geological reference materials for standardization and quality assurance of instrumental neutron...

Journal of Radioanalytical and Nuclear Chemistry, Articles, Vol. 174, No. 2 (1993)229-242

GEOLOGICAL REFERENCE MATERIALS FOR STANDARDIZATION AND QUALITY ASSURANCE

OF INSTRUMENTAL NEUTRON ACTIVATION ANALYSIS

M. D. GLASCOCK, M. P. ANDERSON

Missouri University Research Reactor, Columbia, MO 65211 (USA)

(Received March 31, 1993)

Results are presented from the INAA of 34 elements in NIST and USGS geological reference materials that were analysed relative to multielemental SRM-1633a Coal Flyash standards. The data compare favorably with works reported by other investigators. The application of historical control chat~s for continuous monitoring of quality assurance and deteaion of systematic errors is demonstrated.

The application of ins~unental neutron activation analysis (INAA) to measure trace element concentrations in whole rocks, mineral separates, and inclusions has a long history. 1-4 The concentrations of dispersed (or lrace) elements in rocks are frequently employed by geochemists to examine the mineralogical, chemical, and thermodynamic conditions that were present during rock formation, s Petrogenetic models are based on the fact that the concentrations of dispersed elements in rocks will partition between the coexisting minerals and also between the coexisting minerals and the silicate matrix. The observed differences in elemental concentrations are essential information used in examining different petrogenetic models.

In the INAA of geological samples, it is common to include reference materials for which previous analytical data already exists, such as the United States Geological Survey (USGS) materials BCR-1, GSP-1, and RGM-1 that have a lengthy history of previous analysis. 6-8 Analysis of reference materials serves many purposes including: (1) development, testing, and certification of new procedure, (2) assessment of the precision and accuracy of measurement, (3) identification of systematic errors that occur between analyses taken at different intervals over long periods of time, and (4) intercomparison of results obtained between different methods and by different laboratories. Unfortunately, the inventories for many of these older reference materials are nearly depleted. In addition, these materials often have too much or too little of many elements that would be considered ideal for a general purpose rock standard.

Earlier reports by KOROTEV 9,z~ successfully demonstrated the use of National Institute of Standards and Technology (NIST)certified reference material SRM-1633a Coal Flyash as a multielement comparator standard for INAA. The current paper extends this examination of SRM-1633a as a primary standard by demonstrating its

Elsevier Sequoia S. A., Lausanne Akad~mld Kiadd, Budapest

M, D. GLASCOCK, M. P, ANDERSON: GEOLOGICAL REFERENCE MATERIALS

application for routine determination of up to 34 elements in several USGS and NIST reck standards. The six reference materials reported in this study include the following: $RM-278 Obsidian R~k , SRM-679 Brick Clay, and $RM-688 Basalt Rock issued by the NIST and BCR-1 Basalt, GSP-1 Granediofite, and RGM-1 Rhyolite issued by the USGS.

The data reported here were compiled from thousands of analytical measurements on geological materials carried out over a five-year period at the Missouri University Research Reactor (MURR). During the course of this work, data o,a several of the reference materials were also tabulated as a means of monitoring quality assurance. In parfictflar, data on the SRM-278 Obsidian Rock reference material was critically examined for the purpose of eliminating (or minimizing) sources of experimental error. Thus, the experimental procedures described in the following section are the existing result of that evolution.

Experimen~a']

In general, each geological INAA investigation at MURR involves irradiation of unknown samples and standard reference materials in batches of approximately 50 samples. In every sample batch, four SRM-1633a Coal F1yash samples are irradiated as the prima~y standard. In addition, two SRM-278 Obsidian Rock and one or more additional SRMa are included for quality assurance proposes. The remaining - 4 0 samples in the batch are usually geological or archaeological unknowns submitted to MUtLR for routine INAA.

All of the standard reference materials analyzed are assumed to be homogeneous in thdr original containers. However, all unknown geological specLmens are carefully ground into fine powders and then mixed to obtain homogeneous samples. All geological samples and st,~dards are dried in an oven at !95 ~ for 24 hours before weighing. The dried samples are transferred to a desiccator in order to cool ts room temperature before separate aliquots of each sample are weighed for both the short and long irradiations employed at MURR.

For each short irradiation sample, about 100 mg is weighed to the nearest 0.01 mg h~to a 2/5-dram high-density polyethylene vial. The poly vial is then filled with an inert paper pulp which helps to confine the sample material in the bottom of the vial. A tightly fitting snap cap seals the vial shut. A second 100 nag sample is weighed into high-purity quartz vials made from "Suprasil" quartz (4 mm inside diameter and 6 mm outside diameter). The quartz vials are sealed on both ends using an oxygen torch flame. As much as possible, care is t~ken to produce quartz vials of the same length~ A diamond tipped marking pencil is used to engrave a unique sample identification on both poly and quartz vials.

230

M. D. GLASCOCK, M. -P. ANDERSON: GEOLOGICAL REFERENCE MATERIALS

The short irradiation samples and standards are sequentially irradiated in MURR's Row #1 irradiation position for five seconds each using a thermal neutron flux of 8.1013 n. cm -2. s -1. Following irradiation, samples are allowed to decay for 25 minutes before initiating a 12-minute measurement period. Samples are then placed in a specially designed holder that fixes them a distance of 10 cm from the front of a 25% efficient HPGe detector. The specially designed holder also facilitates inserting a magnetic stirring bar and stirring plate beneath the sample to provide continuous rotation of the sample during measurement. The short-lived elements that are routinely determined from these short irradiations include: AI, Ca, C1, Dy, K, Mn, No, Ti, and V. A typical batch of 50 samples and standards can be processed in about twelve working hours.

The quartz vials used for long irradiation are bundled into batchesof approximately 50 samples as mentioned above. The vials are arranged in a close-packed hexagonal configuration with standards and quality controls dispersed throughout the bundle. The sample bundle is tightly wrapped in aluminum foil such that its extreme diameter does not exceed 5 cm. One, two, or three sample bundles are then tightly stacked inside a 5 cm diameter aluminium can for irradiation.

All of the sample batches reported in this study were irradiated in MURR irradiation position R-2-M for a period of 24 hours using a thermal neutron flux 5- 1013 n �9 cm -2- s -l. During irradiation, the sample bundles are rotated continuously in order to reduce the effects of the reactor's radial flux gradient. Previous work, ll,lz has shown that the relative flux gradient from the exterior of a rotated sample bundle to its center should not introduce an error greater than 2%. All of the irradiations by KOROTEV 9,1~ were conducted in this same MURR irradiation position, however his sample bundles were smaller leading to a smaller radial flux gradient (on the order of 1%).

Following the long irradiation, samples decay for seven days before the vials are cleaned in aqua regia to remove any surface contamination that may have been caused by contact with the aluminium foil or various handling procedures. A first measurement ("middle count") of these long-irradiation samples is conducted 7-8 days after irradiation by measuring each radioactive sample for 2,000 seconds with a 20% HPGe detector coupled to an automatic sample changer. This measurement allows determination of eight medium-lived elements including: As, Ba, La, Lu, Nd, Sin, U, and Yb. Finally, a longer measurement of 10,000 seconds ("long count") on each sample is performed after an additional 2-3 weeks of decay to determine seventeen long-lived elements including: Ce, Co, Cr, Cs, Eu, Fe, Hf, Ni, Rb, Sb, Sc, Sr, Ta, Tb, Th, Zn and Zr.

Table 1 lists the isotopes and gamma rays used to identify and quantify each element. The element concentrations in each unknown and quality control sample are

231

M. D. GLASCOCK, M. P. ANDERSON: GEOLOGICAL REFERENCE MATERIALS

Table 1 Isotopes, gamma rays, and element concentrations in NIST SRM-1633a Coal

Fly Ash employed as a multielement reference standard in this work

Measured Measured Gamma-ray Recommended element isotope energy, keV concentration, ppm

Elements reported from short irradiations:

AI 28A1 1779.0 14.09% Ca 49Ca 3084.4 1.1% (31 38C1 1642.7 77 Dy 165Dy 94.7 14.6 K 42I( 1524.6 1.89% Mn 56Mn 846.8 190 Ha 24Na 1368.6 1650 Ti 51Ti 320.1 0.80% V 52V 1434.1 300

Elements reported from long irradiations:

As 76As 559.1 145 Ba 131Ba 496.3 1320 Ce 141Ce 145.4 168.3 Co 60Co 1332.5 44.1 Cr 51Cr 320.1 193 Cs 134Cs 795.8 10.42 En 152Eu 1408.0 3.58 Fe 59Fe 1099.3 9.38% Hf 181Hf 482.2 7.29 La 140La 1596.2 79.1 La 177Lu 208.4 1.075 Nd 147Nd 91.1 75.7 Ni 58Co 810.8 130 Rb 86Rb 1076.6 134 Sb 124Sb 1691.0 6.15 Sc 46Sc 889.3 38.6 Sm 153Sm 103.2 16.83 Sr 85Sr 514.0 835 Ta 182Ta 1221.4 1.93 Tb 160Tb 879.4 2.53 Th 233pa 312.0 24.0 U 239Np 106.1 10.3 Yb 175yb 396.3 7.50 Zn 65Zn 1115.6 220 Zr 95Zr 756.7 24O

232


calculated relative to the recommended concentrations for SRM-1633a Coal Flyash listed in Table 1. The recommended data were compiled from data reported by KOROTEV 9 and earlier work H conducted in our laboratory. Using methods we described previously, 12 corrections are made to the elements Ba, Ce, La, Nd, and Zr for interferences due to fission product isotopes created by the thermal-neutron-induced fission of natural 235U present in the sample.

Results and discussion

Up to a total of 34 elements can be measured in geological and archaeological materials when employing SRM-1633a Flyash as a primary standard. Even so, SRM-1633a is not an ideal standard because at least five elements (Ca, CI, K, Ti, and Ni) yield such low count rates that the analytical precision is generally poor. In spite of these limitations, the Flyash material is far superior to other rock powders currently known to be available. 9

The inclusion of our short irradiation procedure enables measurement of the elements AI, CI, Dy, Mn, Ti, and V that could not be determined by using the procedures employed by KOROTEV 9. However, the increased decay between the end of the long irradiation and our middle count decreases our sensitivity for the elements As, Au, Br and W which were reported by KOROTEV 9. Of the latter group, we report data for As only.

The results for SRM-278 Obsidian Rock are reported in Table 2 where the means and standard deviations are listed from a total of 218 determinations. Both Ni and V were found to be difficult to measure in SRM-278 with detection limits being reported instead. For comparison purposes, the Table 2 also lists concentration values reported on the NIST certification sheet 13 and more recent data obtained from KOROTEV. 14 For nearly all elements, the agreement between laboratories is excellent. The only element failing to satisfy a Student's t test at the 95% confidence level is Zr for which there is now growing evidence of a systematic difference between laboratories. Altttough the comparison of mean values for Cr and Sr between laboratories finds relative differences of approximately 9%, this not too surprising as both elements are in the lower range of our detectability.

It is noted that the standard deviations for our best elements are approximately 2% versus the approximately 1% reported by KOROTEV: 4 We attribute this to our working with larger numbers of samples in each batch, and thus the effect of the. larger neutron flux gradient across sample bundles we cited earlier. Even so, the overall accuracies of our measurements for the vast majority of elements are quite acceptable for routine INAA of geological and archaeological specimens.

233

M. D. G L A S C O C K , M. P. A N D E R S O N : G E O L O G I C A L R E F E R E N C E M A T E R I A L S

Table 2

Resul ts f rom the analys is o f NIST SRM-278 Obsidian Rock as compared

with N / S T cert if ied values 13 and unpubl ished data from K O R O T E V , 14

All values are in ppm of the element unless otherwise indicated

M U R R NIST K O R O T E V Elemen~

(n - 218) Certif ication (n = 15)

AI%

A s

Ba

Ca%

Cr

CI

Co

Cr

Cs

oy En

Fe%

H f

K%

La

Lu

Mn

Na%

Nd

Ni

Rb

Sb

Sc

Sm

Sr

Ta

Tb

Th

Ti%

U

V

Yb

Zn

Z r

7.45 -+ 0.21

4.3 .+ 0.5

891 .+ 39

0.71 .+ 0.18

63.2 .+ 1.7

641 .+ 86

1.47 .+ 0.03

6.4 .+ 0.5

5.21 .+ 0.10

5.76 .+ 0 .30

0 .777 .+ 0 .016

1.42 .+ 0.03

8.16 .+ 0.35

3.47 - 0.10

30.2 .+ 0.6

0 .659 .+ 0 .019

428 .+ 8

3.44 .+ 0.21

26.1 .+ 2.1

< 16

128 .+ 4

1.57 .+ 0.29

5.03 + 0.12

5195 • 0.15

61 .+ 15

1.30 .+ 0.03

0.98 .+ 0.07

11.9 .+ 0.2

0 .146 .+ 0 .027

4.68 -,- 0.38

< 12

4 .48 .+ 0.11

43 + 4

208 .+ 20

7.49 .+ 0.08 n . d .

n.d. 4.4 • 0.2

(1140) 887 .+ 16

0 .703 _+ 0 .002 0.72 .+ 0.09

(62.2) 61.5 .+ 0.8

n . d . n . d .

(1.5) 1.44 .+ 0.02

(6.1) 5.9 .+ 1.0

(5.5) 5.13 .+ 0.07

n . d . n . d .

(0.84) 0.767 .+ 0 ,008

1.43 .+ 0.02 1.410 .+ 0.011

(8,4) 8.07 .+ 0.14

3.45 .+ 0.02 n, d.

n . d . 30.0 .+ 0.2

(0.73) 0,675 .+ 0 ,015

400.+ 15 n, d,

3.59 .+ 0.04 3.54 .+ 0.03

n . d . 25.9 • 0.9

3.6 .+ 0.3 < 20

127.5 .+ 0.3 126 -+ 2

(1.5) 1.52 • 0.25

(5.1) 4.96 • 0.04

(5.7) 5.78 .+ 0.07

63.5 • 0.1 67 • 3

(1.2) 1.26 • 0.02

(1.0) 0.961 • 0.035

12.4 .+ 0.3 11.71 ~ 0.13

0.146 .+ 0 .004 n . d .

4.58 -+ 0.04 4.55:-+ 0.13

n . d . n . d .

(4.5) 4.46 .+ 0.06

(55) n.d.

n, d. 272 - 31

n. d, - not determined~

2 3 4


E 5.6 ID..

5.4

A

30 level

5 . 0 -

4.oI I _J _ I I I ~ 20 40 60 80 100

QC number

& 1 3o level

1.5~ 2c.[e_vet

I I I I % 20 60 80

QC number

Fig. 1. Historical control charts from the first one hundred analyses (approximately 50 batches) of Cs and Fe in 5RM-278 employed as a quality assurance monitor. The means, 2-~, and 3-0" deviation lines are indicated on each chart

Tn addition to measuring our analytical precision, our use of SRM-278 provides us with a continuous historical record for quality assurance and possible detection of systematic errors between sample runs. A common method of reviewing historical data is the use of graphical conlrol charts as shown in Fig. 1. We routinely track our most sensitive elements (i.e., Ce, Co, Cs, Eu, Fe, Hf, Sc, and Th) by employing this method. Separate decision levels indicating the need for "evaluation" or "corrective action" are indicated by the 2-0"and 3-0- deviation limits, respectively. When the corrective action Hmits for a number of the monitored elements are exceeded, the data for that batch are critically examined to identify and !correct probable systematic errors.

Tables 3 and 4 list our results for SRM-679 Brick Clay and SRM-688 Basalt, respectively, where fewer analyses were performed. In both cases, the NIST certified values 13,15 and unpublished data from KOROTEV 14 have been presented for comparison. Agreement between MURR and KOROTEV 14 is excellent except for the

235


Table 3

Results from the aitalysis o f NIST SRM-679 Brick Clay as compared

with NIST certified values 15 and unpublished data from KOROTEV. 14 All values

are reported in ppm of the element unless otherwise indicated

M U R R NIST KOROTEV Element

(n - 10) Certification (It - 6)

AI% 10.6 • 0.3 11.01 • 0.34 It. d.

As 9.5 __. 0.6 It. d. 9.5 • 0.2

Ba 445 • 48 432.2 • 9.8 439 • 9

Ca ,~ 2000 1628 • 13 1500 • 860

Ce 105.5 • 2.5 (105) 103.3 • 1.2

CI < 1 0 0 n . d . n.d.

Co 25.8 • 0.4 (26) 25.7 • 0.3

Cr 107 • 2 109.7 • 4.9 107 • 1

Cs 9.86 • 0.20 (9.6) 9.66 • 0.05

Dy 6.24 • 0.31 It. d. n .d .

Eft 1.81 • 0.06 (1.9) 1.78 • 0.03

Fe% 9.06 • 0.18 9.05 • 0.21 9.05 • 0,10

l-If 4.59 • 0,14 (4.6) 4,43 • 0.05

K% 2.38 • 0.14 2.433 • 0.047 n, d.

La 49.9 • 0.4 n .d . 49.9 • 0.5

Lu 0.527 • 0.017 It. d. 0,538 • 0.004

Mn 1835 • 35 (1730) a . d .

Na% 0.123 • 0.003 0.1304 • 0.0038 0.128 • 0.004

Nd 47.8 • 5.1 n .d . 43.3 • 2.8

Ni < 7 5 n .d . 5 6 •

Rb 196 + 9 (190) 189 • 3

Sb 0.80 • 0.10 n . d . 0.72 • 0.02

Se 22.7 • 0.4 (22.5) 22.8 • 0.2

Sm 9.16 • 0.10 It. d. 9.13 • 0.08

Sr < 80 73.4 • 2,6 82 • 7

Ta 1.26 • 0.22 n . d . 1.21 • 0.01

Tb 1.23 • 0.19 n . d . 1.20 • 0.02

Th 13.7 • 0.2 (14) 13.46 • 0.12

Ti% 0.52 • 0.05 0.577 • 0.033 n . d .

U 2.55 • 0.55 II. d. 2.71 • 0.13

V 1 6 0 • n. d, i1. d.

Yb 3.82 • 0.13 n. d, 3.68 • 0.04

Zn 104 • 21 (150) n . d .

Zr 138 • 23 n .d . 145 • 14

n. d. - not determined

236


Table 4

Results from the analysis of SRM-688 Basalt Rock as compared with NIST

certif ied values 13 and unpublished data from KOROTEV. 14

Al l values are reported in ppm of the element unless otherwise indicated

MURR NIST KOROTEV Element

(11 = 10) Certification (n - 12)

Al% 9,08 __ 0.34 9.18 _ 0.05 n . d .

As <2 .5 n .d . 2.4-+0.3

Ba 145 -+ 24 (200) 154 -+ 49

Ca% 8.81 • 0,21 (8.70) 8.79 -+ 0.36

Ce 11.7 -+ 0.5 (13.3) 11.95 -+ 0.15

CI < 50 n . d . n .d .

Co 48.3 -+ 0.5 (49.7) 48.5 + 0.4

Cr 326 -+ 4 332 • 9 329 -+ 3

Cs <0 .2 n .d . <0.1

Dy 2.79 -+ 0.18 n .d . n .d .

Eu 0.990 -+ 0.020 (1.07) 0.983 • 0.015

Fe% 7.08 -+ 0.08 7.23 • 0.03 7.21 -+ 0.08

I-If 1.59 _ 0.06 (1.6) 1.55 -+ 0.03

K% < 0.5 0.155 -+ 0.007 n.d.

La 5.14-+0.17 n . d . 5.17-+0.05

Lu 0.292 -+ 0.038 (0.34) 0.304 -+ 0.006

Mn 1334 -+ 50 1290 • 20 n . d .

Na% 1.51 -+ 0.04 1.60 • 0.02 1.59 -+ 0.03

Nd 6.8 -+ 1.2 n . d . 8.4 -+ 1.0

Ni 114 -+ 26 (150) 158 -+ 13

Rb < 6 1.91 :e 0.01 3.8 -+ 1.2

Sb < 0.15 n . d . 0.097 + 0.016

Sc 37.0 -+ 0.4 (38.1) 37.0 -+ 0.3

Sm 2.37 -+ 0.04 (2.79) 2.40 -+ 0.03

Sr 115 -+ 26 169.2 • 0.7 172 -+ 17

Ta 0.266 -+ 0.030 n .d . 0.269 -+ 0.011

Tb 0.48 -+ 0.09 (0.448) 0.499 -+ 0.011

Th 0.27 -+ 0.04 0.33 • 0.02 0.282 -+ 0.019

Ti% 0.61 -+ 0.05 0.700 • 0.006 n .d .

U < 0.6 (0.37) 0.32 - 0.18

V 261 -+ 13 (250) n .d .

Yb 2.11 -+ 0.07 (2.09) 2,03 -+ 0.04

Zn 95 -+ 7 (58) n .d .

Zr < 80 n .d . 59 -+ 10

n. d. - not determined.

237


Table 5

Results f rom the ana lys i s of U S G S BCR-1 Basalt as compared with the data

compi la t ion o f G L A D N E Y 8 and unpubl ished data from K O R O T E V . 14

All va lues are reported in ppm of the element unless otherwise indicated

M U R R G L A D N E Y ' s K O R O T E V Element

(n - 5) Compilat ion (n - 4)

.4,1% 7.43 • 0.28 7.21 • 0.13 n . d .

A s < 2 0.64 • 0.14 n . d .

Ba 658 • 24 678 • 16 670 • 40

C a % 5.12 • 0.18 4.97 • 0.11 4.9 • 0.2

Ce 52.3 • 0.6 53.7 • 0.8 53.9 • 0.8

CI < 80 60 • 16 n . d .

Co 37.2 • 0.3 36.3 • 1.6 37.4 • 1.7

Cr 11.1 • 0.2 16 • 4 10.9 • 0,6

Cs 0 .92 • 0.04 0.97 • 0.13 0.90 • 0.15

Dy 6.35 • 0.25 6.35 • 0.12 n . d .

Eu 1.95 • 0 .03 1.96 • 0.05 1.96 • 0.04

Fe% 9.33 _-. 0 .09 9.38 • 0 .22 9.50 • 0.18

H f 5.09 _ 0.04 4.9 • 0.3 5.2 • 0.3

K % 1.54 + 0.07 1.40 • 0007 1.2 • 0.9

La 25.1 • 0.4 25.0 • 0.8 25.4 • 0.6

Lu 0 .443 • 0 .008 0 .512 • 0.025 0 .482 • 0.009

Mn 1609 • 16 1410 • 90 n . d .

Na% 2.34 • 0.02 2.43 • 0.08 2.40 • 0.05

Nd 31.3 • 3.8 28,7 • 0.6 29.3 + 2.4

Ni < 5 0 1 3 • n . d .

Rb 52 • 1 47.1 • 0.6 52 • 5

Sb 0.64 • 0.05 0.62 • 0.10 0.60 • 0.06

Se 32.1 • 0.2 32.8 • 1.7 32.4 • 1.6

S m 6.48 • 0.09 6.58 • 0.17 6.76 • 0.I1

Sr 328 • 63 330 • 5 390 • 70

Ta 0.75 • 0.04 0.79 --- 0.09 0.78 _ 0.10

Tb 1.02 • 0 .19 1.05 • 0.09 1.02 • 0.07

Th 5.73 • 0 .09 6.04 • 0.60 5.95 • 0.21

Ti% 1.28 • 0.05 1.33 • 0.06 n . d .

U 1.64 • 0 .27 1.71 • 0.16 1.7 • 0.3

V 442 • 13 404 • 40 n. el.

Yb 3 .39 • 0.09 3.39 • 0.08 3 .3I • 0 . I0

Zn 141 • 6 129 • 1 n . d .

Z r 134 • 21 191 • 5 170 • 60

n. d. - n o t d e t e r m i n e d .

238


Table 6

Resul ts f rom the ana lys i s of U S G S G S P - I Granodior i te as compared with the

data compi la t ion by G L A D N E Y 8 and unpubl ished data f rom KOROTEV. 14 Al l values are reported in ppm of the d e m e n t unless otherwise indicated


(n - 5) Compila t ion (n = 5)

AI% 7.90 • 0.33 8.02 • 0.15 n . d .

A s < 1 . 8 n . d . n . d .

Ba 1223 • 60 1310 • 10 1290 • 10

Ca% 1.43 • 0 .12 1.46 • 0.07 n . d .

Ce 420 • 8 406 • 20 437 -,- 9

CI 420 • 47 330 • 20 n . d .

Co 6.57 • 0.15 6.5 • 0.8 6.73 • 0.13

Cr 14.0 • 0.9 13.0 • 2.6 10.9 • 0.2

Cs 0.96 • 0.03 0.95 • 0.16 1.01 • 0.01

D y 5.94 • 0.35 5.4 • 0.4 n . d .

Eu 2.17 • 0.02 2.36 • 0.22 2.21 • 0.03

Fe% 2.91 • 0.03 3.01 • 0.09 2.96 • 0.02

H f 16,4 • 0.4 15.0 • 1.3 16.6 • 1.1

K % 4.59 • 0.20 4.57 • 0.12 4.67 • 0.51

La 182 • 1 8 2 • 13 183 • 3

Lu 0 .207 • 0 .017 0 .220 • 0 .050 0 .235 • 0 .010

Mn 333 • 12 310 • 40 n . d .

Na% 1.96 • 0 .02 2.08 • 0.07 2.12 • 0.02

Nd 210 • 7 190 • 17 207 • 6

Ni < 15 9.8 + 3.2 n . d .

Rb 249 • 7 254 • 2 242 • 6

Sb 2 .96 • 0,11 3.2 • 0.4 3.16 • 0.04

Sc 5.69 • 0.11 6.1 • 0.5 5.92 • 0.05

S m 25.4 • 0.6 26.8 • 2.5 26.5 • 0.4

Sr 247 • 19 234 • 3 251 • 12

Ta 0 .837 • 0 .027 0.91 • 0.14 0.86 • 0.02

Tb 1.24 • 0.03 1.36 • 0.14 1.35 • 0.05

Th 101 • 2 105 • 5 104 • 2

Ti% 0.38 • 0.04 0 .393 • 0 .024 n . d .

U 1.92 • 0.22 2.2 • 0.3 2.30 • 0.06

V 6 0 • 5 3 • n . d .

Yb 1.70 • 0.08 1.7 • 0.4 1.70 5:0.04

Zn 92 • 2 103 + 9 n . d .

Z r 378 • 14 530 • 70 539 • 40

u . d . - n o t d e t e r m . i n e d .

239

M. D. G L A S C O C K , M. P. A N D E R S O N : G E O L O G I C A L REFERENCE MATERIALS

Table 7

Results f rom the analys is o f USGS R G M - 1 Rhyoli te as compared with the data

compi la t ion o f G L A D N E Y 8 and unpubl ished data f rom K O R O T E V . 14

All va lues are repor ted in ppm of the element unless otherwise indicated


(n = 5) Compila t ion (n = 7)

Al % 7.40 -,- 0 .24 7.24 -,- 0.16 n . d .

As 3.1 -,- 0.3 < 5 3.24 • 0.15

Ba 826 • 31 800 • 60 813 + 8

Ca % 0.90 • 0.13 0.81 • 0.07 0.86 • 0.08

Ce 46.0 • 1.2 53 • 7 46.2 • 0.7

CI 783 • 36 480 • 40 n . d .

Co 1.92 • 0~ 2.0 • 0.2 1.95 • 0.03

Cr 2.78 • 0.15 2.6 • 0.4 2.93 • 0.25

Cs 9.78 • 0.18 10.0 • 0.4 9.93 • 0.09

Dy 3.34 • 0.22 (4.3) n . d .

Eu 0 .592 • 0 .016 0.75 • 0.14 0 .597 • 0.008

Fe % 1.24 • 0.03 1.32 • 0.05 1.279 • 0 ,022

H f 6.13 • 0.11 6.0 • 0.5 6.17 • 0.11

K % 3.69 • 0.12 3.64 • 0 . , o 3.57 • 0.33

La 23.3 --- 0.4 (25) 23.1 • 0.3

Lu 0 .382 • 0 .030 (0.42) 0.384 • 0 ,007

Mu 323 • 7 280 • 20 n . d .

Na % 2.92 • 0.08 3.00 • 3.04 • 0,02

Nd 21 • 2 17 • 2 18.2 • 0.3

Ni < 16 < 14 < 20

Rb 145 • 3 157 • 4 149 • 1

S b 0.97 • 0.05 1.3 • 0.2 1.204 • 0 .033

�9 Se 4.31 • 0.09 5.0 • 0.9 4.41 • 0.05

S m 4.34 • 0.08 4.3 • 0.5 4.05 • 0.06

Sr 120 • 10 110 • t 0 106 • 8

Ta 0 .887 • 0 .022 0.95 • 0.10 0 .892 • 0.019

Tb 0 .578 • 0 .020 (0.74) 0.592 • 0.008

Th 13.76 • 0.28 i 6 • 2 14.05 • 0.15

T i % 0.18 • 0.02 0.16 • 0.02 n . d .

U 5.97 • 0.25 (5.84) 5.72 • 0.16

V < 1 0 1 3 • n . d .

Yb 2.54 • 0.17 2.6 • 0.4 2.49 • 0.03

Zn 3 2 • 3 2 • n . d .

Z r 1 5 0 • 7 214 • 14 220 • 13

n. d . - n o t d e t e r m i n e d .

240


elements Nd, Ni, Rb, Sb, U, and Zr. These elements are all at significantly lower

concentrations in these reference standards as compared with SRM-278 and other rock standards. KOROTEV 9 also notes that Sb concenlrations were quite variable suggesting

greater heterogeneity for this element.

I ~] I03F ~] 103

~v ISRM-1633a ~v

SRM-278 ~:t'L,*.~

I0 ~ ~ 10

._LI t I ~ I I t I I t i ~ I I )- I I I i I ) I I I I I I I I I I ),

Lu La Nd Eu T#y Yb Lace Nd smEu T#y Yb .Ce Sm Lu

Fig. 2. REE concentrations in NIST and USGS reference materials normalized relative to chondritic abundances. The chondrite values of NAKAMURA 16 were employed

Tables 5, 6, and 7 present our results for the BCR-I Basalt, GSP-1 Granodiorite, and RGM-1 Rhyolite reference materials as determined in routine analyses. In each case, the MURR results are presented as the means and standard deviations we determined from

the analysis Of five replicates. Concenla'ation values from the compilation by GLADNEY 7,8 and unpublished results from KOROTEV 14 are included for comparison.

In general, the results for compare favorably for most elements. The most notable exceptions are Nd, Sb, and Zr all of which were also difficult elements in the N/ST

standards at these levels. As explained by KOROTEV, l~ an important test of the quality of rare earth element

(REE) concentrations values is the smoothness of plots of chondrite-normalized REE abundances versus atomic number. Figure 2 presents our data for NIST and USGS reference materials reported in this study after normalization by the chondrite values of NAKAMURA. 16 The patterns show the evenness expected.

241


Conclusions

During the course of five years of laboratory work with INAA, we have analyzed several thousand geological and archaeological specimens relative to SRM-1633a which we use as our multielement standard. Our results show that SRM-278 Obsidian Rock and several other NIST and USGS reference materials were analyzed with excellent reliability. For most elements (except Ca, CI, K, Ni, and Ti) the accuracy and precision are excellent. Other standards are recommended if Ca, CI, K, Ni, and Ti are being sought. The use of a single reference material such as SRM-278 for quality assurance, detection, and correction of systematic errors between runs was also demonstrated.

The work reported here was supported by National Science Foundation grants (BNS 88-01707 and BNS 91-02016) to M.D.G. and a Research Experiences for Undergraduates grant from the National Science Foundation to M.P.A. during the summer of 1989. The authors achnowledge the assistance of the Reactor Operating staff and engineers who perform an exemplary job of keeping MURR on its 150+ hours per week operating schedule. The authors also thank R. L. KOROTEV for supplying his unpublished data on many of the reference materials reported in this paper.

References

1. G. HEVESY, H. LEVI, Math. Fys, Meal., 145 (1936) 34. 2. H. BROWN, E. GOLDBERG, Science, 109 (1949) 347. 3. G. W. REED, A. TURKEVICH, Nature, 176 (1955) 794. 4. J. C. LAUL, Atomic Energy Rev., 17 (1979) 603. 5. G. N. HANSON, Ann. Rex,. Earth Planet, Sci., 8 (1980) 371. 6. F. J, FLANAGAN, Geochim.'Cosmochim. Acta, 33 (1969) 81. 7. E. S. GLADNEY, W. E. GOODE, Geostandards Newsl., 5 (1981) 31. 8. E. S. GLADNEY, C. E. BURNS, I. ROELANDTS, Geostandards Newsl., 7 (1983) 3. 9. R. L. KOROTEV, J. Radioanal. Nncl. Chem., 110 (1987) 159.

lO. R. L. KOROTEV, L Radioanal. Nncl. Chem., 110 (1987) 179. 11. C. C. GRAHAM, M. D. GLASCOCK, J. J. CARNI, J. R. VOGT, T. G. SPALDING, Anal. Chem. 54 (1982)

1623. 12. M. D. GLASCOCK, P. I. NABELEK, D. D. WEINRICH, and R. M. COVENEY, Jr., J. Radioanal. Nucl.

Chem., 99 (1986) 121. 13. G. A. URIANO, National Bureau of Standards, Certificate of Analysis, Washington, D C., 1981. 14. R. L. KOROTEV, unpublished data. 15. S. D. RASBERRY, National Bureau of Standards, Certificate of Analysis, Washington, D~ C., 1987. 16. N. NAKAMURA, Geochim. Cosmochim. Acta, 38 (1974) 757.

242

geological reference materials for standardization and quality assurance of instrumental neutron...

Documents