Determination of ruthenium in photographic materials using solid sampling electrothermal vaporization inductively coupled plasma mass spectrometry† Yi Hu,a Frank Vanhaecke,a Luc Moens,*a and Richard Damsa and Ingrid Geuensb aLaboratory of Analytical Chemistry, University of Ghent, Proeftuinstraat 86, B-9000 Ghent, Belgium bR&D Analytic Research, Agfa-Gevaert N. V., Septestraat 27, B-2640 Mortsel, Belgium Received 9th November 1998, Accepted 11th January 1999 Ru was determined in photographic emulsions and films using electrothermal vaporization inductively coupled plasma mass spectrometry (ETV-ICP-MS) after minimal sample pre-treatment (no separation or preconcentration required).The emulsion samples were either (i) converted into a colloidal solution by dissolution in warm water or dilute HNO3 ( liquid sampling) or (ii) dried at 105 °C (solid sampling) prior to analysis. For the analysis of photographic films, only solid sampling, requiring no sample pre-treatment, was used.By optimization of the ETV heating programme, on-line separation of Ru (analyte) and Ir (internal standard) from the Ag matrix was accomplished: #90% of the Ag present was removed prior to the vaporization of Ru and Ir. Quantification of Ru was accomplished by single standard addition, whereby Ir was used as an internal standard. The absolute limit of detection was found to be#1 pg. The results obtained showed a good agreement with those obtained by pneumatic nebulization ICP-MS and/or electrothermal atomic absorption spectrometry after sample dissolution. sampling, sample preparation can be reduced to a minimum, Introduction which significantly reduces the risk of sample contamination During the production of photographic emulsions, trace or analyte losses.Moreover, since the samples are not diluted, amounts of precious metals (e.g., Au, Rh, Ru or Ir) are the detection limits may be improved. In our laboratory, purposely added in order to obtain desirable characteristics in research has been carried out concerning direct solid sampling terms of light sensitivity.1–4 In order to assess the quality of with ETV-ICP-MS.As, Se, Cd, Sb and Hg have been deterthe photographic materials thus produced, accurate quantitat- mined successfully in samples of diVerent origin.12–16 Also, for ive determination of these elements is important. However, liquid samples, the use of ETV as a sample introduction the often low concentrations of these dopants and the heavy technique21–25 oVers some distinct advantages over pneumatic Ag matrix make this analysis diYcult.In addition, separation nebulization: (i) the limit of detection can be significantly of the precious metals from the Ag-containing matrix prior to improved owing to the enhanced sample transport eYciency their determination is often not self-evident. (#80%), (ii) small sample volumes (5 ml ) can be analyzed Inductively coupled plasma mass spectrometry (ICP-MS) is and (iii) on-line matrix removal is sometimes possible when a powerful analytical technique, showing extremely low limits using an appropriate heating programme.Since ETV has of detection, multi-element capabilities, a wide linear dynamic already been extensively used for many years in combination range and a high sample throughput. Originally, ICP-MS was with AAS and ICP-OES, valuable information (e.g., temperamainly intended for the analysis of aqueous samples and ture programming, the use of chemical modifiers) is available hence, pneumatic nebulization is the most widely used method in the literature.for sample introduction. Pneumatic nebulizers owe their suc- In this work, the determination of Ru in photographic cess to their low cost, instrumental simplicity, high sample emulsions was carried out using liquid and solid sampling throughput and good stability. However, pneumatic nebuliz- ETV-ICP-MS after minimal sample pre-treatment (no separaation also shows important drawbacks.The sample transport tion or preconcentration required). Ru in photographic films eYciency is very low (in combination with a spray chamber, was determined using direct solid sampling ETV-ICP-MS. typically 1–2%), while the sample must be dissolved and a relatively large volume of solution is required. Hence, eVorts have been made to couple alternative sample introduction systems to ICP-MS in order to extend its application range. Experimental At present, solid samples can be directly analyzed using laser Instrumentation ablation (LA)5–10 or in some cases also using electrothermal vaporization (ETV)11–20 for sample introduction.The poten- The present study was carried out using a commercially tial of ETV-ICP-MS for ‘solid sampling’ has been recognized available graphite furnace of the ‘boat-in-tube’ type (SM-30 for some years and some publications on the subject have Gru�n Analytische Mess-Systeme, Ehringhausen, Germany), appeared, but most of these papers are concerned with analys- originally developed for ETAAS.This graphite furnace was ing slurries17–20 rather than (dry) solid samples. With solid modified for coupling with ICP-based instruments (ICP-OES, ICP-MS). The description of this modification has been published elsewhere.26 The heating programme of the furnace was †Presented at the 8th Solid Sampling Spectrometry Colloquium, Budapest, Hungary, September 1–4, 1998.controlled by an in-house developed computer program. The J. Anal. At. Spectrom., 1999, 14, 589–592 589Table 2 ICP mass spectrometer: instrument settings and acquisition Table 1 Graphite furnace temperature programme parameters Step Duration/s Temperature/°C ICP-MS PE SCIEX Elan 5000 ‘Drying’ step 10 100–120 ‘Ashing’ step Ia (for matrix removal ) 30 800 Rf power 1 kW Plasma gas flow rate 12 l min-1 ‘Ashing’ step IIb (for matrix removal ) 60 1400 ‘Intermediate’ step 10 200 Auxiliary gas flow rate 1.2 l min-1 Carrier gas flow rate 0.75 l min-1 (to switch the valve to the ‘measuring’ position and allow the plasma Scanning mode Peak hop transient Acquisition points per peak 1 to stabilize) ‘Vaporization’ step 15 1900 Sweeps per reading 1 Readings per replicate 300 ‘Intermediate’ step 40 No heating (to switch the valve to the ‘venting’ Dwell time 30 ms Total measuring time 30 s position after the data acquisition has been completed) ‘Cleaning’ steps 2×5 2400 into the sample holder for subsequent analysis.An advantage aThis pre-treatment step is aimed at removing the film base (PET) of this approach is that the analytes of interest are enriched when analyzing a sample of photographic film. bThis pre-treatment by a factor of approximately five. Photographic emulsion step is intended to remove Ag to the largest possible extent. normally contains #100 mg of Ag per gram of emulsion.Photographic films. For film samples, first the weight per furnace temperature was monitored using an optical pyrometer square centimetre was determined. This was accomplished by (PY20, Gru�n Optik, Ehringhausen, Germany). taking 9 cm2 of film and accurately determining the corre- The ETV system was coupled to a Perkin-Elmer SCIEX sponding weight. In order to determine the Ru content, the Elan 5000 ICP mass spectrometer (Perkin-Elmer, OVenbach, film was cut into small pieces (of about 1.5 mg each) using a Germany) via a 10 mm id silicone rubber tubing.A three-way ceramic knife. These film samples were loaded into a sample valve was used to vent vapours generated during the drying, holder and accurately weighed. Thereafter, the sample holder ashing and cleaning steps. As a result, only vapours generated was inserted into the graphite furnace for subsequent analysis. during the vaporization step of the heating programme were An additional ashing step was used for the film in order to allowed to reach the plasma, so that deposition of vaporized remove the film base (polyethyleneterephthalate or PET), matrix material (on the torch, interface and lens stack), before the vaporization of the analyte, see Table 1.contamination of the inace pump oil and plasma over- Photographic film normally contains #2 gm-2 Ag. loading could be reduced to a minimum. The flow rate of the Ar carrier gas was controlled by means of a mass flow Calibration and internal standardization controller (Model 5876, Brooks Instruments, Veenendaal, The In previous work carried out in our laboratory,12 it was shown Netherlands).that in solid sampling ETV-ICP-MS, single standard addition The settings for the ETV system and the ICP mass specis an accurate, fast and practical calibration strategy. Three trometer are listed in Tables 1 and 2, respectively. The following measurements of the blank, five measurements of the sample isotopes (abundances given in parentheses) were monitored and five measurements of the sample to which an analyte spike since they are free from inter-element isobaric interference: of appropriate concentration is added are suYcient to obtain 99Ru (12.7%), 101Ru (17.0%), 103Rh (100%) and 193Ir (62.7%). an accurate and precise concentration value.Hence, in this work, single standard addition was used for calibration. Standards and reagents In solid sampling ETV-ICP-MS, the use of an internal All reagents used were of analytical-reagent grade or higher standard is often imperative.12 In contrast to pneumatic nebulpurity.Water was de-ionised and further purified using a ization ICP-MS, for which a close match in mass number Milli-Q water purification system (Millipore, Bedford, MA, between the analyte elements and the internal standard is the USA). Ru, Rh and Ir standard solutions were prepared from most important selection criterion,27 similar volatility of the commercially available 1 g l-1 single-element standards, by analyte element and the internal standard is a more important appropriate dilution with 20% HCl for Ru and with 1% HNO3 issue in solid sampling ETV-ICP-MS.In this work, Ir and Rh for Rh and Ir. HNO3 (65%) and HCl (32%) were purified by were tested as potential internal standards, as their volatility sub-boiling distillation in quartz equipment. is similar to that of Ru. The experiments showed that biased results were obtained when Rh was used as an internal Sample preparation standard, probably because of a spectral interference at m/z 103, the origin of which could not be identified. Consequently, Photographic emulsions.Two ways of sample pre-treatment Ir was chosen as an internal standard to correct for signal were used: suppression due to residual matrix eVects (matrix-induced For ‘liquid sampling’ ETV-ICP-MS, approximately 500 mg signal suppression or enhancement) and to improve the pre- of emulsion were first dissolved in 100 ml of warm (#40 °C) cision of repeated measurements (repeatability).The signal of water (or 0.14M HNO3) to form a colloidal solution. the argon dimer (Ar2+), which is always present during plasma Thereafter, a micropipette was used to transfer an appropriate operation, was always monitored as an indicator of residual amount of this solution (10 ml ) into a sample holder, which matrix eVects.28 was inserted into the graphite furnace for subsequent measurement.Although a pure AgCl emulsion can be dissolved in Analysis procedure concentrated ammonia very easily, this dissolution method was not used in this work, because photographic emulsions An appropriate amount of standard solution (Ru) and/or internal standard (Ir) was pipetted into the sample holder and often contain AgI or AgBr, which cannot be dissolved in an ammonia solution. subsequently dried under an infrared lamp prior to sample loading.This preliminary drying facilitates solid sample load- For ‘solid sampling’ ETV-ICP-MS, a solid sample, which is easy to handle, was obtained by drying the emulsion (overnight ing and it has been shown in earlier work12 that—at least for some elements—an identical behaviour of the analyte (i) in at 105 °C). About 0.5 mg of dried sample was directly loaded 590 J. Anal. At. Spectrom., 1999, 14, 589–592the solid sample and (ii) in the standard solution can only be guaranteed under these conditions.When liquid sampling is used, 10 ml of sample solution are inserted into the sample holder by means of a micropipette and dried under an infrared lamp before introduction into the furnace. For solid sampling, the sample holder was first tared using a microbalance (readability of 1 mg; M3P, Sartorius, Go� ttingen, Germany) and subsequently the solid sample was loaded and weighed. Finally, the sample holder was inserted into the graphite furnace with the aid of a pair of tweezers, which can slide on a rail (rigidly mounted in front of the furnace), allowing reproducible loading.Results and discussion Fig. 2 Vaporization curves for Rh and Ir (no ashing step was applied, the heating time is 15 s for each corresponding temperature). In photographic emulsions and films, Ag is present at a high concentration: #100 mg g-1 in emulsion and #2 gm-2 in film, while precious metals, such as Ru, are added at trace or ultratrace levels (usually around 1 mg g-1 Ag).Owing to its high mass number and relatively low ionization potential, Ag gives rise to severe matrix eVects in ICP-MS. Strong suppression of the analyte signals by a 1 g l-1 Ag matrix was observed when pneumatic nebulization ICP-MS was used for the multielement analysis of photographic materials, carried out previously at our laboratory.29 On the other hand, separation of Ag from Ru is diYcult owing to the low concentration of the latter and the complex matrix involved (predominantly AgX and gelatin).One of the advantages of ETV as a means of sample introduction is that on-line separation of the analyte(s) of interest from the matrix element becomes possible if an appropriate heating programme is used, provided that the Fig. 3 Separation of Ru and Ir from the Ag matrix obtained by elements of interest and the matrix elements show a suYciently heating a sample of film II (about 0.1 mg). A 10 ml volume of diVerent volatility. 500 mg l-1 Ru and 50 mg l-1 Ir standard solution was added and an OmniRange setting of 30 was used for the Ag+ signal in order to In order to investigate the possibility of separating Ag from prevent overloading of the detector. Ru and to select an appropriate internal standard, the vaporization curves of Ag, Ru, Ir and Rh were recorded using a 50 mg l-1 standard solution of these elements. The results obtained are given in Fig. 1 and 2. As can be seen from these vaporization curves, Ag starts to volatilize at a temperature of about 1000 °C while Ru, Rh and Ir are volatilized at temperatures of about 1500 °C or higher.Hence, on-line separation of Ag from Ru, Rh and Ir is possible. The vaporization curves of Ru, Rh and Ir agree well with those obtained by Byrne et al.,30 who studied the vaporization mechanism of the platinum group elements (PGEs, e.g., Ru, Rh, Pd, Os, Ir and Pt) in the graphite furnace in detail. They suggest that, except for Os, all PGEs are vaporized as metal vapour sublimed from metal deposited on the graphite surface after oxide decomposition. Next, the ETV heating programme Fig. 4 Signal profiles for Ru, Ir and Ar2+ observed for a sample of was further optimized, with the aim of maximizing the volatiliz- film II (vaporization step). ation of Ag in the ashing step, while minimizing the losses of Ru, Rh and Ir during this step. As a result, the heating programme shown in Table 1 was used in all further work.For both film and emulsion samples, this multi-step heating programme allowed #90% of the Ag matrix to be removed prior to the vaporization of Ru and Ir. The separation of Ru and Ir (used as an internal standard) from the Ag matrix is clearly visible from Fig. 3, which was obtained by heating a ‘real-life’ film sample using the heating programme listed in Table 1. Two peaks are observed for Ag. It is tempting to attribute this behaviour to the occurrence of two diVerent Ag species in the film sample.However, a similar signal profile was observed for an AgNO3 solution. Hence, it is more likely that Ag previously condensed on cooler parts of the furnace and/or migrated to some extent into the graphite the sample Fig. 1 Vaporization curves for Ag and Ru (no ashing step was applied; the heating time is 15 s for each temperature). holder is released at the higher temperature of the vaporization. J. Anal. At. Spectrom., 1999, 14, 589–592 591Table 3 Results in ppm (mg Ru per g Ag); standard deviations are given in parenthesesa Emulsion I Emulsion II Film I Film II Ru added 1 ppm 10 ppm 10 ppm 50 ppm Liquid sampling — 11.0 (0.4, n=5) — — ETV-ICP-MS Solid sampling 1.36 (0.12, n=5) 11.0 (0.8, n=5) 13.3 (0.9, n=5) 52.1 (1.6, n=5) ETV-ICP-MS ETAAS 1.06 (0.09, n=3) 11.1 (0.2, n=3) 12.6, 12.6 45, 49.1 Pneumatic 1.34 (0.03, n=4) 11.4 (0.7, n=4) — — nebulization ICP-MS aThe Ag concentrations for emulsions I and II are 88.8 and 88.6 mg g-1, respectively.Films I and II contain 1.94 and 2.07 g m-2 Ag, respectively. Two samples consisting of a pure AgCl emulsion, provided References by Agfa-Gevaert and doped with diVerent amounts of Ru, 1 T. H. James, The Theory of the Photographic Process, Macmillan, were measured using both liquid and solid sampling ETVLondon, 4th edn., 1977. ICP-MS. Two film samples which were coated with Ru-doped 2 R. S. Eachus and M. T. Olm, Cryst. Lattice Defects Amorphous AgCl emulsions provided by Agfa-Gevaert were also measured.Mater., 1989, 18, 297. Signal profiles for Ru, Ir and Ar2+ obtained during analysis 3 R. S. Eachus and M. T. Olm, Nippon Shashin Gakkaishi, 1991, of a sample of film II are given in Fig. 4. The almost unsup- 54, 294. 4 D. Volman, G. Hammond and D. Neckers, Advances in pressed smooth curve of the argon dimer signal indicates that Photochemistry, Wiley, New York, 1992, vol. 17. the removal of the matrix during the ashing step is eVective 5 P.Arrowsmith, Anal. Chem., 1987, 59, 1437. and that only a minimum amount of matrix is left after this 6 E. R. Denoyer, K. J. Fredeen and J. Hager, Anal. Chem., 1991, step. Both 99Ru and 101Ru were monitored and the total Ru 63, 445A. contents calculated using both signal intensities were in excel- 7 J. S. Crains and D. L. Gallimore, J. Anal. At. Spectrom., 1992, lent agreement, indicating that no significant spectral inter- 7, 605. 8 N. J. G. Pearce, W. T. Perkins and R. Fuge, J. Anal. At. Spectrom., ferences occurred. Hence, either of these two isotopes can be 1992, 7, 595. used for the determination of the Ru content by ETV-ICP-MS. 9 S. F. Durrant and N. I. Ward, Fresenius’ J. Anal. Chem., 1993, The results for both the emulsion and film samples are listed 345, 512. in Table 3. For comparison, the results obtained by ETAAS 10 E. H. De Carlo and E. Pruszkowski, At. Spectrosc., 1995, 16, 65. (Agfa-Gevaert) and by pneumatic nebulization ICP-MS 11 J.Wang, J.M. Carey and J. A. Caruso, Spectrochim. Acta, Part B, (University of Ghent), after taking the sample into solution, 1994, 49, 193. 12 F. Vanhaecke, S. Boonen, L. Moens and R. Dams, J. Anal. At. have also been included. Spectrom., 1995, 10, 81. As can be seen from Table 3, the agreement between the 13 S. Boonen, F. Vanhaecke, L. Moens and R. Dams, Spectrochim. ETV-ICP-MS results and the results obtained by ETAAS and Acta, Part B, 1996, 51, 271. pneumatic nebulization ICP-MS is very good.On each 14 G. Galba�cs, F. Vanhaecke, L. Moens and R. Dams, Microchem. occasion, the diVerence between the average ETV-ICP-MS J., 1996, 54, 272. result and the corresponding reference value (ETAAS result 15 F. Vanhaecke, S. Boonen, L. Moens and R. Dams, J. Anal. At. Spectrom., 1997, 12, 125. or pneumatic neulization ICP-MS results) is <10%. In 16 F. Vanhaecke, I. Gelaude, L. Moens, and R. Dams, Anal. Chim. addition, the precision of the results obtained (average RSD Acta, in the press.<10% for n=5) is also satisfactory. It is important to mention 17 U. Voellkopf, M. Paul and R. E. Denoyer, Fresenius’ J. Anal. that sample inhomogeneity may deteriorate the results Chem., 1992, 342, 917. obtained.31 Hence, in each case a suYcient number of sub- 18 D. C. Gre�goire, N. J. Miller-Ihli and R. E. Sturgeon, J. Anal. At. samples should be analysed to obtain a reliable result. Spectrom., 1994, 9, 605. 19 S. Hauptkorn, V. Krivan, B. Gerken and J. Pavel, J. Anal. At. The limit of detection was determined using an empty Spectrom., 1997, 12, 421. sample holder as a blank and was calculated according to the 20 M. J. Liaw, S. J. Jiang and Y. C. Li, Spectrochim. Acta, Part B, 3s criterion (IUPAC). The absolute limit of detection was 1997, 52, 779. found to be approximately 1 pg. When liquid sampling is used, 21 J. M. Carey and J. A. Caruso, Crit. Rev. Anal. Chem., 1992, this corresponds to a relative value of 0.1 mg l-1, taking 10 ml 23, 397.as a typical sample volume. For solid sampling, this is equival- 22 R. W. Fonseca. and N. J. Miller-Ihli, Spectrochim. Acta, Part B, 1996, 51, 1591. ent to 1 ng g-1, taking 1 mg as a typical sample mass. 23 H. Naka and D. C. Gre�goire, J. Anal. At. Spectrom., 1996, 11, 359. From the results obtained, it is clear that ETV-ICP-MS can 24 C. C. Chang and S. J. Jiang, J. Anal. At. Spectrom., 1997, 12, 75. be considered as a promising technique for the determination 25 L. Yu, S. R. Koirtyohann, M. L. Rueppel, A. K. Skipor and of trace and ultratrace amounts of Ru in photographic emul- J. J. Jacobs, J. Anal. At. Spectrom., 1997, 12, 69. sion and film samples. 26 P. Verrept, R. Dams and U. Kurfurst, Fresenius’ J. Anal. Chem., 1993, 346, 1035. 27 F. Vanhaecke, H. Vanhoe, R. Dams and C. Vandecasteele, Talanta, 1992, 39, 737. Conclusion 28 F. Vanhaecke, G. Galba�cs, S. Boonen, L. Moens and R. Dams, Solid sampling ETV-ICP-MS is a fast and accurate analytical J. Anal. At. Spectrom., 1995, 10, 1047. 29 Y. Hu, F. Vanhaecke, L. Moens and R. Dams, Anal. Chim. Acta, technique for determining trace amounts of Ru in photo- 1997, 355, 105. graphic materials. Since only single precious metals are nor- 30 J. P. Byrne, D. C. Gre�goire, M. E. Benyounes and C. L. mally added to the photographic emulsion as a dopant, it can Chakrabarti, Spectrochim. Acta, Part B, 1997, 52, 1575. be predicted that the content of Ir in photographic materials 31 F. Vanhaecke, J. Diemer, K. G. Heumann, L. Moens and R. can also be determined in the same way, but using Ru as an Dams, Fresenius’ J. Anal. Chem., 1998, 362, 553. internal standard. Furthermore, it is also to be expected that this analytical method can be applied to the determination of the other precious metals. Paper 8/08751B 592 J. Anal. At. Spectrom., 1999, 14, 589