ROBOTIZED ICP-LABORATORY FOR THE ANALYSIS OF PRECIOUS METALS

M. Bril! W.C. Heraeus GmbH Analytical Laboratories Hanau, Germany

A highly automated system, based on inductively coupled plasma (ICP) optical emission spectrometry of aqueous solutions, has been developed for the analysis of precious metals. The system, designated ICP-SIM-ROB, is designed around an ICP simultaneous spectrometer, supported by auxiliary equipment ranging from robot units for the preparation and handling of the sample and reference solutions to a facility for producing machine-readable labels. Under the control of its main computer, the system can be set up for unattended operation and it is even possible for analysis to be initiated on a 'self-service' basis by users not familiar with the detailed operation of the equipment. The modular design enables hardware and software components of the system to be added to or updated in line with evolving requirements and progress in technology. The system is capable of determining the fineness of gold jewellery alloys with the same precision as the manual ISO standard method.

38

Gold Bull., 1995, 28(2)

INTRODUCTION
The development of automated methods of analysis has been eagerly awaited, especially in the precious metals recycling sector since, obviously, the cost of analysis is one factor that can adversely affect the profitability of recycling. So far, however, there has been some hesitancy in exploiting the rationalization potential now possible in the laboratories specializing in noble metal chemistry. Including the necessary process and product analysis for value and quality assurance, the main methods of analysis are essentially: X-ray fluorescente, emission spectrometry for solid samples and ICP spectrometry for aqueous solutions containing precious metals [1-6]. Apart from a small, but significant, proportion of gravimetric methods used for products with a high precious metal content, the wet chemical and fire assay techniques provide a valuable auxiliary service, mainly for ICP spectrometry. In addition to its localized tasks of process control, the analytical laboratory, by describing modular analytical methods with details of know-how and equipment, makes an important economie contribution to contractual matters concerning the trade in goods containing precious metals, such as ore concentraten, certified pure precious metals and platinum-rhodiuin catalyst gauzes. The value of the precious metal content of these goods has, in the international exchange of merchandise, the same importance as the known stable currencies. Accordingly, it follows that at the international level the basis of confidence depends on recognition of the accuracy of the methods for value assessment. This applies in particular to modern methods like ICP spectrometry, whose use on an international basis has been promoted by the standardization process and which has met with a high degree of approval. The important point here is that the international standards Working Group ISO/TC174/WG1 has provided a clear formulation of methods for the determination of platinum, palladium, gold and silver, and has confirmed their quality by round robin tests [7-10]. The applicability of these methods for jewellery alloys is a major factor contributing to the aim of replacing the labour intensive gravimetric methods [11-15]. ICP spectrometry of mineral acid aqueous solutions containing precious metals and base metals has shown itself to be appreciably more selective for the noble metal elements and has consequently opened up a wide field of application in the area of precious metals analysis. On the basis of this and the certitude of the future importance of the method, a large capital investment in ICP was

undertaken at Heraeus, covering the current equipment possibilities including robotization. This was also directed at the objective of setting up a central ICP laboratory facility with a high analytical capacity that would be a first step towards the automated measurement of solutions when the laboratory is unoccupied. The following sections present the ICP information systern and its operation as a fully robotized ICP laboratory, restricting the discussion to the determination of gold in solutions of gold jewellery alloys. This research and development project was promoted by the World Gold Council, within its technological development programme for the gold jewellery industry, with the task of developing the software for this technique in order to determine gold with the same excellence as is possible with cupellation. [*]

1CP METHODS CAPABLE OF AUTOMATION FOR THE DETERMINATION OF PRECIOUS METALS

a) New Contributions to ICP Methods for the Highly Accurate Determination of Precious Metals in Aqueous Solution Continuing from the report of the 1992 congress on the highly precise measurement of precious metal fineness in jewellery alloys by ICP optical emission spectrometry, the correlations described therein can today be confirmed on the basis of practical experience gained fiom the large number of analyses performed [16, 17]. The consequential extension of the methods to the majority of precious metal content determinations of recycled materials is further proof of the universal suitability of the methods provided that representative aqueous solutions can be produced by means of chemical and fire assay preparation techniques. Figure 1 briefly summarizes the principle of the ICP method as applied to gold. Gold fineness values for gold jewellery alloys measured using the manual method of ISO TC174/N71 are given in Table 1 where they are compared with the results obtained by the cupellation method of ISO 11426 [18-25]. In the international standards work no complete agreement as regards the relatively high concentrations of buffer components (CuCl z , NaNO 3 , HCI) in the ICP measuring solution
[*] The present report has been presented at the 2nd European Precious Metals Conference in Lisbon on May 12th, 1995.

ro

Gold Bull., 1995, 28(2)

39

Figure 1

The principle of the Determination of Gold According to the ISO/TCI174/WG1 /N71 Method

Calculations
Nota
The method of interaal standardization is based on the linear relation between the intensity ratios 'Au /1y and the concentration ratios Cap/CY or botter mass ratios m,, /my. Using the same mass of Yttrium (buffer/interml standard solution) to prepare all solutions, it is not necessary to have an exact volume of the measuring solutions. The accuracy of the 100 ml volumetrlc Ilask is satisfactory. The other important advantage of refering always to the same nominal mass of the interml standard is that all calculations can be done with ma,, instead of mn /my om;,,^.

'the s riultaneousiy :registered:.single rneasurements of the gold

Generslly the . data processing unit.proades the .quotients. from

and tie,yttrium- intensitis.

has yet been achieved because of the use of very different spectrometer systems. It should however be stressed that this buffer mixture is intended to stabilize the plasma output of those spectrometer types without necessitating electronic adjustment. It is known that the stabilizing property of similar buffer mixtures is used with DCP methods [26]. The endeavour to simplify where possible the specified ISO standard methods in this respect is a clear indication of the general interest. Moreover, the results of a round robin analysis of gold jewellery alloys carried out internationally using suitable simultaneous spectrometers are stimulating and are shown in Table 2.
of

A contribution worthy of mention with regard to the chemistry of the solvent medium for gold jewellery samples is that of an assay office which on the basis of weighings under 20 mg can determine their carat values with sufficient accuracy. The modified ISO N71 method makes use of amidosulphonic acid to remove the derivative of nitrous acid and, finally, sample and calibration solutions containing sulphuric acid [29]. It should here be remarked that in the original method the hydrochloric acid medium is designed to stabilize the nitrous acid by formation of the NO+-nitrosyl cation and this is certainly more practical for the automation of the method. Also worthy of mention for the standards work is that it was clearly shown that for the determination of silver in silver jewellery alloys according to ISO 9202 [30] using a HNO 3 /HCI solution mixture with a high hydrochloric acid fraction and addition of Cue - to compensate for the possible presente of copper - the ICP method is capable of being carried out with the same accuracy as potentiometry (ISO 11427) [10]. These contributions serve to show the universal cmployment of ICP spectrometry for precious metal analysis and also to increase our onderstanding of the correlations of the method.
b) The ICP Spectrometer Criteria for Automation

The parameters indicated in the ISO ICP standards for fineness determinations can be used as a basis for the procurement of suitable multichannel spectrometers. In
CtL Gem A 332.96 1 U.IS (5)

lo'

CupeUon lW i U d
^

- ^0fo 4 ) (n)
0

ICP 4na1.

AuV.)(n)

B ICPAnaf. Ai[%6j()(n) 333.30 t 0.10 (5) 376,19 O. SS' (5). 585.85 0.75 (6) 585,40 0.30`(5) 58801 040(6) 750.53 0.29(6) 750.73 0.34 (6) 75334 .080',(6) , 751.91 0.20(5) 747.32 0.21 (5) 585.16 *:0.55:(6) sgo 2i X 0 l.4() 747.I9 *0.41(5). 75441 #'0:3.1 (5) 917.23 t 0.80 (6) 989,90 0.66 ffl
" ,

GG CG CG
CG.

CG CG CG
CG

CG CG
WG

333.05 t 0.1.6 (8) 375.86 0.13 :(8) 585.32#0.13(7) 584.88. 1.0.21 (8) 587.74 t 0,1 2 (8)! 750.98 0.07 (7) 750.08 0.16 (7) 753.27'10.12;(8) 751.900.14(8) 747.00 t 0.08 (8)
, ,

3751O.IO(51
585.27 0.10 (5) 585.00 * 0.$ (5). 388..19 `t 0.23 750.70 0.16 (5) 749.68 t0.32 (5) 753.43 t* 0.11(5)

751.90 0.15 (3)

747.44 0.07 (5) 585.00'} OUO (2), 59060035(2)

vv

58488 t 0.14 (8)

590,30 0.14 (7)

332.00.5(5) 375:810.4(4) 585,3 0.5 (6) 1584.9 0.3(5) 587,1. 0.6 (6) 752.20.8(6) 750.3 t 0.5 (6) 753.808(6) 752.0.0.6 (6) 748.1 * 1.3 (6) 583.8*07(6)

Table 1 Comparison of Fineness of Gold in Goldjewellery Alloys by Cupellation (ISO 11426) and ICP Spectrometry
(ISO/N71) [2.3-25, 27, 28]

WG.
CG
CG
;

750.990.13'(8) 75983;0.12 (5) 917.180.11 (7)

5050 t o. (2)

488.57*0.14(8)

917.300.57(2) 908 0000.(2) s (48) =0.215

7'500.3: (r) 13 0,4 () 917.4# 2.1 (6) 92..1.5 () s (9'1) = 0.896

0 0.9 5)

Standard devlation of all single determinations,

5 (123) =0, 1.34

s (90) = 0.518

Au (%)

40

ro Gold Bu11., 1995, 28(2)

Test sample

Cupelladon

Swflzed~
%oaw %

ICP

ICP,

Gen

y (}

%aa^, 2409 2410 2414 2412

24^1:3

% arm
751.,84 337.5,1 583:41 753.37 332:83
750.48

PIW

Ni WG C.G NiWG CG.

2414

9c

9 ct
22Ct

NJ WG
CC.

ce

752:83 335.85 586.1I 753.49 333,07 75:28 76 ii 377,80 917 ti5

0.29 (6) 0..14 (g) 0.13 (8) 0.16(8) 0.13 (8)

o.Q6 (g)

75z.90
335.60 752.40 332.50 7.51 8fi 377.60 9f6.60

a 28. (6).
0.1.5 (6) 0.42 (6) 0.15 (6)

OSr0), 0.69 (6)

0.^2 (6g.
0.69 (6) 029 (6) 4 92 . (y

0.22 q8)
0.04 (8) 0.21 (8)

7s9v

PAAI (6) 041 f6)

0,26 (6) 0.29 (6)

37725

278.72

0:42 (6).'
0.48 (6) 0.73 (6)

75$30; 0.36,(6). 5 335.94 0.28 (6) 586.38 0(6) 753.05 0.21 (6) 332:8O 0:28 (6). 7^5046 10.33 (4) ,

Table 2 International Round Robin Analysis for the Determination ofFineness of Gold in Goldjewellery Alloys - Comparison of Cupellation ISO 11426 and ICP ISO/N71 [3134]

916.82

Test ~e

Germrry W

Cdladon
,

ICP

Gerrrar y H

ICP .

%.A. VZ
2409 2410 1411 2412 2413 241.4 9 ct 9" 22:tt PdWG: NIWG ; Nl WG CC CC CC NiWG CC 735.40 336.00 585_80 75360 33270 750.40 376.80 376.70 91730 0.17 (5) 0.10 (5) 0.12 (5) , 0.05..(5) 002(5),, 0.14 (5) 0.06(5): 0.26 (5) 0,21 (5):

%,Am
775284 334.40 5871.0 752.90 333:10 751.70 _ -

%r
1.5 (4) 1.1 (6) 03 (6) I, I . (6) 11(6) 1.0 (6) -

%u alt
752.59 336.15 586 10 752.20 33170 750.90 37681 377.07 91723

2,
^()) 0:8 0.4(5) 03 (5) 0.5 (5) 05(5) 04r(5), 01(5) 02(5) 08(5)

^2, %.A. %
754i20

0;14-(5),
0.25 (5)

335.70 5861}5 754.00 33310 7.1..27 375,.80. 376.20 9,1.6:30

0.43, (5)' 0.60 (5) Oi 1 I (5):. 0.I3:(5). 0.10 (5) 0.18 (5) 1,07s),

the precise planning of a highly automated ICP laboratory, the layout of the spectrometer is very important. For an instrument to be used as an ICP information source, as may be required by the complexity of the analytical problems of precious metal recycling, a dual ICP spectrometer with vacuum facility, 114 element line channels and 2 scanner arrangements (quartz refractor plates) was envisaged. In the approved Paschen-Runge arrangement, the optical layout is based on concave gratings with 1510 Tines/mm, focal lengths of 0.75 m, secondary slits provided with order filters and special photomultipliers [34]. By scanning element line backgrounds, measuring and interfering line effects can be monitored in a cycle of minutes. This is indispensable, given the complexity of recycling analysis. The approved fire assay and chemical separation methods using NIS, Cu, Pb, Sn, H 2 S, Zn, Te are not selective enough with respect to the possible performance of the ICP instrument described. For this reason a subtractive interference correction must be con- sidered which of course requires the determination on their detection lines of interfering elements such as Ni, Cr, Fe, Zr, Mo and V. The Rowland circles of both spectrometers are interspersed with several Tine positions for Y, Sc, Be, Sr so that a great ana-

lytical flexibility is possible based on the theory of the internal standard. In contrast to the ISO methods for fineness determinations describing the total signal technique for precious metal and standard lines, the net signal technique must be used for solutions with several elements present in similar concentrations. Fortunately, with our ICP system a precision in the region of parts per thousand can be attained which even comes up to that of the ISO standard methods. This performance is only achievable if the physical stability of the equipment can be guaranteed as well as the intrinsic excellence of the analytical method. As regards the sample solution, multiplicative intensity effects must be controlled by the addition of buffer substances such as topper. These intensity effects act on short and long wavelength element lines in the presente of elements like Cu, Zn, Al and Fe in a very marked and variable manner. Mineral acid content should also be considered in this respect and it alone can have an effect on the trace analysis of precious metals for their purity certification. Within the framework of development and the care and control of this central ICP information system, there should also be included the performance testing of other spectrometer techniques. It should be noted

(9 Gold Bull., 1995, 28(2)

41

here that, above all, the optical resolution of 20 pm specified in the ISO standards for the given stability of the ICP excitation of the system represents the basis of the system performance. It is understandable that for the described conditions compromises were necessary with the dispersion under consideration and the necessity for the disposition of measuring lines in higher orders and consequently a special filter technique was necessary. With our current knowledge and the required performance of precieus metal analysis it would be mere speculation to say that instruments with an optical resolution in the region of 8 pm in the first order and the flexibility of sequential spectrometers or the latest echelle' grating optics, perhaps with semiconductor detection, would have been able to create the proposed arrangement more economically. A more important consideration is whether, with the modelar construction of the entire ICP information system, components of the equipment and the methods can possibly be replaced by compatible components of greater performance, and this is, in fact, the case. With the desired target of the highest possible degree of automation of the entire installation, the most important requirement is for consistency of the element concentration dependent ICP emission intensities of element lines for sample and calibration solutions. The stability of the high frequency output plays an important part here. Obviously the electronic control system employed achieves optimum resonance conditions with the passing of time so that the stability of the aerosol excitation in the argon plasma torch is not impaired by small alterations to the composition of the sample solutions. In this way, the ICP system can also satisfy, apart from the necessary condition of precision, the condition sufficient for the gravimetric excellence of accuracy in the parts per thousand region. This quality criterion is demonstrated by means of precieus metal concentrations between 50 and 200 mg/l with the emission lines shown in Table 3. The reliability of the nebulizer is crucial for uninterrupted operation of the ICP spectrometer when the laboratory is unoccupied. Although the ISO precision method had been developed originally with the Meinhard nebulizer, the cross flow nebulizer, less sensitive for unavoidable solid particles in the sample solution, has proved successful. Attention should be paid to not placing high and low element concentrations next to each other unnecessarily since, otherwise, memory effects occur when they are nebulized. It is also important not to carry out calibrations in sequences which result in solutions influencing one another. With the given stability of ICP excitation, this problem can be obviated by, for example, lengthening the pre-integration time [35, 361.

Table 3

The Precious Metal Line Channels Installed in the Dual ICP Multichannel Spectrometer and their Optical Parameters (Grating 1510 lines/mm)
Element Emission Line

(n-)

Order

Resolution

(nni)

Spectr.

:Au
Ag Pa

Au

!lu

Au
Pd

Au Ag Ag
,

242.7 2675 267.5 328.0 338.2

1 41.8
-

2
2. 2 2 1 1

0.0:24
0.016' 0,024 0.048:' 0.024 0.048

1.
1 2 1 1

Pd Pd: Pt
Pt

340.4 3609 363.4

214.4 265.9

2 I' 2
2 2 2 2

0 016 0048 0024

0.016

I'
1

0.024

Pt

pt
Pt
pt

299.7
306.4

1
1 2 2 I 1 1 1 1 1 1 1 1 1 1

265.9
252.0

air.

14h

Rh'
Rk Ir Ir Ir r

343.4.
3.69.2: 212.6 215.7 224.2. 292.4

Ir

Ru.
Os

Ru:
Ru Ri Os Os Os

240.2:
349!8

3728
189.8 225.5 290.9

0.024 0.024 3 0.016 2 0.024 20.024 1 0:040: 2 0.016 3 0.016 2 0.024 2 0.024 2 0.016 1 0.048' 2 0024 2 0,024 3 0,016 1 0.048

c) Automation of ICP Solution Spectrometry for the Determination of Precious Metals

How the performance of an ICP spectrometer of the type presented is distinguished with respect to the precision of internally standardized emission signal quotients and correspondingly accurately attainable concentration values becomes clearly evident from the limits of manual preparation of the sample solutions. If it is intended to accurately measure gold to 1 %o, the sample of about 100 mg must be weighed out to at least 0.1 %o, especially in the case of jewellery alloys. As sample solutions from 1 to 100 g are aliquoted or weighed out both a microbalance and an analytical balante must be employed. The constant changing from one weighing system to another, including the identification of the samples and recording their masses, soon show the limits of manual operation. If one succeeds in methodically taking action to define work patterns, robotization is the

42

ro Gold Bull., 1995, 28(2)

technical consequence of building up an analytical information system that we have named ICP-SIM-ROB meaning Inductively Coupled Plasma SlMultaneous Spectrometer ROBot [37, 38]. An essential prerequisite for successful functioning of this ICP analysis automation is the intrinsic excellence of the automatically performed techniques. The main criterion for this excellence is the theory of the internal standard. If analyte and internal standard element are combined in the sample solution always by exact weighing then the analyte is transported through the system, to be excited in the plasma torch, as a standardized mass and is largely independent of the possibly locally conditioned volume. In other words, accidental concentration effects are excluded. The limits of this analytical concept are easily checked by investigating, for example, the reproducibility of the quotient of the ICP intensities of gold lines Au 267 nm and Au 242 nm by means of a gold solution with about 100 mg/l gold; the scatter is about 0.03 % and is due to the noise in the electronic light detection system [39]. The following three basic methods, each with internal standardization, form the framework of automated ICP analysis: 1. Precision analysis with bracketing calibration and automated production of low and high calibration standards. 2. Precision analysis after automated production of two target concentrations and calibration each time with an automatically produced low and high calibration standard pair. 3. Process analysis with 10, 100 and 1000-fold dilution of the sample solution and calibration with a low and a high standard. These three methods are derived from the approved procedure of solution spectral analysis, with manual preparation of the sample solutions replaced by the robot. Further work routines or combinations of these basic methods are at present not defined, however promising they might be. Apart from the replacement of manual functions, indispensable requirements for the determination of precious metal contents are the quality assurance of the resulting actions and their uninterrupted documentation with the help of the data processing capacity of this ICP information system. Robotization means the thorough utilization of the high value computing centre when the laboratory is unoccupied and 'self-service analysis' by personnel not familiar with the system. Of course one must ensure that

the system adheres to a priority principle in order, for example, to complete the analysis cycle that is in progress. In this respect, a priority handling sequence of suggested solution types is provided with sample registering using a coded authorization. It should be understood that the automated operational modes include standard and internal laboratoryspecific computations and also, for example, the automatic reprofiling of the dual spectrometer. Consequently it would seem to be evident that an effective degree of automation was aspired to for the planned full utilization of this ICP computing centre and the provision of personnel.

THE TYPES OF OPERATION OF THE ICP-SIM-ROB SYSTEM AND ITS CONSTRUCTION

With all possible implementation of automation using current means, the actual purpose of the central ICP measuring installation must not be overlooked. On the one hand, this is the quickest possible automated handling of samples with analytical ICP routine programs. On the other hand, the manual operation of the ICP spectrometer and also the possibility of simultaneously carrying out development of the method while the analytical program is running are crucial points for precious metal analysis. In summary, the ICP-SIM-ROB system can be operated in the following ways by pressing the appropriate function keys: as a manual ICP spectrometer on line with a robot on line with the autosampler on line in a convenient sampler routine for precision analysis following the ISO standard (off line robot) (The last operation i. e. sample preparation by the robot with the ICP spectrometer uncoupled is incorporated in software planning.) The most interesting type of operation of the ICPSIM-ROB system covered in this article is the one which selects an analysis method on the basis of dialogue questioning at the registering station. For the frequently recurring analytical requirements of the do-it-yourself operator the logical link-up with the robot for sample preparation and with the ICP spectrometer for the required analytical program is initiated here. It is not nec-

Gold Bull., 1995, 28(2)

43

Figure 2 The Components of the ICP-SIM-ROBSystem

ICP spectrometer ICP sample bek aspiration pipette rock shaker standard solution process solution and intermediate solution h: cover rack 50 ml bottle i: cover rack 20 ml vial p rocess solution k: balante 1 1: balante 2 a: b: c: d: e: f: g: m: n: o: p: q: r: t: bolde receptien 1 bolde receptien 2 pipette reception 1 pipette reception 2 pipette waste vial dispenser

i b e9

w v

standard stock solution u: sample belt entrance v: screw tapping station w: barcode scanner

s: vial waste

rii
k

O o rl a o

^o s

essary that the operators have a special knowledge of methods. The operation of the entire system for automated analysis is initiated by means of a function key. This also activates the sample registering facility. Figure 2 is a schematic representation of the entire equipment. Figure 3 shows the functional arrangement.

The hierarchical structure with the supervisor PC as control computer for the entire installation, information collecting point and calculating point for analysis results is represented in Figure 4. The entire installation takes up about 30 m 2 of floor space. Figure 5 is a general view of the equipment.

Figure 3

The Funtcional Arrangement of System Components of the ICP-SIM-ROB Information System

CD CD
5

III

1718

21

2 O . I
i iRI s^^s^

6
7

15

25

--------L TI---- -----

i

i
i^Trxox

iAl^aiw

19
--

^-:cnv

^^
O

_ 8

22
23

1b
LPT2

20

3 4
12

10

24 13

26

27

^ 14

29 28

1: Sample registering 11: LRCS system for the preparation of samples III: Supervisor station IV: ICP spectrometer

3: 4: 5: 6: 7:

manual sample designation sampler circulation designation LRCS-PC control of basic functions RS232 link for data

transfer weights

running commands, dilution results,

1: results printer 2: barcode printer

8: robot control

robot application including handling of the samples and transport of samples to lCp 10: 2 analytical balances : sample belt to LRCS system 2: barcode reader station 13: barcode reader station 14: manual transport of undissolved samples 9:

I5:

calculation of the analytical resuits

16:

supervisor software for the entire installation and

22: 23: 24: 25: 26: 27: 28: 29:

2 manual barcode readers 17: ICP onlineloflline 18: single resuits 19: results printer 20: remote results printers 21: ICP control PC

selection of methods results: intensity, concentration sample transfer to !CP by the robot system gtaphit methods development ICP control ICP autosampler or manual operation printer

44

Gold Bull., 1995, 28(2)

A UTOMA TION CONSIDERATIONS

Figure 4

CASE OF
GOLD JEWELLERY ALLOYS

The Hierarchical Structure for the Supervisor PC of the ICP-SIMROB System

ISO/TC174[WG1/N71 specifies a standard method that has the intrinsic qualities required for automated runs. Moreover, the results of manual ICP analysis of gold barcnd prnter according to this method are comparable ICP ^^ resuits printer 2 with those of cupellation techniques (Taresults printer 1 barcode scanner bles 1 and 2). ICP spectrometry is there^batance 1 baiance 2 5i rnat ic $5 fore a valid alternative to cupellation. As bottie tumabie dUuters 1-8 ppetting/metering arm evidenced by the distinctly lower preciICP sipper sion of the ICP results, the preparation of the shaker sample and calibration solutions by manvial dispenser solvent sucking ual weighing appears to represent the crualarm cial limitation of the method. This and screw capping station other considerations provided the motisample belt vation to resort to automation to overcome this problem. vised by the data processing system, the balances and The weighing and dissolving of sample scrapings of associated scanners being connected on line. In this way, gold jewellery alloys are carried out manually. The starting with the registration of samples, data manageweighing operations described in the standard method ment is able to document sample identification, sample 100 mg sample with the microbalance and 100 g soweights, sample preparation, etc. and to make available lution with the analytical balance are strictly superthe results of all process instructions. Using pure gold instead of the sample material, the calibration stock solutions are prepared in the same way. Procedures such as the numerical evaluation of Au fineness are derived from the N71 standard method. Al! that needs to be mentioned here is that a jewellery alloy sample with 585 %o gold leads finally to a sample solution with a concentration of perhaps 58.5 mg gold/kg; the two bracketing standard solutions with 50 and 60 mg/kg gold are produced from standard stock solutions

with 500 and 600 mg/kg gold with the help of the robot to give a 10-fold dilution by exact weighing on analytical balances.

Figure 5

General View of the ICP-SIM-ROB Installation

ro Gold Bull., 1995, 28(2)

45

Figure 6 External Microbalance and Analytical Balance of the ICP-SIM-ROB System

EQUIPMENT COMPONENTS OF THE ICP-SIM-ROB SYSTEM EMPLOYED TO DETERMINE GOLD IN GOLD JEWELLERY ALLOYS
To input the sample details, the operator finds the appropriate display on the registration PC screen for the required precision analysis of a gold jewellery sample. If the system has recognized the analytical task, the printer of the registration PC produces labels with a bar code that identifies the sample, contains instructions for the procedure to be followed by the robot system and provides the basis for the entire data management of the analytical program. In this connection, the registering PC also communicates with the external microbalance and the analytical balance (Fig. 6) which are set up at suitable remote locations; at these balances, by means of the bar code, the necessary sequence of tare and gross weighings of sample and sample solutions is found, displayed on the screen and transferred to the system using the input keyboard.

The sample as well as the standard stock solutions nomenclature of the N71 standard method are in standardized sample bottles and carry the bar code issued by the registering PC. Pressing a button causes the bottles to be transported into the system by means of the sample input conveyor belt. The robot grasps the bottle and puts it on a turntable to read the bar code and identify it by means of a scanner. The robot places the sample and standard stock solutions on the appropriate bottle rack for temporary storage. To prepare the ICP sample solutions from the stored stock solutions, these are introduced with 10-fold dilution into vials, small sealable bottles, which are automatically dispensed from a special magazine. The empty vial is first transported by the robot to one of the two internal analytical balances. To weigh the sample or stock solution into the vial the robot delivers a disposable pipette to the automated pipette metcring station for aliquoting. Following this, the required auxiliary solutions CuC1 2 , NaNO 3 , HCI including the yttrium internal standard solution are weighed out by means of metering pumps (Fig. 7). After combining all necessary solutions the vial, sealed by the capping station, is shaken to homogenize the contents, opened again and with the transport device taken under the pneumatically lowered `sipper' connected to the peristaltic pump; this sucks in the sample solution for nebulizing in the ICP plasma. The transport device has altogether 4 sample receivers so that calibration and measurement can be carried out in carousel operation. Vials not yet required for ICP measurement are stored on the rack.

Figure 7 Metering Pumps of the ICP-SIM-ROB Installation

46

Gold Bull., 1995, 28(2)

THE AUTOMATED DETERMINATION OF GOLD FINENESS OF GOLD JEWELLERY ALLOYS USING THE ICP-SIM-ROB SYSTEM AND EVALUATION
As shown in Tables 1 and 2, evidence is provided that gold in gold jewellery alloys, as they are partly characterized by ISO standard 9202, can be determined by ICP within the laboratory and also by an international round robin analysis. The expenditure of effort for the ISO-ICP method compared with cupellation ISO 11426 has been discussed from numerous aspects. Finally, from the development of the ICP method, the question arises of how to devise an alternative analytical route as good as cupellation to encompass all precious metals. Allowance should be made for the fact that, using cupellation, no other precious metal can be determined as accurately as gold can. As mentioned above, the automatic measurement of gold was already of interest from the point of view of clarifying whether it can attain manual precision or even exceed it. Preliminary experiments already prove that aliquoting with the robot system, i. e. metering of solutions, is accurate enough. In this connection, the two typical solution extractions by the robot were carried out several times, once with the help of disposable pipettes by means of an air volume sucking and expelling the solution through the metering pumps and, for comparison, metering by immediate transfer of the solution by the metering pump, as is the case, for example with the addition of the internal standard solution. Table 4 shows the result. One realizes from the weighing results of metering that an exact weighing of metered volumes is indispensable. As, however, completely automated analysis with reduced precision can be of interest, weighings can be replaced by volumes onder certain conditions to speed up analysis.

Figure 8

Multiple Valve Assembly for the Transport of Calibration Solutions Selected by the Control Computer of the ICP-SIM-ROB-Installation

A particular feature of the ICP-SIM-ROB system is the multiple valve assembly (Fig. 8) which is employed to obtain, for example, the rough gold value. As specified in the ISO N71 standard method, this rough value serves to Eind the bracketing standard concentrations nearest to it; these two standard solutions low and high are thereupon prepared from the appropriate stored standard stock solutions by the robot. The valve unit contains altogether 20 single valves with a common outlet to the peristaltic pump. The inlet side can be connected to the appropriate calibration solution. The valve unit is controlled by the analytical program from the control terminal of the ICP spectrometer. With automated operation, the functional and numerical operations are initiated or performed by the supervisor PC after it has collected the necessary data in the course of the analytical procedure. Its printer issues the analytical results and their statistical data together with identification of the sample. A second printout is issued by the printer of the registering PC, but without statistics.

Fpetng .

Meng

Ra

4.884
0.0674 1.4

g
g %

X
5e v

9.82484:
0.00144 0 0l5

g
g

58

Table 4

Comparison of Pipetting with Disposable Pipettes and Metering a Solution by Pumping

so Gold Bull., 1995, 28(2) ..

47

Finally the automated determination of gold, starting from solutions of gold jewellery alloys is described: First the nickel white gold sample 590 Ni-WG (see Table 1) was analyzed for gold in six separate, manually prepared sample solutions using the convenient sampler method which mainly depends on ISO method N71; the bracketing calibration solutions were manually prepared in just the same way. As is known from the standard method, the single measurement of a solution depends at any given time on five integrations and consequently on five intensity quotients IAu267.5 nm/IY371.0 nm. The approximately 30 minute analysis using the sampler mode produces the analysis printout reproduced in Table 5. Essentially, the efficiency of the ICP computing tentre can be seen here in what, at present, is probably its most effective application, certainly with manual preparation of all measuring solutions. In the final result one

can see that a gold value has been eliminated by Nalimov's outlier test [40]. This outlier test has, above all, the practical significance of making the final result plausible since, with the high accuracy requirement, relatively large deviations can occur through manual errors. It is evident that in the case under consideration the statistica) conditions of ISO method N71 aisfied. (In thef3r example of ant of this convenient sampler mode is given. From such printouts, the results of all gold line channels of the dual simultaneous spectrometer can be extracted see Table 3). The following printouts should show whether by automatic handling of the gold determination, especially with automated dilution, a more favourable result could be obtained. For the four analysis printouts (Table 6) four runs of a sample batch were undertaken with six solutions with separate weighings each time of gold jewellery alloy 1403 (Table 1). The drift test sample (`Wiederfindungsprobe in the printouts) and the brackering standards were also freshl 1r repared Bach time. Anelysls Number: 233 Soiutlon: Sample 4 The drift test sample is a solution Ragfatering: 01/25/9316:29:07 Rep.at 1: 591.127 whose gold concentration is 10 % Data: 01/26193 Time: 11:08:07 Awrege 14: 591,127 Une: Au2675 Solutlon: Sample 5 higher than the low bracketing standSolutlon: Sampl 1 Rep.at 1: 590.912 Repeat 1. 589.926 Averege 05: 590.912 ard solution. A comparison of expected Averege 01: 589.926 Solutlon: Sarnpb 6 Solutlon: Sample 2 and actual values is carried out on this Repeat 1: 590.525 Rep.at 1: 590.486 Averege 16: 590.526 drift test sample solution before and afAverege 12: 590.486 %red: 0,0721878 m Solutlon: Sapte 3 Narmov's-Outllar Tast (95 %): ter the measuring cycle of the sample Repeat 1: 590.866 1.t Run: Averay. of 6: 590.64 Av.rage 13: 590.866 Abs. Standard Div.: 0.42637 batch with a time interval of 3 hours Outiler No. 1: Sample 1 with 589.926 and the per cent deviation printed out %red: 0.0460947 2nd Run: Averege of 5: 590.783 at the top of printouts. This compariAbs. Standard Div.: 0.27232 End of Outlier Tut son is absolutely necessary for automated installations to evaluate unHERAEUS Centra) Analysis Dept. ' "' Analysis Protocol predictable drifts. The drift test sample Analyticel Lab. Dept. : GSC-TSL Coat Center No.: 629 P1,one Number: is calibrated like the sample solutions Client Coat Center No.: Phone Number: Job Number with the same low and high bracketing Analysis Number: 233 standards; it is individually calibrated Registering: 01/25/9316:29:07 for each precision measurement cycle Date: 01/26/93 Time: 11:08:07 and represents, at present, the basis of Sample: 590W9 ICP-Method: eumeas confidence for the unsupervised operaElement Content Au (Meis Line: Au2675) in %o Average of 5 Samples: 690.783 tion of the robot tentre. The case of an 0.27232 Aba. Standerd Deviation automated precision measurement cycle with six solutions degending on single weighings of the 9 ct sample (Table 1) not giving the required basis Table 5 of confidence should be mentioned Original Printout Automated Determination of the Fin eness of Gold (Table 7). A value correction cannot be in a Nickel White Gold Jewellery Alloy According to ISO/N71 with the undertaken with certainty on the basis ICP Sampler Mode and a Manual Preparation of Sa mple and of deviation of the drift test sample as Calibration Solutions; Gold Content Determined by C aspellution: a possible error is not clearly attributed to it. On the other hand, also in this Italy U 590.30 %o, s=0.14 %o (7); Germanv A 590. 6 %o (2)

48

(' Gold Bull., 1995, 28(2)

Nr. 1
Element: Au267b W Ied.Aind ung rprobe: 100.158 Element Content: 588,64 596.74 597.78 591.25 588,63 595.82 Averege: 587,45 Standerd Dev4tnn d neen: 2.118 195: 1,96 1.814 00.34

Run Nr. 2
Element: Au207b

W lederlindung 1probe: 100,399

Element Content:

100,00

585,88 586,00 687,33 591,12 585,70 583.98 Aver.9e: 580,68 Sunderd Den eti n d men: 2.008 A5: 2,43 1.814

Element Content: 666.54 586,74 587,78 ri non: 1.718 0 e95: 568,63 585.82 0,70 1.575 Avara9e: 688,70 Sunde,d D-i b.n

Element Contant: 585,88 588,00 587.33 ri mee: 1.685 0 25: 585,70 583,98 1,20 1.575 Average: 585,77 Standerd Deomtinn

Element Contant. 588,54 586.74 Averpe; 580,43 ei m.x: 1.698 0 0 95: 586,63 585,82 0,42 Standerd D.vl.tinn

Element Content: 585,88 Averepe: ,1 mey: 688,00 588,22 1.708 587.33 0 95: 585,70 0,75 1.845 0 Standerd Ded.Ybn

1 645

Element Content:

Element Content:

586,64 586,74
ri men: 1.271

0 t95.

588,53 0,10 1.409

585,86 ei eux:

686.00 1.253

0 95:

585,70 0,15 1,409

Averega: 686,84 St.ndard Devletien

Avee5e: 585,86

Standerd Devletlnn

Table 6 Original Printouts of 4 runs for the Automated Determination of the Fineness of Gold of a Gold jewellery Alloy with the ICP-SIM-ROB Information System with the Presentation of 6Manually Prepared Sample Stock Solutions; the Fineness of Gold Determination by Cupellation: Italy U 585.32, s=0.13 (7) %o Au Germany A 585.27, s=0.10 (5) %oAu (rWiederfindungsprobe' - Drift Test Sample, % relative)

Element: Au267b Wiederlindntgsprobe: 100.228 Element Content: 587,85 586,91 585,48 592,48 585,98 564,40 Averege: 687,18 ri max: 2.035 Standerd Devietion r95: 2,85 1.814 99,943

Element: Au267b Wiederfindungsprobe: 100,166 Element Content: 685,75 584,95 585.48 604,18 583,95 585,96 Averege: 588,37 ri mem: 2.227 Standerd Devietion r95: 7,78 1.814 100,17

Element Content: 687,85 688,91 585.48 ti max: 0 585,98 584,40 1,32 Averege: 588,12 Standerd Devletidn

Element Content: 685,75 684,95 585,48 0 583,96 585,98 0,80 Avefege: 685,21 Standerd Devietlon

formed consecutively, produces the group of gold values which show prac585,75 564,95 585,46 0 585,96 0 tically no significant outlier according Averege: 585,53 Standerd Devietion 0,44 ri mes: 1.630 195: 1.645 to the normal distribution. The excellence of the automatic determination of gold as a main component in gold example, the consistency of the measurements and an jewellery alloys by ICP spectrometry may be readexceptionally high precision of 0.13 %o gold can be ily understood from this investigation. The limit of seen. A valuable aid for the final presentation of the gold repeatability with a confidence of 95 % is s 1 %o gold; content, for fully automated as well as sampler mode opthus, the method satisfies the conditions of the ISO eration, is the Nalimov test for outliers which, perstandard method [41].
1,459 e95: 1.575 ei mes: 1.761 r95: 1.675 Element Content:

9ement &j26'7b Wiederfindungsprobe 37469 AYea Ed347 2I ri r+hx2, [29

: ;

i 00.676

I00,63

374fl

374.41 r95 11,814'. 374 43;

37458,.

37426

8 4.

Stan+dartl devistlon:0.5,5

Eleme le content: 374.69

37+ 211

374.50'

37426.

Arage 374:
n host; 1424:

Ssaridaurd:dev dcn0.'.lO.

Table 7

Element conrsnt:
0
374.23 Aval e 374.35 ri max 1.319

Standart1 Olgvr4 n ,0,1.3

r95. 1445.

3744z

374.5.0

Automated Determination of the Fineness of Gold of a Gold jewellery Alloy with the ICP374.26: SIM-ROB Information System with Error Display of the Drift Test Sample ('Wiederfindungsprobe') % relative [Cupellation: Italy U375.86, s=0.13 (8) %oAu]

Col Gold Bull., 1995, 28(2)

49

CONCLUSIONS AND PROSPECTS

The ICP-SIM-ROB development project consists of a multi-faceted task concerned with measuring instruments for the analysis of elements in aqueous solution. The relatively high cost equipment of the installed ICP information system has the scope necessary to make a significant contribution in exploring this type of spectrometric measurement technique. The precious metal analysis objective sets the standard for the analytical performance capacity of this ICP computing centre. The system components were designed for the required excellence of the determinations as well as for a high analysis throughput. In addition, unoccupied laboratory time could be made accessible for sample preparation and analysis by automation. The ICP-SIM-ROB computing centre makes it possible to determine gold, for example, with the same precision as manually, making use of automated calibration and aliquoting of presented solutions. The knowledge gained can be readily employed for the equally precise determination of the precious metals platinum, palladium, iridium, rhodium, silver and ruthenium. In its present state of development, ICP measurement technology may be applied as a modular component of equipment to grade the precious metal content of most of the refined goods encountered in precious metal recycling. The chemical and Eire assay conversion techniques make an essential contribution in the preparation of suitable aqueous solutions for ICP measurement. If the necessary precious metal solution concentrations can be successfully prepared, then, by assisting with representative sampling, the samples are capable of being analyzed with a predictable relative accuracy of 0.2 %. For the planned capacity of 500 measuring solutions per day, ascertaining the identity of the sample material and method oriented data plays a crucial role in the ICP information system for guaranteeing control. Suitable ICP spectrometers with the required long term stability of plasma excitation and corresponding signal measurement technology allow themselves to be integrated into a robot system. The robot, including the system components, facilitates the preparation of the solution for ICP measurement from the supplied sample solution with the help of analytical balances. A bonus is that the precision attained enables the consumption of chemical reagents to be minimized. The operations carried out at present as a 'self-service analysis', as shown by example for the precision analysis of gold, are limited to defined tasks [42-44]. With the hardware onder con -sideraton,hclfietnqu,adlog-

led, flexible operation of the ICP-SIM-ROB system, is possible. The off line technique is an operation of the robot centre where, for example, during the operation of the ICP spectrometer/sampler mode combination, solutions are prepared economically from the sample solutions fed in. Sampler mode operation is at present the most favourable variant of the system for full utilization of productivity but, of course, with manual preparation of measuring solutions. Despite the current need and the possibilities available, rationalization is at present being undertaken only cautiously. In particular, this has led to a `fashionable' standstill in the implementation of the robotized laboratory facilities. This condition is even more difficult to understand when it is demonstrated that the route followed to automation effectively represents the possibility of offering the analytical service at reasonable cost. It is, of course, desirable that the more widespread use of these methods should be promoted in order to force the best equipment into the market through competition. The impression should not be given that for the high value ICP-SIM-ROB computing centre all problems regarding this technology have been resolved. On the contrary, this equipment, apart from its routine cmployment, serves to possibly improve the analytical service in the area of precious metals. Unfortunately, solutions of problems of a method or apparatus type are not obtainable `off-the-peg' but require thorough development work. The trend for equipment manufacturers to become less interested in this for thetn less lucrative market for this exacting analytical technology is a further sound argument against wholesale consideration `of the necessity for chemical investigations and the associated budget ... '[45].

ACKNOWLEDGMENTS
Apart from thanks to my co-workers and tolleagues at the Central Analytical Facility of W. C. Heraeus GmbH, I am indebted to very many people and organizations around the world too numerous to mention individually in this article for their contributions to the development of analysis methods for precious metal products and also for the corresponding collaboration in the area of jewellery alloys. My thanks are due to all those who have promoted in every possible way the progress of this work and that of the standards. I should like to thank especially Mr. K.-H. Wiedemann who is entrusted with project management of the ICP-SIM-ROB and also Mr. Klaus Belendorff, Mr. Er-

50

ro

Gold Bull., 1995, 28(2)

hard Berg and Mr. Ronald Rhr who have helped in an advisory capacity with the realization and with looking after the technique. I expressly thank the World Gold Council which has promoted this project within its technical development programma, and, above all, for the co-operative support of Mr. Robert T. Owen, London.

REFERENCES
1. H.-H. Liischow, Edelmetallanalytik eine Ubersicht ber bewahrte klassische und moderne Methoden, Heft 69 der Schriftentreihe der GDMB Gesellschaft Deutscher Metallhtten- und Bergleute, ClausthalZellerfeld, 1993 S. Wissmann, U. Nordheim, Proceedings 17. IPMIKonferenz, Rhode Island, USA, 1993, 117 M. Brilt, Bericht ber das Geschaftsjahr 1988; Fachvereinigung Edelmetalle e. V. Dsseldorf, 1989, 63-84 M. Brilt, Proceedings 11th IPMI Conference Brussels, 1987, 369 M. Brilt, PTB-Bericht des 78. Seminars PhysikalischTechnische Bundesanstalt, Braunschweig, 1989 M. Brilt, 4. Spectro-Seminar `Qualititssicherung durch Qualitiitskontrolle'; `Der analytische Einsatz eines Mini-RBntgenfluoreszenzspektrometers zur Bestimmung der Komponenten von Edelmetallerzeugnissen', Dillingen, Saar, 1994 `Determination of Gold in Gold Jewellery Alloys ICP Solution Spectrometric Method Using Yttrium as Internal Standard Element', Doe. ISO/TC 174, rev. N 71, Gouda, 1992 `Determination of Palladium in Palladium Jewellery Alloys ICP Solution Spectrometric Method using Yttrium as Internal Standard Element', Doe. ISO/DIS 11495, rev. N 69, Gouda, 1992 `Determination of Platinum in Platinum Jewellery Alloys ICP Solution Spectrometric Method using Yttrium as Internal Standard Element', Doe. ISO/DIS

2. 3. 4. 5. 6.

7.

8.

9.

11494, rev. N 67, Gouda, 1992

1994

10. 11.

H.-M. Lschow, Meeting ISO/TC174/WG1, Prag,

14. `Determination of Platinum in Platinum Jewellery Alloys Gravimetric Method after Precipitation of Diammonium-Hexachloroplatinate', Doe. ISO/DIS 11210, rev. N 14, Gouda, 1992 15. Determination of Platinum in Platinum Jewellery Alloys Gravimetric Determination by Reduction with Mercurous Chloride', Doe. ISO/DIS 11489, rev. N 22, Gouda, 1992 16. M. Brilt , Precious Metals Horizon 2000, Nice, Proceedings, 307, cd. by Eurometaux, 1992 17. M. Brilt, Metall47, 1993, 630 18. W. W^ilchli, P. Vuilleumier, Assaying Gold by Cupellation', Aurum, 29 (1987), 56-64; Gold Technology, 3 (1991), 19-27 19. The Gold Institute, Washington DC, The Fire Assay of Gold, January 1985 20. E. E. Bugbee, Textbook of Fire Assaying; cd. J. Wiley & Sons, Inc., New York, 1940 21. N. Petit, Meeting ISO/TC 174 WG 1, Gouda 1992 22. Sh.M. Carson, Proceedings International Precious Metals Conference (IPMI), Brussels 1987, 453-67 23. C. Knosp, personal communication, Pforzheim, 1992 24. K. Aldinger, personal communication, Pforzheim, 1992 25. D. Caloni, personal communication, Arezzo, 1993 26. M. Maleiro-Reymao, P. Baucorps, M.-H. Montech, Precious Metals Horizon 2000, Nice, Proceedings, S 291, cd. by Eurometaux, 1992 27. M. Brilt, K.-H. Wiedemann, Metall, 45(7), 1991, 656-62 28. M. Brilt, K.-H. Wiedemann, Gold Bulletin, 25(1), S. 13-26; World Gold Council, Genf, 1992 29. V. Pelt, J.; Mitteilung Meeting ISO/TC 174/ WG I, Prag, 1994 30. ISO 9202, Jewellery Fineness of Precious Metal Alloys, Ist edition, Genf 1991-09-15 31. Applied Research Laboratories SA. Manual, 3580/ 1186, Ecublens (Schweiz), 1984 32. Bldorn, W.: SPECTRO Analytical Instruments, Kleve 33. Shimadzu Corporation, Tokio 34. Thermo Jarrell Ash Corporation, Manual ICAP 61 E, Part Number 128832 - 01, Franklin, USA, 1991 35. W. Birken, W. Schreiber, Therrno Jarrell Ash, Offenbach, 1993

'Derermination of Palladium in Palladium Jewellery Alloys Gravimetric Determination of Palladium with Dimethylglyoxime', Doe. ISO/DIS 11490, rev. N 16, Gouda, 1992 12. `Determination of Gold in Gold Jewellery Alloys Cupellation Method (Fire Assay)', Doe. ISO 11426, rev. N 87, Gouda, 1992 13. `Determination of Silver in Silver Jewellery Alloys Volumetric (Potentiometric) Method', Doe. ISO
11427, rev. N 88, Gouda, 1992

36. 37.

38. 39.

U. Jerono, D. Marquardt, personal communication, Hanau, 1995 M. Brill, K.-H. Wiedemann, Proceedings Colloquium Analytische Atomspektroskopie, CANAS '93, S. 1005-11 U. Leist, W. Rothermel, A. Seyffer, B. Klein, H. Diehl, ISRA Systemtechnik GmbH, Darmstadt, 1994 M. Brilt, P. Cassagne, European Winter Conference on Plasma Spectrocheinistry, Dortmund, January
1991

Gold Bull., 1995, 28(2)

51

40. V.V. Nalimov, `The Application of Mathematical Statistics to Chemical Analysis', Pergamon Press, Oxford, 1963 41. R. Kaiser, G. Gottschalk, 'Elementare Tests zur Beurteilung von McRdaten', Nr. 774; Bi Wissenschaftsverlag, Mannheim, 1972 42. Heraeus Edelmetall-Bijouterie-Legierungen , Hanau, 1986

Heraeus Werkstoffe Datenmappe, Geschaftsbereich Metalle, Hanau, 1985 44. Heraeus Handbuch fur Edelmetall-Dentallegierungen, Hanau, 1992 45. F. Schneegans, Frankfurter Allgemeine Zeitung, Blick durch die Wirtschaft: `Die blinden Flecken...', 11. Jan. 1995 43.

APPENDIX
Exemples for automated determination of the fineness of gold jewellery alloys according to ISO/N71 with the Convenient Sampler Mode and all gold and yttrium line channels installed in the Dual ICP Simultaneous Spectrometer (yttrium as internal standard). Au 267.5 nm, Au 242.7 nm, Au 197.8 nm, Y 371.0 nm (spectrometer 1) Au 267.5 nm, Y 371.0 nm (spectrometer 2) Table 8 gives a summary of the fineness determination for a sequence of gold jewellery alloys using the Convenient Sampler Mode of the ICP-SIM-ROB Information System. The detailed printout for the 14 ct alloy is given in Table 9.

Table 8

Summary of Fineness Determination of Gold (%o) in Standard Coloured Gold Jewellery Alloys with ICP and the Main Detection Line Au 267.5 nm
Gold content Test allo%, %d AM , 8 ct 9 ct 3761904 I4 ct 585(1403 I8 ct 7501853

333/832

Cest 1
te 2; c t3

333.3,63 333.225
333.195 333.747A 333.404 333.454
-

375.635 375.883::
376.734! 376.576 375.679

tes 5 te; 6+
Ave~ abs. Standard dev:% Au rel. Standard deK: (5) Gupgllar n G rri ny:

376.249
376.118 0.476 0.127

585.685 585.102 685,3.65 585.210 585.437 585.550 585.392 0.215 0.03.7

747.480 747.223 747.82

791.050,
746.51.9. 5 746.096 747.030 0.630 0.084

333.340 0.102 0.031

Cupelltlon iriy

332~96

ICP-An lysitspectr mterA: lCP-Art(y i apectromter A

3.32.001 0.50; 333.30. *0.10

333.05 0. 16

*0.15

37577

375.86
37580

.0, I3

0.10

376.19 0.25

*0.40,

585.33 30;13 58530 - 0.50 585.85 _0.75

585.27 '310:10

74744 0.08

748.10; #1.30
147.32 *0.21

747.00

0.08

52

(' Gold Bull., 1995, 28(2)

Table 9

Original Printout for the Determination of the Fineness of Gold of 14 ct Gold Jewellery Alloy Nr. 1403

Lh . Au2676 Au2427 Au1978 Au267A

PML Nt. 1 685.684998 585.827026 585.844971 595.976013

PM Nr. 2 585.101990 584.23400e 581.0.43030 584.927979

PML Nr. 3 585.364990 585.206970 585.739990 585.650024

PML Nr. 4 585.2)0022 584.465027 581.822996 585.197021

PML N,. 5 585.437012 585.745972 586.017029 585.947021

PML Nr. 6 585.549988 585.794983 583.197988 585.763000

Au 'km 595.392029 585.211975 583.953979 585.577026

. Au 1,,. 0.214840 0.709610 2.207080 0.424640

LI4 Nr.: 789

Registrierung: 0 1/1 719 5 1 2:32:39 AiaIysedatum 01/17/95 Clhrzeit: 15:09:29
L kvr Au2675 L84urq: Probs 1 Repeat 1: 585.685 PML- Mitntw.rt 91. 685.885 L8surq: Prob. 2 R.pwt 1: 585.102 PML- Mitbiwert 82: 585.102 L8.ung: Pro. 3 R.pwt 1: 585.365 PML-Mittdw.n 93: 595.365 LO.ung: P,ob. 4 R.Pwt 1: 585.21 PML-Mittdwat #41; 585.21 LO.unp: Prob 5 R~t 1: 585.437 PML-Mits tw.rt 95: 585.437 LS.urq: Prob. 6 R.Ia 1: 685.55 PMLMittelwert 06: 585.55 %nd: 0.0367002 Ndimov Auurhwnest (95%): 1. Laut: Mittelw. von 6: 585.392 .bs. Sundud.Ew.: 0.21484 End. Au.r.iowt..t. Ri(an .):1.4 9652 Ri(95): 1.814 Ri(99): Lni.: Au2427 L8surq: Probe 1 R.pt t: 585.827 PML-Mitalwwt 81: 375.63 L6.un9: Prob 2 R.peat t: 584.234 PML-Mittelw.n 92: 584.234 L8sung: Probe 3 R.p.t t: 585.207 PML-Mitt.twt 83: 585.207 Ld.,s : ProW 4 R.pwt 1: 584.465 PML-Mitt.lwut 841 : 584.466 L8surq: Probe 5 R.p..t 1: 585.746 PMLMitt.lwut 15: 585.746 L6sunp: Probs 6 Repeat 1: 585.795 PML-Mitnhvt I6: 585.795 %rad: 0.121257 Ndimov Auwtartt (95%): 1. L.ut: 1rrn.1w. von 6: 585.212 .b.. Strdeedlbw.: 0.70981 Ende Au,.iu.r1Mt. fG .n.: 1.510027 Ri(9 : 1.314 RUU ): Lk4i; Au1978 Lbsunp: Pwb. 1 R.pwt 1: 585.845 PML-Mindw.rt t1: 585.845 Lotung: Pro s 2 Report 1: 581.043 PML-Mindwut 82: 581.043 Lthurq: Probe 3 Rpwt 1: 585.74 PML.Min.lwrrt 03: 585.74 L8s : Probs 4 R.P.n 1: 581.883 PML-Mittdwvt 041: 581.883 L6.urp: Probs 5 R.p.m 9: 586.017 PML-Mindwut 85: 586.017 L8.urp: Probe 6 Report 1: 583.198 PML-Min.lwut 96: 683.198 %,W: 0.377954 Ndinwv Auwei"ert.at (95%): 1. L": Mitt.lw. vun 6: 583.854 ab.. St.rW.rdWw.: 2.20708 Eng, Atwr.i..Ntnt. Ri(wt.):1.444 98 Ri(95):1. 8 14 RN99 1: LWe: Au267A L8.vy: Prob. t R.p lt 1: 585,976 PML-Mtt.twrrt 81: 685.976 LMung: Prob. 2 Report 1: 584.928 PML-Mittelwwt 02: 584.928 LMurp: Prob 3 R.peet t: 585.65 PML.Miitatwwt 03: 585.65 LMurp: PrOm 4 R.put 1: 685.197 PML-Min.twert 841: 585.197 L64un8: PrOm 5 RPeat 1: 585.947 PML-Min.lwert 65; 685.947 LS$ung: Prob. 6 R.peat 1; 585.763 PML-Min.w.rt 06: 585.763 Stad: 0.0725165 Narunov Autniss.rt.t (95%): 1, Lmk Mittdw. von 6: 585.577 Mw. StW.rd.bw.: 0,42464 End. AunlswtMt. Ri .n.: 1.87379 Ri(95): 1.814

1.814

1.814

1.814

Ril 91:

1.914

HERAEUS ' Z.ntr W AnMytik AnatyuNabor Abt. Auft,. g. Autrpsnummar Ltd. Nr.: 789 R.yistri.nung: 0111 719 5 1 2:32:39 Anslys.datum:01117n95 U)vtit: t509:29 Prat,.: 585/1403 1cP-M.t),od.: Yi.,stof ELrrnnt8.M7t Au (Hevpttrir Au2675) 1n )oo Mitt.lwert Ob., 6 PML: 585.392
.bs. Si.nd.rdobwwctwrp

An&ywwcbin Ko.)..telie: 629 Kut.rst.IN: Tei.Nr.: T.I. Nr.:

GBC-TSL

0.21484

ro

Gold Bull., 1995, 28(2)

53