Anal. Chem. lg82, 5 4 , 2611-2612
2811
Preparation andl Stable Isotope Determination of NBS-16 and NBS-17 Carbon Dioxide Reference Samples Tyler B. Coplen" and Carol Kendall U.S. Geologlcal Survey, 432 National Center, Reston, Vlrglnia 22092
In September 1976, a consultants' meeting was convened by the International Atomic Energy Agency to discuss interlaboratory calibration of stable isotope measurements in natural samples and to review stable isotope measurements of water standards. Participants recognized that to improve agreement of results olbtained in different laboratories there was a need for stable isotope reference samples other than water (1). In response to this need, two carbon dioxide reference samples, NBS-116and NBS-17, were prepared for oxygen and carbon stable isotope ratio calibration of mass spectrometers. Becauise the preparation of several hundred isotopically homogeneous carbon dioxide reference samples proved to be a nontrivial exercise and because this method can be used to prepare gaseous laboratory reference samples for internal use to keep an accurate check on drift of the working standard (Z),the preparation technique is presented below. NBS-16 was prepared from carbon dioxide taken directly from a tank of Air Products liquified carbon dioxide, produced as an industrial byproduct. NBS-17 was prepared from carbon dioxide taken directly from a tank of Chemetron Corp. (Carbon Dioxide Division, Dallas, TX) gaseous carbon dioxide, obtained from a carbon dioxide well in the southwestern United States. This selection ensured that there would be a large difference in carbon and oxygen isotope ratios between the two NBS reference gases. Sample purification and aliquot preparation were carried out on a vacuum line shown in Figure 1. After opening all 20 500-cm3volumes to the manifold and evacuating the line, the system is filled to approximately 70 kPa (10 psia) with carbon dioxide from the desired tank. The carbon dioxide is purified by three cycles of freezing with liquid nitrogen and sublimation with a dry ice slurry. The carbon dioxide is frozen into the Hoke stainlenci steel cylinder, the Nupro valve is closed, and the carbon dioxide is allowed to sublime and to homogenize at ambient temperature for at least 4 h to eliminate isotopic inhomogeneities produced by freezing and subliming the carbon dioxide. The volume of the stainless steel cylinder was selected so that the pressure of the carbon dioxide remains less than 6 MPa to prevent formation of liquid carbon dioxide. The carbon dioxide is expanded into the 20 500-cm3volumes and it is homogenized by mixing with a calculator-operated mercury piston (3) for 30 min. The stopcock on each 500-cm3volume is closed. Each of these volumes should now contain high purity carbon dioxide identical in carbon arid oxygen isotopic abundance. To fabricate containers for each aliquot of carbon dioxide, 6 mm 0.d. glass tubing is cut into 40-cm lengths and sealed at one end. These tubes can be opened later using the tube cracker of Des Marais and Hayes (4)or Coleman (5). Twenty tubes are loaded into the vacuum line and evacuated. Carbon dioxide from the first 500-cm3volume is expanded into the manifold and 20 tubes and it is homogenized by mixing with the mercury piston for 110 min. The stopcock for each tube is closed (see Figure l), the carbon dioxide is frozen for 5 min, the manifold is evacuated, noncondensible gases are pumped away, and each tube is sealed by a hand torch with dual flame tips. As shown in Figure 1, the volume between each tube and its associated stopclock on the manifold is relatively large compared to the size of ithe manifold; consequently, most of the carbon dioxide from each 500-cm3 volume ends up in
~-
Table I. Carbon Isotope Ratio Abundances of NBS Reference Samples, in o/M,, Relative to PDB ref sample
no. of prepns
NBS-16 NBS-17 NBS-20
std dev, 10
6 '3C
4 8
0.03 0.02
11
0.02
-41.61 -4.48 -1.06
Table 11. Oxygen Isotope Ratio Abundances of NBS Reference Samples, in o//,
no. of std dev, b ref sample prepns 10 s'~OV-SMOW~ 61 8 0 ~ ~ ~ - ~ ~ t 3.89 -36.09 NBS-16 3 0.01 t21.99 -18.71 3 0.04 NBS-17 t26.64 -4.14 7 0.03 NBS-20 a 6 l80relative to the V-SMOWreference water (1 ). The oxygen isotopic fractionation factor of 1.0412 recommended by Friedman and O'Neil ( 7 ) was used to calculate 6 relative to carbon dioxide data in this column. produced by reaction at 25 "C of 100%phosphoric acid of, carbon with PDB calcite, assuming ~ ' * O ~ D B . C O dioxide evolved by reaction at 25 "C of 100%phosphoric acid with NBS-80 (Solenhofen limestone) is -4.14°/ao (8).
tubes. Each tube should contain about 300 wmol of gas. Carbon dioxide from each of the other 19 500-cm3volumes is loaded into 6-mm tubes in the same manner. Oxygen and carbon isotope ratio determinations of samples from the same 500-cm3sampling volume and from different sampling volumes prepared with the apparatus in Figure 1 are identical within the precision of the analysis, d=0.05%0 and f0.07%0for carbon and oxygen, respectively. The carbon and oxygen isotope abundances of NBS-16 and NBS-17 were determined on a double-focusing double-collecting mass spectrometer (6) and are shown in Tables I and 11. Oxygen isotope results have been normalized so that 6l80of SLAP is -55.5%0 relative to V-SMOW as recommended by Gonfiantini ( I ) . Mass spectrometric corrections were made after Mook and Grootes (9). Comparison of these two reference samples with NBS-18, NBS-19, NBS-20, and several other stable isotope reference samples is given elsewhere (10). In ow fiist attempt at preparing gaseous reference samples, we expanded high-purity carbon dioxide (99.995%) from a tank through a high-quality diffusion resistant regulator directly into glass tubes. We heated the tank of liquified carbon dioxide above the critical point (31 "C) to eliminate isotope fractionation during phase change. The carbon dioxide was frozen into 6 mm 0.d. glass tubes. We assumed viscous flow would eliminate any isotopic fractionation among the aliquots. After preparation of more than 100 tubes, the oxygen isotope ratio variability among a random selection was greater than 0.5%0. The reason for this variation is unknown. The next attempt at preparation of references employed the apparatus in Figure 1except that the Hoke stainless steel sampling cylinder was not used. Freezing and subliming during the purification process produces such isotopic inhomogeneity throughout the system that 30 min of mixing with a mercury piston is insufficient to homogenize the carbon dioxide in the 500-cm3volumes. The variability in the oxygen isotope abundance of carbon dioxide from different 500-cm3
Thls article not subject to US. Copyrlght. Publlshed I982 by the Amerlcan Chemlcal Soclety
Anal. Chem. 1982, 54, 2612-2613
2812
TO
THERMOCOUPLE GAUGE
DIFFUSION PUMP
7,( ),
1
1 ~
through the volumes for 30 min is sufficient to eliminate measurable isotopic inhomogeneities in the system. For preparation of hydrogen and nitrogen gases it is desirable that both the volume of the manifold and the volume between each tube and its associated stopcock (item D in Figure 1) be as small as possible. Hydrogen and nitrogen are sealed in 9-mm 0.d. glass tubes with a hand torch without freezing or sorbing the gas. This line has now been used to prepare hydrogen, nitrogen, and carbon dioxide reference samples that are isotopically homogeneousas measured by state-of-the-art isotope ratio mass spectrometry.
1 ~
l l
ACKNOWLEDGMENT
I '
d -
TOTAL OF 20 RESERVOIRS AND GAS EXTRACTORS
Glass vacuum preparatlon line for preparation of homogeneous carbon dioxide reference samples: (A) Hoke 500-cm3stainless steel sampling cylinder with a burst pressure of at least 13 MPa; (B) Nupro 4BK-series stainless steel valve with KeCF stem tip; (C) 50km3 ghss sampling volumes (total of 20); (D) 100-cm3glass volumes (total of 20); (E) size 7 Ace-Thread connector with FETFE O-ring (Ace Glass Inc., Vlneland, NJ) for 6-mm 0.d. tubing or size 11 Ace-thread connector for 9-mm 0.d. tubing (CaJonO-ring Ultra-torr unions are also ('I4 In.) glass tube about 40 cm long. satisfactory); (F) 6"or 6%" Figure 1.
volumes was 1%. Consequently, subliming the carbon dioxide in the stainless steel cylinder is a critical step in preparing homogeneous reference samples. Although an extra expense is involved, some users may wish to install a stainless steel bellows pump in the system to cycle gas through it. Then additional stopcocks and interconnections are needed so that all 500-cm3volumes can be placed in series with each other and with the pump. Cycling gas
We thank John Laughlin, Jessica Hopple, Vicki Ober, Judy Rowe, and Freddie Leighty for sample preparation and analysis of these reference samples. We appreciate the reviews of this work by our colleagues I. Lynus Barnes and P. Morales.
LITERATURE CITED Gonflantlnl, R. Nature (London) 1079, 271, 534-536. Cralg, H.; Gordon, L. I. In "Stable Isotopes In Ocenaographlc Studies and Paleotemperatures"; Tonglorgl, E., Ed.; V. Lischl: Plsa, 1965; pp 9-130. Coplen, T. B. Anal. Chem. 1981, 53, 940-942. Des Marais, D. J.; Hayes, J. M. Anal. Chem. 1976, 4 8 , 1651-1652. Coleman, D. D. Anal. Chem. 1981, 53, 1962-1963. Coplen, T. B. Int. J . Mass Spectrom. Ion Phys. 1973, 1 1 , 37-40. Friedman, I.; O'Neil, J. R. Geol. Surv. Prof. Pap. ( U . S ) 1977, 440KK, 12 p. Craig, H. Geochlm. Cosmochim. Acta 1957, 12, 133-149. Mook, W. G.; Grootes, P. M. Int. J . Mass Spectrom. Ion Phys. 1973, 72, 273-298. Coplen, T. 8.; Kendall. C.; Hopple, J., submitted for publlcatlon in Nature (London).
RECEIVED for review July 13, 1982. Accepted September 3, 1982. Use of trade names and trademarks in this publication is for descriptive purposes only and does not constitute endorsement by the U.S. Geological Survey.
Elution of Disposable Octadecylsilane Cartridges with Hydrophobic Organic Solvents Sharon L. Pallante," Martln Stognlew,' Michael Colvln, and Danlel J. Llberato DepaHment of Pharmacology and Experimental Therapeutics, Johns Hopkins School of Medicine, 7 2 5 North Wolfe Street, Baltimore, Maryland 2 1205
The determination of various xenobiotic substances in biological fluids or other complex matrices usually involves protein precipitation, filtration, and solvent extraction in order to prepare the samples for analysis by either spectrometric or chromatographic techniques. These cleanup procedures are required to remove a multitude of interfering compounds. The removal of these interfering substances is essential in order to obtain a pure sample by subsequent procedures or, if quantitative studies are required, a signal-to-noise ratio of sufficient intensity to ensure accurate and reproducible measurements. Recently, prepacked octadelylsilane columns or cartridges (ODSC) have seen wide use in sample cleanup. Organic substances are adsorbed onto the ODSC from biological fluids and subsequently recovered by elution with water or mixtures of water and water misible solvents (e.g., methanol and acetonitrile) (1). In use, the ODS column or cartridge has already proven its superiority over liquid-liquid extraction for the removal of organic salts and sugars from biological samples. Present address: Pharmaceutical Division, Ciba-Geigy Corp., Suffern, NY 10901. 0003-2700/82/0354-2612$01.25/0
While this procedure is extremely useful and provides mixtures of organic substances free of the biological matrix, it would be highly desirable to have a technique which would also remove the less polar components prior to the more polar ones and thus provide partial purification as well as recovery from the biological or environmental matrix. In the process of isolating polar conjugates of drugs from urine, blood, and enzymatic incubations (all of which contain compounds with a broad range of solubilities), we devised such a procedure based on the use of the ODS cartridges. This system involves aqueous adsorption onto the ODSC followed by selective removal of compounds with a water immiscible organic solvent. An elution with ether, for example, is capable of removing those ether soluble compounds which have, in the past, been eliminated only by liquid-liquid extraction. In conventional ODSC procedures (elution with water-MeOH) these same compounds have coeluted in our glucuronide-containing methanol fraction. This technique therefore has the advantage of combining filtration, extraction, and adsorption into one step resulting in faster, less expensive, and in some cases a more efficient cleanup method. The method will be illustrated here by the 0 1982 American Chemical Soclety
