Instrumental Setup for Simultaneous Total and Speciation Analysis of

Apr 25, 2017 - Search; Citation; Subject .... Although analysis of metals and metalloids, such as arsenic, is widely spread in many different fields, ...
0 downloads 0 Views 463KB Size
Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Technical Note

A new instrumental set-up for simultaneous total and speciation analysis of volatile metal(loid)s in gas and liquefied gas samples Laura Freije-Carrelo, Mariella Moldovan, Jose Ignacio GarciaAlonso, Thuy Diep Thanh VO, and Jorge Ruiz Encinar Anal. Chem., Just Accepted Manuscript • Publication Date (Web): 25 Apr 2017 Downloaded from http://pubs.acs.org on April 29, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

A new instrumental set-up for simultaneous total and speciation analysis of volatile arsenic compounds in gas and liquefied gas samples Laura Freije-Carreloa, Mariella Moldovana, J. Ignacio García Alonsoa, Thuy Diep Thanh VOb, Jorge Ruiz Encinara* a b

Department of Physical and Analytical Chemistry, University of Oviedo, Julián Clavería 8, 33006, Oviedo, Spain Borealis Polyolefine GmbH, St.-Peter-Straße 25, 4021 Linz, Austria

*Author for correspondence: Dr. Jorge Ruiz Encinar. E-mail: [email protected]

ABSTRACT: Although analysis of metals and metalloids, such as arsenic, is widely spread in many different fields, their analysis in gas and liquefied gas samples is still a challenge. A new GC-ICP-MS set up has been developed for their simultaneous total and speciation analysis in gas and liquefied gas samples without the need of a preconcentration step. An arsine in nitrogen standard was used for optimization and evaluation of the system. Good linearity and detection limits in the very low ppt level for both total and speciation analyses were found. Liquefied butane pressurized under nitrogen and doped with arsine and a propylene real sample from a cracker plant were analyzed using both external calibration and standard additions methods. The good match between both quantifying approaches demonstrated almost negligible matrix effects, even for the total analysis. Application of the approach to check repartition of volatile elements or species between gas and liquid phases was performed in the real propylene sample. Finally, its potential applicability for the simultaneous total and speciation analysis of other elements, such as Hg, was also proved.

Volatile metal(loid)s elements and compounds are present in many different types of gas and liquefied gas samples, such as natural gas or biogas.1,2 Unfortunately, their analysis is still a challenge mainly due to the lack of reliable analytical methods that achieve the required detection limits (ppb-ppt level) and the low availability of gas phase standards.3 Arsenic is one of the most prevalent elemental impurities in such gaseous and liquefied gaseous samples, being of particular interest in many fields. It is highly toxic and poses significant health and safety risks.4 It is present in natural gas, gas condensates and biogases, causing environmental pollution when these gases are burned.2,5 Additionally, it has to be monitored in the polymer industry since its presence as impurity in monomers, such as ethylene or propylene, can affect the final polymer characteristics. It is also known that even trace concentrations of volatile arsenic compounds can cause catalyst poison during such gas processing, increasing the costs for gas exploitation and processing industries.6,7 Importantly, its toxicity depends on the chemical species, being arsine the most toxic one.8 Thus, reliable analytical methods for both total and speciation analysis of arsenic traces in gaseous samples are required. Methods for total As analysis are usually based on collection of the sample in different filters or solvents, such as mixed-cellulose ester filter9 or acid solutions.10,11 Indirect methods are also often employed for total analysis of other metalloids or metals. For example, ISO 6978-1:200312 and

ISO 6978-2:200313 methods for total analysis of mercury in natural gas require the use of chemisorption on iodine impregnated silica gel or collection on supported gold sorbent,

respectively. However, these indirect methods are time consuming (at least 240 min for sampling collection are needed); require high sample volumes, ranging from 15 to 480 L depending on the method and detection limit required and, of course, do not provide speciation information.9,10,11 Additionally, a preconcentration step is also commonly required in speciation methods.14 In NIOSH 6001 method, arsine is collected in solid sorbent tubes with activated coconut shell charcoal. Unfortunately, this method is not completely specific for arsine because other arsenic compounds may be collected on the sampler and would be erroneously reported as arsine.15 Chemotrapping methods have also been applied using different liquids or solids to chemo-trap arsenic species. For example, silver nitrate impregnated silica gel tubes were used for trapping trimethylarsine in natural gas4 and arsines emanating from soils.16,17 Cryotrapping, followed by gas chromatography coupled to inductively coupled plasma mass spectrometry (GC-ICP-MS) or atomic fluorescence spectrometry, was also applied for arsenic speciation in gas standards18,19, biogas5 and gases released from lake sediments.20 These indirect methods require thus the use of chemical reactants and are long and prone to error. However, direct methods are not often applied and just a few of them are compiled in the literature. Volatile arsines in synthetic gaseous samples and air were directly analysed by GC-MS.21 GC was also used with a dielectric barrier discharge detector for analysis of arsine in hydrocarbons.22 Finally, ICP-MS was applied for direct analysis of arsine in nitrogen.23 1

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 6

Figure 1. Diagram of the GC-ICP-MS instrumental set-up developed for simultaneous total and speciation analysis.

It is evident that the use of GC coupled mostly with MS- based detectors for direct arsenic speciation is a very powerful choice. However, it hampers total arsenic determination in gas samples as species that do not elute from column (i.e. particulate matter) might not be detected. Therefore, despite of the variety of individual techniques for total and speciation analysis of arsenic in gaseous samples separately, methods to carry them out simultaneously have not been reported so far. Herein, a conceptually new instrumental approach for such purpose is presented. Liquefied butane doped with arsine and a real propylene sample from a cracker plant were used to validate the new set-up. Moreover, its potential applicability to the total and speciation analysis of other elements, such as Hg, is also shown.

EXPERIMENTAL SECTION Gas standards and samples Arsine in N2 standard (68 ± 20 ppb, v/v), named hereafter AsH3/N2 std., was obtained from Linde AG (Munich, Germany). Liquefied butane pressurized with N2 to 200 bar and doped with arsine was obtained from ISGAS (Texas, USA). Unfortunately, stability and concentration of such synthetic butane sample could not be guaranteed. Real propylene sample was obtained from a cracker plant (Borealis, Stenungsund, Sweden). It contains both gas and liquid propylene phases pressurized to 38 bar, and each phase was extracted from the bottle depending on the exit of the bottle used. Helium doped with 1% Xenon was obtained from Praxair (Madrid, Spain). Instrumentation A gas chromatograph GC 7890B (Agilent Technologies, CA, USA) was modified to be equipped with two interconnected gas sampling valves. A J&W GS-GasPro column (60 m, 0.32 mm i.d.) was used. The GC was coupled to a 7900 ICP-MS by means of a heated GC-ICP-MS interface (Agilent Technologies). A gas dilutor GasMix Aiolos II and a vaporizer

chamber (Alytech, France) were used for online dilution of the samples and standards.

New instrumental set-up A commercial GC-ICP-MS instrument was modified for simultaneous total and speciation analysis of gas and liquefied gas samples. A diagram of the new GC-ICP-MS configuration is shown in Figure 1. Standard/samples and the gas used for dilution are connected to the gas dilutor. If a liquefied gas is analysed, a vaporizer chamber (fixed at 120 ºC) is placed before the gas dilutor inlet for its vaporization before on-line dilution. A 3-ways valve allows to direct the dilutor outlet gas flow to the GC through the gas sampling valves (GSV) 1 and 2 (solid line in the low part of Figure 1). GSV1 and GSV2 are internally connected, so they are loaded at the same time. When turning to the injection position, GSV1 volume is injected through the injection port of the GC to the analytical column. In that way, speciation analysis of the different target species is carried out. Simultaneously, GSV2 volume is introduced directly into the GC-ICP-MS interface using an inert transfer line (fused silica deactivated, 5 m, 0.32 mm i.d.) placed inside the oven and therefore heated to the initial temperature of the GC program. Such gas volume is quickly transferred to the GC-ICP-MS interface arriving to the specific and sensitive ICP-MS detector before the dead volume of the GC injection. Such direct transfer allows performing a multielemental Flow Injection Analysis (FIA) of the ICP-detectable elements present in the gaseous samples. The exit of both the inert transfer line and the analytical column are connected by means of a two holes ferrule to the GC-ICP-MS interface. At this point, they are mixed on-line with the carrier Ar gas flow (previously heated in the pre-heating pipe located inside the GC oven) and transferred to the ICP-MS plasma. In this way, total and speciation analysis of the element of interest can be done at the same injection in a simple and reproducible way. Alternatively, the same 3-ways valve can be used to divert the dilutor outlet gas flow directly to the GC-ICP-MS interface (dashed line in right upper part of Figure 1). Using a cross 2

ACS Paragon Plus Environment

Page 3 of 6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry piece this flow is mixed with the ICP-MS Ar dilution gas before entering the pre-heating pipe. This solution allows: (i) to optimize the ICP-MS detection system by monitoring the continuous signal of the analytes present in the gas matrix and (ii) to carry out a fast screening analysis in order to find out whether the samples contains or not the target elements and therefore, if subsequent quantitative analysis is necessary. Additionally, a second 3-ways valve has been also installed (left lower part of Figure 1) for the control of the type of GC carrier gas. In that way, He/Xe can be employed as GC carrier gas in order to carry out the optimization of the torch position and tune the ICP-MS signal. Finally, an auxiliary electronic pressure control (Aux EPC) controls the addition of a constant pressure of nitrogen into the ICP-MS through the GC-ICP-MS interface when necessary.

Procedures Optimized GC and ICP-MS parameters are summarized in Table S-1. External calibration curves from 0 to 68 ppb were obtained using the gas dilutor. It consists of two mass flow controllers (MFC), calibrated for the desired gases: MFC 1 was used here for AsH3/N2 std. and MFC 2 was used for Ar. The external calibration curve was achieved by mixing a constant flow of AsH3/N2 std. (30 mL/min) with increasing flows of Ar (from 0 to 743 mL/min). Samples were first connected to the vaporizer chamber, whose exit was as well connected to MFC 1 of the gas dilutor. In this way, samples can be diluted on-line with Ar. An additional MFC 3 was used for standard additions quantification in order to internally validate our quantitative results obtained by external calibration. The outlet flow from the gas dilutor was mixed online with the flow coming from the MFC 3 before entering the GSVs using a Tpiece. In that way, sample, standard and dilution flows can be controlled and mixed online. In each point of the curve, a constant flow of the sample was mixed online with a flow of N2 and a flow of AsH3/N2 std. Total flow was kept constant by increasing the AsH3/N2 std. flow while decreasing the N2 one.

RESULTS AND DISCUSSION Optimization and evaluation of the system First, the pre-heated Ar flow (called optional flow, see preheating pipe in Figure 1) used to bring the analytes from the exits of the analytical column and transfer line to the plasma was optimized together with the addition of other gases. In that sense, it is known that the addition of N2 to the plasma can lead to a significant sensitivity enhancement for arsenic.24 Therefore, an auxiliary electronic pressure control (Aux EPC) was installed for this purpose. Although direct introduction of the AsH3/N2 std. could be used for optimization purposes (see dashed line in Figure 1), this option is not recommended for optimizing plasma conditions. Note that using this direct introduction, matrix and analytes reach the plasma at the same time, obtaining plasma conditions similar to total analysis. However, plasma conditions during speciation analysis are slightly different because analytes and matrix are separated in the analytical column and reach the plasma at different time. Therefore, injections of the AsH3/N2 std. were made using the GSVs and the corresponding areas for both total and speciation were computed using different dilution gas – N2 combinations. Best sensitivity was achieved adding to the plasma 0.30.4 L/min of dilution gas and 1-2 psi of N2, which corresponds

to about 10 mL/min of N2. Areas obtained for both total and speciation analyses under these optimized conditions were 17 times higher than those obtained without adding N2. For the optimization of torch position and lenses voltages, a flow of AsH3/N2 std. diluted in Ar (1:20) was directly and continuously introduced from the gas dilutor into the ICP-MS through the pre-heating pipe of the GC-ICP-MS interface (dashed line in Figure 1). Simultaneously, a continuous signal of Xe could be also monitored using He/Xe as GC carrier gas. Optimization curves for torch position and lenses are shown in Figures S-1 to S-9 and optimum values are given in Table S-1. Interestingly, the continuous As and Xe signals showed similar behaviour. Thus, it seems that torch position and lenses voltages could be optimized daily using Xe signal without the need of adding a continuous flow of the AsH3/N2 std.

Figure 2. Chromatogram of arsine in nitrogen (68 ppb, v/v).

As an example, a chromatogram of 68 ppb (v/v) of AsH3/N2 std. and a N2 blank obtained under optimum conditions is shown in Figure 2. As can be seen, a first flat-topped peak was obtained at 0.2 min, corresponding to the total arsenic eluted from the inert transfer line. Later on, a second peak was observed at 3.8 min corresponding to the speciation analysis (arsine eluted from the column). Therefore, total and speciation analysis could be made simultaneously within the same injection. It is worth mentioning that matrix effects could be expected for total analysis as the elements of interest and the sample matrix elute at the same time. In contrast, no matrix effects are expected for speciation analysis as matrix sample and arsine are separated in the column. In this specific case of the analysis of AsH3/N2 std., negligible matrix effects are expected as ICP plasma is already loaded with N2, which is the matrix of the standard. In fact, no As signal perturbation was observed when injecting pure N2 as blank (Figure 2). The difference in the peak area observed for total and speciation analysis is due to the split ratio used (1:6) in the GC injector. Since no adsorptions or losses in the inert transfer line are expected for AsH3, the comparison between the experimentally obtained and the expected theoretical ratios can be used to assess if the total As observed is exclusively in the form of arsine or if other As species are present in the sample. Integration of peak areas in each point of a calibration curve from 0.2 to 68 ppb (n=13) showed a mean recovery of 89 ± 11 % of the AsH3/N2 std. This indicates that, as expected, As is exclusively present as arsine in the AsH3/N2 std. Note that an accurate and precise split ratio is assumed for such mass balance calculation. Moreover, arsenic coelutes with sample matrix in total analysis. Therefore, sensitivity could not be 3

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

exactly the same in both conditions. These two facts could explain the significant uncertainty (12% RSD) observed for the calculated recovery. A calibration curve from 0.2 ppb to 68 ppb was built for total and speciation analysis (Figure S-10 and S-11, respectively). A triplicate injection for 0.2 ppb was made by mixing the lowest flow of arsine that can be set in the gas dilutor (1.9 mL/min) with a highest flow of Ar (743 mL/min), showing a RSD value of 27% RSD. In contrast, RSD obtained were much lower (around 5%) for triplicate injections from 3 to 68 ppb. Method showed a very good linearity within the tested range of concentration for both total (r2=0.9996) and speciation analysis (r2=0.998). Detection limit (DL) was calculated from three times the standard deviation on the chromatographic base line divided by peak height obtained for a known amount of As/AsH3. DL of 2 and 12 ppt for total and speciation analyses, respectively, were obtained. To the best of our knowledge, these are the lowest DL for direct total As and AsH3 speciation analysis in gaseous samples reported so far.

Analysis of liquefied butane doped with arsine For validation purposes, liquefied butane pressurized with N2 to 200 bar and doped with arsine was analysed as a synthetic model sample. The analysis of liquefied gases requires for the use of the vaporization chamber before the gas dilutor inlet. Vaporization chamber was set at 120 ºC and connected on-line with the MFC 1 of the gas dilutor. Initially, it was necessary to proof that arsine was not adsorbed onto the walls of the vaporizer chamber. For this purpose, a triplicate injection of the AsH3/N2 std. was made without using the vaporizer, obtaining areas of 20543 ± 390 and 4337 ± 39 for total and speciation analysis, respectively (uncertainties are given as one standard deviation). When the vaporizer was connected, areas of 20760 ± 675 and 4259 ± 185, respectively, were obtained. Therefore, no losses of As seems to occur when the vaporizer chamber is connected. Pure unspiked liquefied butane was used as blank and injected as well. In contrast to the N2 blank used before (Fig. 2), a small signal was observed for total analysis. This could be due to matrix effects produced by the coelution of the organic butane solvent matrix or to the presence of traces of other elements that could lead to polyatomic interferences (i.e. 40 Ar35Cl). In order to check this, three replicates of both blank and doped butane (both diluted 1:4 in Ar) were analysed using increasing flows of He (from 0 to 4 mL/min) in the collision cell. Total As areas were then computed under the different cell conditions (Figure S-12). The best sample to blank areas ratio was obtained using 2 mL/min He in the cell. A chromatogram of the liquefied butane (diluted in Ar 1:7) obtained under the optimized conditions is shown in Figure S-13. As expected, besides the peak for total arsenic in the void, only one peak of arsine, eluting at 3.8 min, was observed for speciation of liquefied butane. DL for total analysis was then calculated from three times the butane blank standard deviation divided by slope of the calibration curve and turned out to be 20 ppt. Interestingly such DL increased up to 160 ppt when He was not used in the collision cell. Therefore, the use of the He cell allowed minimizing the occurrence of polyatomic interferences. The DL obtained for speciation analysis was 35 ppt. As mentioned before, blank signal observed could be also due to matrix effects caused by coelution of the organic matrix

Page 4 of 6

during total analysis. In other to shed some light on it, besides external calibration, standard additions were carried out by mixing online a constant flow of liquefied butane with different flows of N2 and AsH3/N2 std., keeping constant the total flow. Calibration curves obtained for external calibration and standard additions are shown in Figure S-14 and S-15, respectively. As can be seen, slopes for total calibration curves for external calibration (y=1712x–336) or standard additions (y=1642x+23278) were very similar, proving that matrix effects for liquefied butane were not significant. However, as shown in Table 1, mean total As concentrations obtained using both calibration approaches were slightly different (7%) in contrast to speciation where results were identical. Nevertheless, both values are statistically indistinguishable taking into account the expanded uncertainty at 95% confidence level. These uncertainties were calculated by Kragten spreadsheet

procedure taking into account the uncertainties from the sample signal and those from the slope and intercept of the calibration graph at 95 % confidence level.25 As expected, uncertainties obtained for external calibration curves are lower than those obtained for standard additions. Table 1. Quantitative results (ppb As, v/v) obtained for liquefied butane. Expanded uncertainties (95% confidence level) are given (not including the uncertainty of the AsH3/N2 std.) Total analysis

Speciation analysis

External calibration

92 ± 4

95 ± 4

Standard additions

99 ± 9

94 ± 19

Method validation in terms of quantitative results was difficult due to the lack of reference methods or certified reference materials. Nevertheless, liquefied butane synthetic sample was analysed by the producer (ISGAS) and later on by an independent lab obtaining results of 218 and 40 ppb, respectively. These differences in the quantitative results highlight the challenge of analysing gas and liquefied gas samples and the need of developing new approaches for this purpose as the one presented herein. Please note that results given in Table 1 using both external calibration and standard addition methods are within the concentrations mentioned above. Finally, another point to consider is the high uncertainty of the AsH3/N2 std. given by the producer (68 ± 20 ppb), that was not taken into account in the expanded uncertainty results shown in Table 1. Quantitative results obtained when propagating such uncertainty are given in Table S-2. Unfortunately, the new uncertainty increased up to around 35% RSD. It is clear that further efforts should be made in this field to improve the reliability and stability of the gas and liquefied gas standards. Interestingly, the butane sample was also doped with mercury. Although no standard was available to carry out its quantification, we wanted to demonstrate that the instrumental setup could be used for the simultaneous analysis of any elemental volatile species. The obtained chromatogram is shown in Figure S-16. As can be seen, both total and speciation Hg signals could be easily detected. However, peak profile observed was worse than As (Figure 2) likely due to unspecific adsorption of Hg species along the valves and devices from the gas cylinder to the ICP-plasma. It is clear that the use of inert materials, such as Sulfinert®-treated tubing, would likely reduce such Hg adsorptions, improving peak profiles. Differ4

ACS Paragon Plus Environment

Page 5 of 6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry ent dilutions of the butane sample in Ar were made (from 1:22 to 1:7) and areas for both total and speciation analysis were computed, showing a linear response for both total and speciation analysis (Figure S-17). Therefore, it seems that using an appropriate Hg gas standard, simultaneous total and speciation quantitative analysis of Hg could be made using the instrumental set-up here developed.

Analysis of a propylene real sample For the oil and gas producing industry, it is important to know not only the total As content but also the partition of the occurring As species between the gas and liquid phases.6 Interestingly, both gas and liquid phases of a sample can be analysed using the instrumental set up here developed. To verify its feasibility to this purpose, a real sample containing a mixture of liquefied and propylene gas was analysed using the instrumental conditions optimized for butane analysis. Either liquid or gas phase could be extracted from the bottle sample depending on which of the valves, located in the top or bottom sides of the cylinder, was used. Chromatograms of the gas and liquid phase (diluted 1:3 with Ar) are shown in Figure S-18A. An arsenic peak for total analysis was obtained and again only a single peak of arsine, eluting at 3.8 min, was observed for speciation. Quantification of both gas and liquid propylene phases was performed using external calibration. For the gas phase, 54 ± 4 ppb and 53 ± 13 ppb for total and speciation analysis, respectively, were obtained, whereas 34 ± 3 ppb and 31 ± 5 ppb for total and speciation analysis, respectively, were obtained for the liquid phase. Uncertainty is computed again as expanded uncertainty at 95% confidence level without taking into account the AsH3/N2 std. uncertainty. Interestingly, concentrations obtained for liquid phase are significantly lower than those found for the gas phase analysis, probably due to the high volatility of the arsine species. In order to check again matrix effects, analysis of the gas phase of the propylene sample was carried out by standard additions as well, obtaining 53 ± 12 ppb and 49 ± 11 ppb for total and speciation analysis, respectively. Thus, matrix effects were not apparent at all in this case. Mercury presence was also monitored in parallel in this real sample (Figure S-18B) and, in contrast to the doped butane sample, it could be hardly detected in both phases.

CONCLUSIONS A new GC-ICP-MS set up has been developed for the direct simultaneous total and speciation analysis of volatile species containing ICP-detectable elements in gas and liquefied gas samples. Two interconnected gas sampling valves allow injection of the sample in the analytical column (speciation) and in an inert transfer line (total) at the same time. Detection limits in the low ppt levels were obtained for both total and speciation analysis without the need for any preconcentration step. These two facts made the method really time saving (5 min per injection) and convenient for its implementation in industrial laboratories, where fast analysis are required in order to accept or deny a stock of gas. Liquefied butane pressurized with N2 and doped with arsine and propylene from a cracker plant were successfully analysed, proving the applicability of the developed system to real sample analysis. An instrumental approach to allow quantification by standard additions was specifically developed. It was

found that matrix effects were almost negligible, for both total and speciation analysis in the analysed samples. However, such standard addition possibility could be extremely useful for heavier samples, where much more significant matrix effects could be expected (i.e. hydrocarbon mixture). Interestingly, the multielemental nature of the ICP-MS detection opens the door to simultaneous quantitative analysis of other elemental species (i.e. Hg) under the condition that adequate standards are available. The only limiting factor of this approach is the high uncertainty of the arsine gas standard. This point deserves much more attention from the metrological community if these determinations are regulated in the future.

ASSOCIATED CONTENT Supporting Information Instrumental parameters, calibration curves, chromatograms of the liquefied butane and propylene sample and quantitative results taking into account the uncertainty of the standard (PDF).

ACKNOWLEDGMENT This research has been supported by Agilent Technologies, Borealis and Fundación Universidad de Oviedo (project FUO-EM 02215). Josep M. Sanjuán from Agilent Technologies Spain is thanked for his help during instrumental development.

REFERENCES (1) Feldmann, J. J. Anal. At. Spectrom. 1997, 12, 1069–1076. (2) Caumette, G.; Lienemann, C. P.; Merdrignac, I.; Bouyssiere, B.; Lobinski, R. J. Anal. At. Spectrom. 2009, 24, 263–276. (3) Carter, K. E.; Gerdes, K. Spectrochim. Acta Part B At. Spectrosc. 2013, 85, 34–44. (4) Uroic, M. K.; Krupp, E. M.; Johnson, C.; Feldmann, J. J. Environ. Monit. 2009, 11, 2222–2230. (5) Pinel-Raffaitin, P.; Le Hecho, I.; Amouroux, D.; Potin-Gautier, M. Environ. Sci. Technol. 2007, 41, 4536–4541. (6) Krupp, E. M.; Johnson, C.; Rechsteiner, C.; Moir, M.; Leong, D.; Feldmann, J. Spectrochim. Acta - Part B At. Spectrosc. 2007, 62, 970–977. (7) Ivandini, T. A.; Yamada, D.; Watanabe, T.; Matsuura, H.; Nakano, N.; Fujishima, A.; Einaga, Y. J. Electroanal. Chem. 2010, 645, 58–63. (8) Cullen, W. R.; Reimer, K. J. Chem. Rev. 1989, 89 (4), 713–764. (9) OSHA, Inorganic Arsenic in Workplace Atmospheres, Method Number ID-105, T-ID105_FV-01-9105-M, Division of Physical Measurements and Inorganic Analysis, OSHA Technical Center, Utah, 2007. (10) U. S. Environmental Protection Agency, Part 60, Method 29, Determination of Metal Emissions from Stationary Sources, Washington, D.C., 1994. (11) Xu, R.; Tang, D. Z.; Yan, Q. T.; Xu, H.; Wang, S. Y.; Tao, S.; Han, Z. X. Energy and Fuels 2015, 29, 3863–3869. (12) International Organization for Standardization ISO 6978-1:2003 Natural gas - Determination of mercury - Part 1: Sampling of mercury by chemisorption on iodine, Geneva, Switzerland, 2003. (13) International Organization for Standardization ISO 6978-2:2003 Natural gas - Determination of mercury - Part 2: Sampling of mercury by amalgamation on gold/platinum alloy, Geneva, Switzerland, 2003. (14) Chen, M. L.; Ma, L. Y.; Chen, X. W. Talanta 2014, 125, 78–86. (15) NIOSH National Institute for Occupational Safety and Health Manual of Analytical Methods, Method 5001, Arsine, 1994 (16) Mestrot, A.; Feldmann, J.; Krupp, E. M.; Hossain, M. S.; RomanRoss, G.; Meharg, A. A. Environ. Sci. Technol. 2011, 45, 1798–1804. (17) Mestrot, A.; Uroic, M. K. Environ. Sci. Technol. 2009, 43, 8270– 8275. 5

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(18) Ilgen, G.; Huang, J.-H. J. Anal. At. Spectrom. 2013, 28, 293–300. (19) Feldmann, J.; Naëls, L.; Haas, K. J. Anal. At. Spectrom. 2001, 16, 1040–1043. (20) Yuan, C.-G.; Zhang, K.; Wang, Z.; Jiang, G. J. Anal. At. Spectrom. 2010, 25, 1605-1611. (21) Pantsar-Kallio, M.; Korpela, A. Anal. Chim. Acta 2000, 410, 65– 70.

Page 6 of 6

(22) Gras, R.; Luong, J.; Hawryluk, M.; Monagle, M. J. Chromatogr. A 2010, 1217, 348–352. (23) Suh, J. K.; Kang, N.; Lee, J. B. Talanta 2009, 78, 321–325. (24) Hu, Z.; Gao, S.; Liu, Y.; Hu, S.; Chen, H.; Yuan, H. J. Anal. At. Spectrom. 2008, 23, 1093-1101. (25) Kragten, J. Analyst 1994, 11, 2161–2165.

For TOC only

6

ACS Paragon Plus Environment