pb 0.99974 (0.004) 0.99972 (0.006) 0.99972 (0.004) 0.99969 (0.003) 0.99976 (0.004) 0.99978 (0.005) 0.99974 0.00006 0.003 1 .OOO 16 0.00036 207Pb:2MPb 0.46685 (0.005) 0.46692 (0.005) 0.46696 (0.006) 0.46693 (0.003) 0.46690 (0.003) 0.46694 (0.005) 0.4 6 6 9 2 0.00008 0.008 0.46707 0.00020 Table 3 Analysis of NIST SRM Pb-983; values in parentheses are the %SE Sample 208pb:204pb 207pb:204pb 206Pb1204pb 1 36.90 (0.10) 192.8 (0.10) 2708.5 (0.10) 2 37.07 (0.09) 193.6 (0.10) 2719.7 (0.10) 3 37.23 (0.16) 194.7 (0.16) 2735.4 (0.16) 4 37.15 (0.10) 194.3 (0.09) 2729.9 (0.10) 5 36.90 (0.12) 193.0 (0.12) 27 12.0 (0.12) 6 37.22 (0.15) 194.5 (0.16) 2732.2 (0.16) Mean 37.08 193.8 2723.0 20 0.30 1.6 22.4 O/oRSD* 0.40 0.4 1 0.41 NIST value 36.71 191.9 2695 20 2.04 10.5 145 *%RSD calculated from la.208Pb:206Pb 0.013623 (0.010) 0.01 3620 (0.008) 0.01 3609 (0.007) 0.01 3608 (0.004) 0.01 3604 (0.009) 0.01 3623 (0.005) 0.01 36 15 0.0000 1 7 0.062 0.01 36 19 0.000024 207pb1206pb 0.07 1 177 (0.004) 0.07 1 165 (0.005) 0.07 1 17 1 (0.004) 0.07 1 179 (0.002) 0.07 1 178 (0.002) 0.07 1 176 (0.004) 0.07 1 174 0.00001 1 0.008 0.07 120 1 0.000040 discrimination are shown in Tables 1,2 and 3 respectively. The measurement of the 205Tl:203Tl ratio was used to correct the measured values of 208Pb:204Pb 207Pb:204Pb 206Pb:204Pb 208Pb:206Pb and 207Pb:206Pb ratios. The value of the 2osTl:203Tl ratio of 2.387 f 0.0010 (30) is given by Dunstan et a1.7 The confidence in each sample measurement (inter- nal precision) is expressed as per cent. standard error (YoSE).The confidence in the mean of the six samples is expressed both as a per cent. relative standard deviation (YoRSD) and as a 95% confidence level (20). The NIST certified value of each isotopic ratio together with its 95% confidence level is also given. The mass discrimination correction applied in this work assumes that the factor (1+C) for lead and thallium depends only upon mass and is therefore of the same magnitude. To check this assumption the mass bias values are calculated for thallium and Pb-982 using eqn. (I) i.e. 1 + CTI=[2.387 1/(20ST1:203T1)measJ0~5 (2) I +Cpb982=[1.OOO16 l(208Pb:206Pb)meas]0.5 (3) where 2.3871 and 1.00016 are certified 205Tl:203Tl and 208Pb-982:206Pb-982 ratios respectively. The correction factors calculated for each sample of the Tl:Pb-982 solution are shown in Table 4 together with their mean values.The difference between the means is within experimental error and therefore justifies the use of the correction. The measurement of isotopic ratios of SRMs Pb-98 1 Pb- 982 and Pb-983 reveal levels of internal and external precision and accuracy comparable to that obtained by TIMS. Analysis times are approximately 100 s per sample compared with at least 25 min per sample by TIMS. The levels of external precision obtained using the thallium correction technique are superior to those quoted by NIST. By reducing both sample uptake rate and analysis time sample usage has been reduced from approximately 1.6 pg in the previous study' to approximately 200 ng in this study. Analysis time has been reduced from 300 to 100 s.The levels of internal and external precision are similar in both studies proving the effectiveness of using thallium as an internal correction for the measurement of lead isotopic22 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY FEBRUARY 1993 VOL. 8 Table 4 Calculated bias correction factors for 205T1:203T1 and 208Pb-9 8 2 206Pb-9 82 Sample Bias factor 20sT1:203T1 Bias factor 208Pb-982:206Pb-982 1 1.01044 2 1.01 132 3 1 .O 1066 4 1.01 121 5 1.01 112 6 1.01 157 Mean 1.01 105 2a 0.00084 %RSD* 0.042 * %RSD calculated from la. 1.01022 1.01 198 1 .O 1 045 1.01098 1.01094 1.01 138 1.01099 0.00 1 27 0.063 ratios. Further reduction in sample usage could be achieved with the recirculation of the large volume of nebulized solution which does not reach the plasma.Isotopic ratio measurements of SRMs Pb-982 and Pb-983 agree with the certified values quoted by NIST within 20. The measurement of SRM Pb-981 agrees with NIST for all ratios except 20sPb:206Pb. The NIST certified value is 2.168 1 k 0.0008 while a value of 2.1662 k 0.0002 is mea- sured here a difference of twice the sum of the two 95% confidence errors. This discrepancy has also been noted by Platzner,s where thermal ionization analysis of SRM Pb- 98 1 revealed a 208Pb:206Pb ratio of 2.16605 k 0.00063. Values of 2.1650 k 0.001 8 and 2.1630 k 0.0020 also deter- mined by thermal ionization were reported by Hamlin et uL9 and Gulson et u1.,l0 respectively errors are given as 20 in each case. Neodymium Isotope Ratio Measurement The left hand columns of Table 5 give details of the analysis of six samples of pure La Jolla neodymium. Each sample was analysed for 20 measurements each of 10 s duration.Comparison of the measured 146Nd:145Nd ratio with a value of 2.07179 was used to determine the value of mass bias which was then used to correct the 143Nd:144Nd measured value. This correction procedure was applied after each 10 s measurement and is similar to that applied to the lead and thallium mixtures. The 143Nd:144Nd ratio is shown for each sample together with the confidence in its measurement internal precision expressed as a %SE. The mean of the six samples is given at the bottom of the column together with the %RSD and as a 95% confidence level (20). Comparison of the measured value of 0.5 1 1825 f 0.000039 with the La Jolla value of 0.5 1 1859 show the accuracy of the technique.The right hand columns of Table 5 give details of the Sm I 143 144 145 146 147 Relative atomic mass Fig. 1 Schematic diagram illustrating the isotopic components of the neodymium-samarium mixture; black shading represents the isotopes of the interfering species and grey shading the isotopic ratio of analytical interest results of analysis of six samples of a La Jolla neodymium- samarium mixture. A graphical representation of each isotopic component is given in Fig. 1. The 144Sm interferes with 144Nd and must therefore be corrected. Comparison of the measured 146Nd:14sNd ratio with a value of 2.07 179 was used to determine the value of mass bias. The 144Sm interference with 14*Nd was corrected with the measure- ment of 147Sm assuming that the 147Sm:144Sm ratio equals 4.8389.lI The measured 143Nd:144Nd ratio corrected for interference was then corrected for mass bias.Each of these corrections was performed after each 10 s measurement. The mean 143Nd:144Nd ratio determined for each sample is shown with its %SE. The mean of the six samples is given at the bottom of the column together with its %RSD and 20. Comparison of the mean value of 0.5 1 1854 +_ 0.000058 with the established value of 0.5 1 1859 shows the accuracy of the technique and the effectiveness of the proposed correction methods for mass bias effects and interfering species. Hafnium Isotope Ratio Measurement The left hand columns of Table 6 give details of the analysis of six samples of pure JMC 475 Hafnium. Each sample was analysed for 100 measurements each of 5 s duration.Cornparison of the measured 179Hf:177Hf ratio with the accepted value of 0.732512 was used to determine the value of mass bias this value was then used to correct the 176Hf:177Hf measured values. This correction was applied as the measurement proceeded. The corrected 176Hf:177Hf ratio is shown for each sample together with the confidence in its measurement %SE. The mean of six samples is given at the bottom of the column together with its %RSD and 20 Table 5 Analysis of La Jolla neodymium and a La Jolla neodymium-samarium mixture; 143Nd:144Nd ratio normalized to 146Nd:145Nd= 2.07 179 147Sm:144Sm assumed to be 4.8389 1.0 ppm La Jolla neodymium 0.5 ppm La Jolla neodymium-samarium Sample l43Nd1144Nd O/o S E 143Nd1144Nd %SE I 2 3 4 5 6 Mean 2a %RSD* Accepted 0.5 1 1830 0.5 1 1794 0.51 1855 0.5 1 1824 0.51 1820 0.51 1827 0.5 1 1825 0.000039 0.0038 0.5 1 1859 0.0023 0.5 1 1840 0.0027 0.5 1 1838 0.0020 0.5 1 1888 0.001 7 0.5 1 1845 0.00 16 0.51 1891 0.0022 0.51 1819 0.51 1854 0.000059 0.0057 0.5 1 1859 0.003 1 0.0035 0.004 1 0.003 1 0.0053 0.0036 *RSD calculated from la.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY FEBRUARY 1993 VOL.8 23 Table 6 Analysis of JMC 475 Hafnium and a JMC 475 Hafnium-lutetium mixture; 17aHf:177Hf ratio normalized to 179Hf:177Hf=0.7325 17aLu:17sLu assumed at 0.026525 Sample 1 2 3 4 5 6 Mean 20 %RSD* Accepted 1.0 ppm JMC 475 Hafnium 176Hf:177Hf YoSE 0.282 187 0.0020 0.282 188 0.00 18 0.282 189 0.00 18 0.282203 0.0035 0.282 187 0.0029 0.2822 10 0.0029 0.282 194 0.000020 0.0035 0.282 195 *RSD calculated from la.1.0 ppm JMC 475 Hafnium and 0.5 ppm lutetium 176Hf I77Hf 0.282208 0.282205 0.282202 0.2822 18 0.2822 1 1 0.282233 0.2822 13 0.000023 0.0040 0.282195 %SE 0.0026 0.0028 0.003 1 0.0029 0.0028 0.0029 value.The mean value of 0.282 194 f 0.000022 agrees with the value of 0.282 195 f 0.00001 5 determined by Patchett.Iz The right hand column of Table 6 gives details of the results of six samples of a JMC 475 Hafnium-lutetium mixture. A graphical representation of each isotopic com- ponent is given in Fig. 2. The 176Lu interferes with 176Hf. Comparison of the measured 179Hf:177Hf ratio with the accepted value of 0.7325 was used to determine the value of mass bias.The I’l6Lu interference with 176Hf was corrected with the measurement of 175Lu assuming that the 176Lu:175Lu ratio equals 0.026526.13 The measured 176Hf:177Hf ratio corrected for interference was then cor- rected for mass bias. Each of these corrections was performed as the analysis proceeded. The mean corrected 176Hf:177Hf ratio determined for each sample is shown with its YoSE. The mean of the six samples is given at the foot of the column together with its %RSD and 20. Comparison of the mean value of 0.2822 13 -t 0.000023 with the established value of 0.282195 again shows the accuracy of the tech- niques and the effectiveness of the proposed correction methods for mass bias effects and interfering species. Conclusion The effectiveness of this novel MS system for the high accuracy and high precision measurement of isotope ratios has been shown.Isotope ratios have been measured to a Lu I 175 176 177 178 179 Relative atomic mass Fig. 2 Schematic diagram illustrating the isotopic components of the hafnium-lutetium mixture; black shading represents the isotopes of the interfering species and grey shading the isotopic ratio of analytical interest precision comparable to that obtained by TIMS. Two methods of mass bias correction have been demonstrated and discussed. The first method involved contamination of a lead sample solution with thallium to allow an internal correction for mass bias a technique which would be very difficult if not impossible by TIMS. This technique also removed the need for a separate calibration analysis.The second method utilized the stable isotopic components of both neodymium and hafnium to correct internally for mass bias this method was also extended to include the correction for interfering species. This has demonstrated the effectiveness of the proposed methods for high accuracy and high precision isotope ratio measurements. The use of more efficient sample nebulization systems such as ultra- sonic or desolvation techniques will further reduce sample consumption by up to an order of magnitude. References 1 Walder A. J. and Freedman P. A. J. Anal. At. Spectrom. 1992 7 571. 2 Date A. R. and Gray A. L. Applications of Inductively Coupled Plasma Mass Spectrometry Blackie Glasgow 1988. 3 Jarvis K. E. Gray A. L. and Houk R. S. Handbook of Inductively Coupled Plasma Mass Spectrometry Blackie Glas- gow and London 1992. 4 Ketterer M. E. Peters M. J. and Tisdale P. J. J. Anal. At. Spectrom. 1991 6 439. 5 Longerich H. P. Fryer B. J. and Strong D. F. Spectrochim. Acta Part B 1987 42 39. 6 Faure G. Principles of Isotope Geology Wiley New York 1986. 7 Dunstan L. P. Gramlich J. W. Barnes I. L. and Purdy W. C. J. Rex Natl. Bur. Stand. 1980 85 1. 8 Platzner I. Int. J. Mass Spectrom. Ion Processes 1987 77 155. 9 Hamilin B. Manhes G. Albarede F. and Allegre C. J. Geochim. Cosmochim. Acta 1985 49 173. 10 Gulson B. L. Korsch M. J. Cameron M. Vaasjoki M. Mizon K. J. Porritt P. M. Carr G. R. Kamper C. Dean J. A. and Calvez J. Y. Int. J. Mass. Spectrom. Ion Processes 1984 59 125. 1 1 DeBievre P. Gallet M. Holden N. E. and Barnes I. L. J. Phys. Chem. Ref Data 1984 13 865. 12 Patchett P. J. Geochim. Cosmochim. Acta 1983 47 81. 13 DeBikvre P. Gallet M. Holden N. E. and Barnes I. L. J. Phys. Chem. Ref Data 1984 13 872. Paper 2/0228 7E Received May 5 1992 Accepted October 15 1992