Article pubs.acs.org/ac
Development and Verification of Air Balance Gas Primary Standards for the Measurement of Nitrous Oxide at Atmospheric Levels Michael E. Kelley,* George C. Rhoderick, and Franklin R. Guenther Gas Metrology Group, Chemical Sciences Division, Materials Measurement Laboratory, National Institute of Standards and Technology, 100 Bureau Drive, MS8393, Gaithersburg, Maryland 20899-8393, United States S Supporting Information *
ABSTRACT: The Gas Metrology Group at the National Institute of Standards and Technology (NIST) became active in developing primary standards at ambient levels of N2O in the 1980s, and this has continued through to the present. In recent years, interest in NIST-traceable standards has increasednot only at the ambient level of approximately 325 nmol mol−1 (ppb) but at micromole per mole (ppm) levels as well. In order to support two in-process dry whole air standard reference materials (SRMs 1720 and 1721) and the NIST Traceable Reference Materials (NTRM) program, a project was implemented in the Gas Metrology Group to produce a complete suite of new primary standard materials (PSMs) of N2O with synthetic air (O2/N2) as the balance gas. Six levels of dilution, approximately 1 order of magnitude apart, were gravimetrically prepared and verified. Each level serves as the “parent mix” for the next level. This discussion describes the process of producing each level and then verifying its amount-ofsubstance fraction. Expanded uncertainties, k = 2, of 0.025% relative to the gravimetric amount-of-substance fraction were obtained at the ambient level. One standard from the final group of standards at the ambient amount-of-substance fraction level was compared with standards from the National Oceanographic and Atmospheric Administration and the Scripps Institution of Oceanography, two organizations experienced in gas standards preparation and ambient whole air measurements, and shows agreement to 0.07 nmol mol−1 (0.02% relative) and 0.20 nmol mol−1 (0.06% relative), respectively.
N
depends on the tropospheric measurements made by laboratories around the world. However, it is difficult to utilize the many data sets from institutions due to the lack of harmonized calibration scales at low uncertainties. To facilitate accurate measurements of ambient levels of N2O, the Gas Metrology Group of the National Institute of Standards and Technology (NIST) developed a new suite of gravimetric primary standard mixtures (PSMs) covering the range of present and slightly higher amount-of-substance fraction (asf) of atmospheric level N2O. These PSMs would then be used to certify newly developed NIST-traceable standard references materials (SRMs). Additionally, the NIST PSMs would be compared to other key laboratories holding N2O calibration scales, including the National Oceanographic and Atmospheric Administration (NOAA) (NOAA-2007 scale)7 as the central calibration laboratory for the World Meteorological Organization (WMO)6 and the Scripps Institution of Oceanography (SIO) 1998 scale (SIO-98).11,16 This paper describes the development of the NIST N2O
itrous oxide (N2O) is the third most important anthropogenic greenhouse gas after CO2 and CH4, as well as an ozone-depleting specie. It is currently present in the atmosphere at a level of approximately 325 nmol mol−1 (ppb), an increase of ≈55 nmol mol−1 over preindustrial abundance of ≈270 nmol mol−1.1,2 Another greenhouse gas, carbon dioxide (CO2), is present at a significantly higher level, namely, 394 μmol mol−1 (ppm).3 With estimated global emissions of 18.8 Tg (teragrams) of N per year,4 N2O is of great concern due to its global warming potential (GWP). This is the ability of a gas to trap heat in the atmosphere compared with the ability of CO2 to trap heat. N2O has a GWP 300 times that of CO2 on a molecule-to-molecule basis measured over a 100 year time scale.5,6 Nitrous oxide has a lifetime in the atmosphere that has been estimated at 120−150 years.7,8 It is now known to be the major emitted ozone-depleting substance and is predicted to be the largest through the 21st century.9 Currently it is increasing at a rate of approximately 0.6−1.0 nmol mol−1 year−1.10−13 Preindustrial increases in the abundance of tropospheric N2O are thought to be related to changes in agricultural practices.14,15 In the modern era, human influence related to food production is suggested to be the dominant source while other sources include gas-phase oxidation of NH3, biomass burning, oceanic production, and industrial emissions.13 Understanding the distribution and behavior of N2O sources This article not subject to U.S. Copyright. Published 2014 by the American Chemical Society
Received: February 11, 2014 Accepted: March 28, 2014 Published: March 28, 2014 4544
dx.doi.org/10.1021/ac500581b | Anal. Chem. 2014, 86, 4544−4549
Analytical Chemistry
Article
Table 1. Composition of Synthetic Air Balance Gas Composition Table of Zero Air Balance Gas component nitrous oxide (measured) nitrogen (difference) oxygen (measured) carbon dioxide (measured) argon (measured)
cylinder no. SG628086 (μmol mol−1) 0.0001 789544 210083 0.44 373
± ± ± ± ±
0.0001 67 66 0.02 10
cylinder no. H2449089Y (μmol mol−1) 0.0001 789473 210144 0.57 382
± ± ± ± ±
0.0001 51 50 0.02 10
0.0001 789223 210397 0.40 380
± ± ± ± ±
0.0001 47 46 0.01 10
cylinder no. 2435542Y (μmol mol−1) 0.0001 789662 209955 0.48 382
± ± ± ± ±
0.0001 35 34 0.02 10
single-pan high-sensitivity balance (Mettler XP26003L Comparator), (2) an automated turntable that supports three cylinders and the reference mass, and (3) a computer to control the weighing process. The reference is a 10 kg ultraclass mass which serves as the calibration mass for the balance and as the control mass. The control mass is periodically remeasured throughout the weighing cycle to determine if small changes are occurring anywhere in the measurement system. The computer software allows the user to adjust (1) the settling time which allows the balance to come to equilibrium and (2) the number of mass measurement cycles performed on each cylinder (typically 10). During each cycle as many as three PSMs are weighed followed by the reference mass. Standard deviations as small as 1 or 2 mg (for 10 runs) are normal. This procedure is done at each step of the filling process: after the cylinders are evacuated, after the addition of the minor component, and after the addition of the balance gas. These results are then transferred to a spreadsheet where buoyancy corrections based on the temperature (T), pressure (P), and humidity (H) readings are applied and the value assignment and standard uncertainty are calculated. At this point the cylinders are rolled for 90+ min to homogenize the mixtures. Calculating the Amount-of-Substance Fraction. The asf of N2O in a PSM is based on the mass (in grams) of N2O and the total mass of all components, which will have units of grams (g) divided by the total grams of all significantly contributing components in the PSM. The grams of N2O divided by its molecular mass (MM) converts to moles (m). Moles of all other species are determined in the same manner. For any component (x)
standards suite and the comparison to other internationally accepted scales.
■
cylinder no. T352821 (μmol mol−1)
METHODOLOGY
The foundation for all NIST compressed gas primary standards (PSMs) is the kilogram as described in the Système international d’unités (SI).17 High-load, high-sensitivity balances allow very accurate masses of gas to be “weighed” in highpressure gas cylinders. As long as the starting materials are wellcharacterized and cylinder wall interactions are minimal, it is possible to produce gas mixture standards whose compositions are known very accurately. For the production of this entire suite of N2O/air gas standards only two starting materials were used: 99.9999% electronic grade nitrous oxide and a synthetic air consisting largely of oxygen (O2) and nitrogen (N2). The amount of N2O found in this grade of air was extremely low less than 0.1 nmol mol−1 (ppb). Argon was present at a level of 380 μmol mol−1, and carbon dioxide was present at a level of 0.5 μmol mol−1. The N2 was determined by subtraction of the asf of all other components. Since O2 represents ≈21% of the balance gas, and ultimately the molecular mass, it is critical to accurately determine its asf with low uncertainty. At the time, NIST PSMs and SRMs (SRM 2659a) could not meet the uncertainty levels necessary. Therefore, new PSMs of O2 were prepared on a high-precision balance to obtain low uncertainties in both the O2 primary standards and asf determinations of O2 in the balance air. Most critical, however, is the determination of trace N2O in the balance gas thereby obtaining a very low overall uncertainty in the molecular mass of the balance air and the PSM. Very low uncertainties in the trace N2O (±0.1 nmol mol−1) were achieved by analyzing the air by gas chromatography/electron capture detection (GC/ECD). A 10.0 nmol mol−1 N2O in air balance standard was volumetrically produced from a 207 nmol mol−1 PSM to evaluate the synthetic air balance gas. Three replicate injections of the 10.0 nmol mol−1 N2O standard followed by three replicate injections of the synthetic air cylinder were obtained, and the average and standard deviation of those peak area responses were calculated. The 10.0 nmol mol−1 N2O produces a chromatographic peak with sufficient area to obtain meaningful average area counts and standard deviations. The standard deviation was used as the uncertainty (precision) and was also used as the uncertainty for the trace amount in the synthetic air balance gas determinations. Table 1 shows the analyzed composition of the four synthetic air cylinders that were used to prepare the ambient level suite of N2O PSMs. The purity of the electronic grade N2O was verified by the producer. The Gas Metrology Group did not have the capability to quantify the typical contaminants, e.g., N2 and Ar, at the nanomole per mole level. All of the cylinder mass measurements were done on a system that consists of (1) a
xm = xg /xMM
(1)
The quantity of N2O will come primarily from the parent gas and a small (if any) amount from the balance gas. All other constituents from both the parent and the balance gas must be accounted for. This includes any gas species that is expected to be present and occurs in quantifiable amounts. For this suite of PSMs, the non-N2O species measured include O2, Ar, and CO2. The amount of N2 present was determined by difference, i.e., (100% − total of all other measurable species). Other compounds were not present at levels (