Subscriber access provided by University of Otago Library
Article
Direct Measurement of Trace Elemental Mercury in Hydrocarbon Matrices by Gas Chromatography with Ultraviolet Photometric Detection Ronda Gras, Jim Luong, and Robert A. Shellie Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.5b02895 • Publication Date (Web): 11 Oct 2015 Downloaded from http://pubs.acs.org on October 27, 2015
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 5
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
Direct Measurement of Trace Elemental Mercury in Hydrocarbon Matrices by Gas Chromatography with Ultraviolet Photometric Detection Ronda Gras(1,2,3), Jim Luong(1,2), Robert A. Shellie(2,3)*. AUTHOR ADDRESS 1 Dow Chemical Canada ULC, Highway 15, Fort Saskatchewan, Alberta, T8L 2P4, Canada 2 Australian Centre for Research on Separation Science (ACROSS), University of Tasmania, Private Bag 75, Hobart, Tasmania 7001, Australia 3
ARC Training Centre for Portable Analytical Separation Technologies (ASTech), University of Tasmania, Private Bag 75, Hobart, Tasmania 7001, Australia
KEYWORDS elemental mercury, gas chromatography, ultraviolet detection ABSTRACT: We introduce a technique for direct measurement of elemental mercury in light hydrocarbons such as natural gas. We determined elemental mercury at parts-per-trillion level with high precision [< 3% RSD (n=20 manual injection)] using gas chromatography with ultraviolet photometric detection (GC-UV) at 254 nm. Our approach requires a small sample volume (1 mL) and does not rely on any form of sample pre-concentration. The GC-UV separation employs an inert divinyl benzene porous layer open tubular column set to separate mercury from other components in the sample matrix. We incorporated a 10-port gas-sampling valve in the GC-UV system, which enables automated sampling as well as back flushing capability to enhance system cleanliness and sample throughput. Total analysis time is < 3 2 2 min, and the procedure is linear over a range of 2 to 83 µg/m [R = 0.998] with a measured recovery > 98% over this range.
0
Elemental mercury (Hg ) exists in the Earth’s crust within a concentration range from 10 ppb (w/w) to 20 ppm 0 (w/w). Hg can be released into the environment from a variety of natural sources like forest fires and volcanic eruptions, or via anthropogenic activities such as fossil fuel combustion and metal production [1]. Natural gas production also frequently generates hydrocarbon streams contain0 0 ing traces of Hg [2]. Hg can have a negative impact on industrial equipment; for instance, mercury in Liquefied Natural Gas (LNG) can form amalgams in cryogenic aluminum exchangers. This can cause corrosion and led to industrial mishaps like that encountered in 1973 in Skikda, Algeria where heat exchangers unexpectedly ruptured and caused catastrophic damage to this world scale LNG plant [3]. Since mercury can also have a negative impact on human health, numerous incentives have been issued to control and limit mercury emissions. A number of approaches for the determination of mercury are reported in the literature. These include Inductively Coupled Plasma Mass Spectrometry (ICP-MS) [4,5], Inductively Coupled Plasma Optical Emission Spectrometry (ICPOES) [6], Microwave Plasma Optical Emission Spectroscopy (MP-OES) [7], and Cavity Ring-Down Spectroscopy (CRDS) [8]. Cold Vapour Atomic Absorption Spectroscopy (CVAAS) [9-12] and Cold Vapour Atomic Fluorescence
Spectroscopy (CVAFS) [13-14] are by far the most common techniques used for mercury determination. These techniques are popular due to their sensitivity and simplicity in addressing liquid and solid matrices where mercury can be found like polluted water, fish, mining, and industrial waste to name a few. CVAAS and CVAFS are approved methods by health and environmental agencies such as ASTM, OSHA, USEPA and others. Many of these technologies rely on enrichment techniques using elemental gold or elemental silver to form an alloy commonly referred to as amalgam [15]. Subsequently, the amalgam formed is thermally desorbed to attain low limits of detection (LOD) for the targeted analyte [16-17]. These pre-concentration approaches typically require 10 mL to 1 L of sample to reach the reported detection limits, 3 3 which range between 0.2 ng/m to 1 ng/m . While these techniques are sufficient in determining the presence of total mercury, they often do not have the capability to speci0 ate the type of mercury involved such as Hg against inorganic/ionic mercury compounds like cinnabar, a reaction between elemental mercury with sulfur and other mercury species. The methods involved typically require sample preparation such as acid digestion. Yet another constraint is gold amalgam which is used for gaseous sample or gas liberated from other matrices. Amalgams are formed rapidly
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
in the presence of air, but not in an inert atmosphere which can exist in nitrogen purged reactors or vessels or hydrocarbon pipelines [18]. Thus, there exists an unmet need for a practical, simple, 0 and flexible direct measurement method for Hg in hydrocarbons such as natural gas or other light hydrocarbons (less than C4). This method should also be convenient and reliable enough for deployment in remote areas or in quality control laboratories. Here we introduce a GC-UV system developed for direct measurement of parts-per-trillion level of elemental mercury in natural gas or light hydrocarbons. Analyzing elemental mercury by gas chromatography as not been strongly considered as an analytical technique yet mercury possesses favourable vapour pressure that makes it very suitable for the technique described especially when used with a UV detector. Our approach uses a small sample volume (1 mL), which is injected directly into the GC-UV system without recourse to any sample pre-concentration. Separation of 0 Hg from the sample matrix is accomplished by using a pair of divinyl benzene porous layer open tubular (PLOT) columns housed in metal capillaries. The PLOT columns provide appropriate chromatographic selectivity and we marry this with back flushing capability to ensure system cleanliness and maximize sample throughput. The total analysis time using our approach is 126 s. Detection is achieved using a Reducing Compound Photometer (RCP) for selective detection of elemental mercury. RCP is also known as Reduction Gas Detection (RGD), and was conceived as a GC detector for determination of trace carbon monoxide and hydrogen for the electronics industry [19,20]. The compounds of interest are detected by proxy by photometric detection of Hg0 which is liberated according to Equation 1. Other compounds such as olefins elicit a respectable response by post column reaction with the packed bed of a mercuric oxide [21-23]. II
Xg + Hg Os -> XOg + Hg
0 g
(Eq. 1) 0
Since our target analyte is Hg , we eliminated the redundant mercuric oxide bed but retained other components of the RCP. We will show that peak tailing caused by reactor kinetics or geometry [21] is eliminated, which leads to highly satisfactory performance.
•
EXPERIMENTAL SECTION
A Peak Performer 1 (PP1) gas chromatograph (Peak Laboratories, Mountain View, CA) was used for all experiments. The mercuric oxide reactor was removed from the PP1 GC platform and replaced with deactivated metal tubing with identical dimensions. The detector was set at 120 o 0 C for all experiments. The Hg in the sample passes directly into a quartz optical cell, where it is measured by its atomic absorption of light from a mercury lamp source at 253.7 nm. The PP1 GC was fitted with two 5 m × 0.53 mm i.d. × 10 µm PoraBOND Q PLOT metal capillary columns (Agilent Technologies, Middelburg, The Netherlands). The columns were connected to a Valco C10 UWP rotary valve using Valco FSR 0.8 adapters (VICI-Valco Instruments, Houston, TX, USA). The valve was fitted with a 1 mL sampling loop. The first PoraBOND Q PLOT column was used as a precolumn for back flushing purpose while the second PoraBOND Q PLOT column was employed as the main analytical column. A modified split / splitless inlet (Agilent Tech-
Page 2 of 5
nologies, Little Falls, USA) was interfaced upstream from the valve in the analytical flow path to facilitate manual injection. Slow manual injections were conducted using 1 mL gastight Hamilton syringe (Hamilton Company, Reno, NV, USA) to avoid inlet over-pressurization where the sample can contact the septum resulting in losses. The inlet was fitted with a 4 mm borosilicate glass liner (Agilent Technologies, USA) and operated in direct injection mode in all experiments at a temperature of 150°C. Connections between the inlet and the valve as well as from the valve to the detector were made using deactivated 0.25 mm i.d., deactivated Ultimetal plus tubing (Agilent Technologies, Middelburg, the Netherlands). The column oven was set isothermally at 45 °C for all experiments and nitrogen carrier gas was provided at 30 mL / min. and the inlet temperature was 150 °C. In this configuration, the unit can be operated in automated or unattended operation mode. 0
Hg was obtained in the local laboratory and was placed in a sealed vial filled with mercury free air which was equilibrated in a water filled Dewar flask. The injection of known amounts of air saturated with mercury vapour into the analytical system was used for calibration. The vapour densities 0 of Hg in air at 1 atm and various T were calculated by using the ideal gas law and mercury vapour pressure fitted to a linear least square equation in log vp versus 1/T [12]. 1 mL o 0 of air at 20 C and 1 atm at equilibrium over Hg l is 13.19 ng 0
of Hg g.
RESULTS AND DISCUSSION
• 0
Hg has a substantially high vapour density, making it suitable to be characterized by gas chromatography. Recent advances in surface passivation have resulted in much more inert tubing intended for use as a transfer line or chromatographic components like sampling valve bodies, rotors, and inlet surfaces [24,25] combatting the tendency of 0 Hg to react with untreated metal surfaces. In the present investigation we used deactivated metal tubing between the inlet and the sampling valve, and between the sampling valve and the detector. These transfer lines were heat traced at 120 °C to avoid sample condensation. Metal capillary columns were also chosen to enhance durability and to maximize on-line compatibility. Divinyl benzene based PLOT columns offer a high degree of inertness and provide good separation of elemental 0 Hg from chromatographic interferences. Two PoraBOND Q PLOT columns were selected for the present application. The first was employed as a pre-column, while the second acted as an analytical column. The columns were housed in a small oven, which was operated isothermally at 45 °C the lowest temperature the column oven module is able to reliably maintain. We elected to use isothermal operation to maximize throughput by eliminating post-analysis cool down and re-equilibration. To ensure the selectivity of the analytical system is not compromised, a number of analytes commonly encountered in natural gas were evaluated. No chromatographic interference was encountered with hydrogen, carbon monoxide, air, argon, carbon dioxide, methane, ethane, and propane even at the double digit percent levels. 1,3-butadiene (tR = 76 s) is a potential interference since it absorbs light at 254 nm, however it is well resolved from mercury (tR = 46 s) under the conditions used. Heavier hydrocarbons such as butane, pentane, and hexane are well retained by the PoraBOND Q stationary phase. The
2 ACS Paragon Plus Environment
Page 3 of 5
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
high degree of selectivity presented an opportunity to incorporate a back flushing step in the analytical workflow to enhance system cleanliness and to reduce total analysis time. To this end, we employed the rotary injection valve for back flushing. Figure 1A shows the valve in Position A for load and back-flushing mode while Figure 1B shows the valve in Position B for inject mode. For manual injection the rotary valve starts in position B and the injected sample flow through the injection loop. By rotating the valve to Position 0 A at 46 s, the Hg peak passes through the detector without being affected by the baseline perturbation caused by valve actuation, while 1,3-butadiene is back-flushed to vent along with any heavier hydrocarbons present in the sample.
For comparison, Figure 3 shows the GC-UV result for a 1 mL manual injection of a sample containing 1,3-butadiene without back flushing mode engaged. Butadiene is eluted from the column as shown in the figure.
Figure 1. A) Diagram of valve position in “load/and backflushing” of the pre-column mode. Figure 3. Chromatogram of 1 mL injection 1% 1,3-butadiene in nitrogen.
Figure 1. B) Diagram of valve position in the “inject” mode.
Analytical figures of merit for the GC-UV approach are highly satisfactory. Repeatability of less than 3% (n = 20) was obtained at two different concentrations, namely 83 3 3 0 µg/m and 8.3 µg/m of Hg in natural gas. We determined 3 our approach to have a detection limit of 1.7 µg/m and a 3 3 linear range of quantification 2 µg/m to 83 µg/m with a correlation coefficient of 0.9981. Average recovery of mercury was measured over using 5 spiked samples ranging from 5 to 60 µg/m3 and found to be 98% (n=3). Figure 4 shows a chromatogram of a commercially available natural 3 0 gas sample spiked with 8.3 µg/m of Hg . Note the great sensitivity obtained with respectable chromatographic symmetry. While other techniques offer lower detection limits, this approach requires small sample size, no preconconcentration and can function in air or inert atmospheres such as nitrogen or helium.
3
0
Figure 4. Chromatogram of 8.3 µg/m of Hg spiked in commercially available natural gas. Figure 2 shows a chromatogram of a 1 mL manual injection 3 0 of a standard containing 83 µg/m of Hg in air. The baseline perturbation at 59 s is due to the valve actuation.
•
CONCLUSIONS
We successfully developed and implemented a GC ap-
3
0
Figure 2. Chromatogram of 83 µg/m of Hg in air.
proach using metal divinyl benzene PLOT columns and a 0 modified RGD for the determination of Hg . This approach is suitable for a variety of matrices including inert atmosphere, ambient air, natural gas, and light hydrocarbons of 0 up to butane. Ultra-trace detection of Hg in various matrices was conducted directly without sample preconcentration; with a complete analysis time of less than 2 min and with a sample volume of 1mL. By combining gas chromatography using selective col0 umns for Hg with ultraviolet photometric detection, great
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
selectivity and sensitivity for the analytical system can be achieved. The approach has a high degree of reliability, and is easy to perform. Future work includes the incorporation of a sample enriching device using gold or silver trapping medium to further enhance method sensitivity.
AUTHOR INFORMATION Corresponding Author *Associate Professor Robert A Shellie e-mail
[email protected] tel +61-3-6226-7656
Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.
ACKNOWLEDGMENT This article is dedicated to Dr. Martine Stolk on the occasion of her retirement for being an inspiration in the development of science and technology at the Dow Chemical Company. Special thanks to Alexander Lowe of Peak Laboratories, Clayton Bleile of Chrome Enterprises, Dr. Taylor Hayward of Kelly Services on assignment to Dow Chemical, and Kate Heymans of Dow Chemical Canada for their encouragement and support. This project was partially funded by the Analytical Technology Center’s Technology 2014 Renewal and Development Program and the Australian Research Council Training Centre for Portable Analytical Separation Technologies (Project IC140100022). Peak Laboratories is gratefully acknowledged for the use of the GC system. Robert Shellie is the recipient of an Australian Research Council Australian Research Fellowship (Project DP110104923).
REFERENCES
Page 4 of 5
(4) Taylor, V.; Carter, A.; Davies, B., Jackson, P. Anal Methods. 2011,3, pp 1143–1148. (5) United States Environmental Protection Agency, EPA Method 200.8 1994. (6) United States Environmental Protection Agency, EPA Method 200.7 2001. (7) Sanz, J.; de Diego, A.; Raposo, J.C.; Madariaga, J.M. Anal. Chim. Acta., 2003, 486, pp 255-267. (8) Fain, X.; Moosm, H.; Obrist, D. Atmos. Chem. Phys. 2010, 10, pp 2879-2892. (9) United States Environmental Protection Agency, EPA Method 245.1 2011. (10) United States Environmental Protection Agency, EPA Method 7470A 2005. (11) United States Environmental Protection Agency, EPA Method 7471B 2007. (12) Long, S.J., Scott, D.R., Thompson, R.J. Anal. Chem. 1973, 45, 2227-2232. (13) United States Environmental Protection Agency, EPA Method 245.7 2012. (14) United States Environmental Protection Agency, EPA Method 1631E 2007. (15) Sabri, Y.M.; Ippolito, S.J.; Tardio, J.; Bhargava, K. J. J. Phys. Chem. C., 2012, 116, pp 2483-2492. (16) American Society for Testing and Materials, Method D6350-14, West Conschohocken, Pennsylvania, USA 2014. (17) American Society for Testing and Materials, Method D5954-98, West Conschohocken, Pennsylvania, USA 2014. (18) Arizona Instrument LLC, operating manual for Jerome J431 gold film sensor mercury analyzer, Arizona, USA 2012. (19) Ostrander, C. US patent 4,411,867, October 25, 1983. (20) Ostrander, C.; Oharra, D.; McDowell, C.; Hartmann, S., US patent 6,368,560, April 9, 2002. (21) Lin, M.; Gras, R.; Luong, J.; Bleile, C.; Shellie, R.A. LC-GC North America, 2015, 33 (5), pp 332-338. (22) Cao, X.L.; Hewitt, C.; Waterhouse, K. J. Chromatogr. A 1994, 679, pp 115-119. (23) Cao, X.L.; Hewitt, C.; Waterhouse, K. Anal. Chimi. Acta 1995,300, pp 193-200. (24) Smith, D. Thin Film Deposition – Principles and Practice, McGraw-Hill, New York 1995. (25) Dobkin, D.; Zuraw, Z. Principles of Chemical Vapour Deposition, Kluwe Academic Publishers, Dordrecht, 2003.
(1) UNEP Chemical branch The Global Atmospheric Mercury Assessment: Sources, Emissions, and transport, UNEPChemicals, Geneva, 2008. (2) U.S. Environmental Protection Agency (USEPA) Mercury in Petroleum and Natural gas: estimation of emission from production, processing, and combustion, EPA-600/R-01-066, 2001. (3) Kinney, G.T. Oil & Gas J. 1975,73, pp 37-45.
4 ACS Paragon Plus Environment
Page 5 of 5
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
Authors are required to submit a graphic entry for the Table of Contents (TOC) that, in conjunction with the manuscript title, should give the reader a representative idea of one of the following: A key structure, reaction, equation, concept, or theorem, etc., that is discussed in the manuscript. Consult the journal’s Instructions for Authors for TOC graphic specifications.
5 ACS Paragon Plus Environment
