Anal. Chem. 1989. 67.
1996
I
A = BLANK
-
20
I -1.0
1
-0.9
I
I -0.7
-0.8
POTENTIAL
/
-0.6
-
I -0.5
-0.L
V fi Ag/AgCI
Figure 5. Figure to show the stripping voltammograms obtained for the blank and a series of lead calibration solutions.
three bubblers in series were used to collect it. In each bubbler there was 100 cm3 of acid. In subsequent determinations most of the lead was found in the first bubbler and none was found in the last bubbler. The nitric acid destroys the TML
+ HX (CH3)SPbX + HX (CH3)dPb
+
+ CHI (CH3)2PbX + CHk
(CH3)SPbX
----*
Under extreme conditions inorganic lead(I1) compounds are formed (14). It is essential that all the organolead be converted to inorganic lead, because anodic stripping voltammetry only detects inorganic lead. After several experiments it was found that all the TML could be converted to inorganic lead by placing it in a sealed digestion vessel, lined with Teflon (Berghol, West Germany) and heating it in an oven at 100 "C for 3 h. Once it was known that all the lead could be converted to inorganic lead, a sample of the diluted vapor standard could be analyzed. The standard atmosphere was generated with an air throughput of 730 cm3 min-'. After dilution, the sample flow rate through the collection solution was 500 cm3 min-'. A 0.5-cm3 portion of the sample that had been collected in acid and digested was diluted to 10 cm3. This was then determined and the concentration of lead in the acid was found
1996-1998
to be 6.75 ng ~ m - The ~ . sample had been collected for 5 min and the diffusion rate was 6.44 X lo-'' kg ss1. There was a 30 times dilution before collection of the sample in 100 cm3 nitric acid. This meant that theoretically there should have been 6.44 ng cm-3 lead present. The error in the result was therefore 4.8%. The ASV analysis confirmed that the vapor generator was operating successfully. The stability of the system was also shown to be acceptable with the error in the slope of the calibration curve being 1.5% over 31 days. CONCLUSIONS The apparatus described here is capable of providing very low vapor concentrations of toxic compounds. For this type of apparatus with a diffusion path length of 49.5 cm, a calibration range of approximately 4 mg m-3 to 10 pg m-3 TML would be possible at 50 "C. At 30 "C the range would be 2 mg mW3to 5 pg m-3 TML. The equipment ensures that the toxic liquid is safely contained but, rapid production of different vapor concentrations is still possible by gas stream mixing. In using this method the diffusion coefficient of the compound may also be determined without making an extra measurement. Although the system was designed for tetramethyllead, it can be easily adapted to other compounds by altering the appropriate parameters. R e g i s t r y No. Tetramethyllead, 75-74-1.
LITERATURE CITED Lucro, D. P. Calibration in Air Monitoring, 7 ; A.S.T.M.S.. Special Techniques Publication 568; ASTM: Philadelphia, PA, 1976; p 301. Nelson, G. 0. Controlled Test Atmospheres, Principles and Techniques; Ann Arbor Publishers: Ann Arbor, MI, 1971. O'Keefe, A. E.; Ortman, G. C. Anal. Chem. 1986, 3 8 , 760. Fortuin, J. M. Anal. Chim. Acta 1986, 75, 521. Methods for the Preparation of Gaseous Mixtures; BS4559; British Standard Institution: London, U.K., 1970. Selected Methods of Measuring Air Pollutants ; WHO Offset Publication No. 249; World Health Organisation: Geneva, Switzerland, 1976. Farmer, D.; Humphrey, J. Safe to Beathe; Klngwood Publications Ltd.: 1984; p 78. Lucero. D. P. Anal. Chem. 1971, 4 3 , 1744. Barratt, R. S.Analyst (London) 1981, 106, 187. Savitsky, A. C.; Sigga, S. Anal. Chem. 1972, 4 4 , 1712. Desty, D. N.; Geach, C. J.; Goldup, A. Gas Chromatography;Butterworth: London, 1960. Atshuiler, A. P.; Cohen, I. R. Anal. Chem. 1960, 3 2 , 802. Gilliland, E. R. Ind. Eng. Chem. 1934, 26, 681. Willemsens, L. C. Organolead Chemistry; International Zinc Research Organisation: New York, 1964.
RECEIVED for review November 30, 1988. Revised May 10, 1989. Accepted May 15, 1989. The authors wish to acknowledge the Occupational Medicine and Hygiene Laboratories of the Health and Safety Executive (Sheffield, U.K.) for supporting this research.
Preparation of Organic Matter for Stable Carbon Isotope Analysis by Sealed Tube Combustion: A Cautionary Note Michael H. Engel* and Rick J. Maynard
School of Geology and Geophysics, The Energy Center, 100 East Boyd Street, The University of Oklahoma, Norman, Oklahoma 73019 INTRODUCTION
The utilization of sealed tube combustion methods (1,2) continues to increase in popularity as a more time-efficient and less costly method for the conversion of organic carbon to carbon dioxide for stable carbon isotope analysis. Com-
* Author to whom correspondence should be addressed. 0003-2700/89/0361-1996$01.50/0
parisons of dynamic combustion and sealed tube combustion have demonstrated that, with appropriate precautions, both methods provide comparable results (1-3). In the course of preparing several samples of reference material NBS 22 for stable carbon isotope analyses using sealed tube combustion, it was observed that with increasing elapsed time subsequent to combustion, the 613C values of 0 1989 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 61, NO. 17, SEPTEMBER these samples became up t o 3% lighter (i.e., depleted in carbon-13). In this paper, we report the results of a series of experiments that document this observed carbon isotope depletion effect and suggest a strategy for sample preparation t o avoid this problem.
EXPERIMENTAL SECTION One hundred eleven samples of NBS 22 (-3 WLeach) were placed in individual Pyrex tubes (22 cm X 7 mm id.) that contained 5.5 g of copper oxide wire. The tubes and copper oxide were preheated (to 550 and 850 O C , respectively) and then cooled to room temperature prior to loading the samples. Next, the tubes were evacuated, sealed, and combusted at 550 "C for 2.5 h. Subsequent to combustion, the tubes were stored at room temperature. A t various times, ranging from 1 to 274 days after combustion (Table I), several tubes were opened with a tube cracker ( 4 ) and the COz gas was cryogenically isolated from the individual tubes for stable carbon isotope analysis as previously described (5). The sampling ampules containing the COzsamples were transferred directly to the inlet system of a Finnigan Delta E mass spectrometer for stable carbon isotope analysis. On several occasions some of the sealed tubes were reheated for 2.5 h at 550 "C just prior to stable carbon isotope analysis. It has been speculated that, subsequent to combustion, copper carbonate may precipitate as a surface film via prolonged contact of COz and HzO with the copper oxide wire in a sealed tube. Additional experimentswere conducted to verify this phenomenon and to evaluate its possible effect on the stable carbon isotope composition of COP After analysis of the COz gas isolated from the 259 and 274 day old tubes (Table I), the individual copper oxide samples isolated from these tubes were reheated to 550 "C. The resultant COz samples were isolated and analyzed for their respective stable carbon isotope compositions as described above. Also, three additional tubes containing NBS 22 and copper oxide wire were combusted and, after being kept at room temperature for 112 days, were pinched in half. The bottom halves, containing gas and copper oxide, and the top halves, containing only gas, were reheated (550 "C) prior to the stable isotope analysis of their respective COz components. Assuming that the observed carbon isotope depletion effect for COz is directly associated with the reaction of COz and HzO with the copper oxide wire, an attempt was made to evaluate whether this apparent COz fractionation could be diminished by increasing the volume of the combustion tube. Aliquots of NBS 22 and copper oxide (approximately identical with the amounts used in the above experiments) were placed in 34 cm X 0.7 mm i.d. Pyrex tubes and combusted as described above. Two of these samples were analyzed for their stable carbon isotope compositions immediately subsequent to combustion. After storage at room temperature for 96 days, the COz gas in 11 of the tubes was analyzed for stable carbon isotope compositions. The copper oxide wire in the bottom half of each tube was resealed while the tube was still under vacuum in the tube cracker and then reheated for 2.5 h at 550 O C . The resultant COz was isolated and analyzed as above. Three of the long tubes were reheated just prior to stable isotope analysis of the COz gas. RESULTS AND DISCUSSION Stable carbon isotope values for samples of NBS 22 that were combusted and then stored for 1-274 days prior t o analysis are compiled in Table I. The average 613C value for samples that were stored for up to 22 days prior to analysis was -29.80 f 0.03%. However, the average 613C value for samples that were stored for 37-274 days prior to analysis was -30.83 f 1.0%. The combusted NBS 22 samples that were stored for 274 days prior t o analysis were depleted by more than 3 L relative to average 613Cvalues for NBS 22 that have been reported in the literature, Le., -29.81 f O.06%0 (3) and -29.73 f 0.09'7~(6). Statistical analysis revealed that a linear relationship exists between the stable carbon isotope values of the NBS 22 samples and storage time subsequent to combustion (Pearson's correlation coefficient = 0.98). I t is important to note, however, that the average 613C value for samples that had been stored for up to 259 days, but were
1, 1989
1997
Table I. Stable Carbon Isotope Values for NBS 22 Samples time elapsed after initial combustn, days 1 2 3 4 5 6 7 8 10 12 13 21 22 31 34 37 39 40 41 46 55 61 68 69 88 89 259 274
813C,L,for samples reheated just prior to anal.
813C, %o -29.82 -29.85 -29.81 -29.82 -29.81 -29.76 -29.80 -29.79 -29.85 -29.77 -29.77 -29.80 -29.83 -30.08
-
-30.06 -29.92 -30.04 -29.85 -29.99 -30.36 -30.30 -30.83 -30.87 -31.08
-
f 0.04 f 0.01 f 0.03 f 0.02 f 0.05 (1) 0.03 f 0.03 (1) f 0.03 f 0.01 f 0.04 f 0.04 f 0.13
*
(13)" (3) (13) (6) (3) (5) (11) (3) (2) (5) (2) (4)
-29.82 (1)
f 0.15 f 0.01 f 0.01 f 0.01 (1) f 0.02 (1) f 0.15 0.01
*
-
(2) (2) (3) (2) (3)
-29.83 -29.82 -29.81 -29.85 -29.82
(3) (3)
(1)
-33.27 (1) -33.21 f 0.42
-
(5)
f 0.01 (2) f 0.01 (4) (1) f 0.01 (2) f 0.01 (2)
"The number in parentheses indicates the number of samples that were analyzed. -, sample not analyzed. Table 11. Stable Carbon Isotope Values for NBS 22 Samples time
combustn tube
length, cm 22 22 34 34 34
elapsed after initial combustn, days 259 274
1 96 96
613C,
L
reheated CuO COz gas" -33.27 -33.21 -29.84 -29.89 -29.86
(1)
f 0.42 (5) f 0.00 (2) f 0.05 (11) f 0.02d (3)'
wireb -25.98 (1) -26.15 f 0.21 (4) -e
-24.16 f 0.42 (7)
-
a 613C value for COz gas present in tube. 813C value for COz gas that was recovered by reheating the CuO wire. e -, sample not analyzed. d Sample reheated just prior to stable carbon isotope analysis. 'The number in parentheses indicates the number of samples that were analyzed.
reheated to 550 "C just prior to analysis, was -29.83 f 0.01'3~ (Table I). Subsequent to the analysis of the NBS 22 gas samples that were stored for 259 and 274 days, the copper oxide wire that remained in the tubes was resealed under vacuum and heated to 550 "C. Whereas the original COz components isolated from these tubes were depleted in carbon-13 relative to NBS 22 samples that were stored for shorter time intervals (Table I), the COz samples recovered by reheating the copper oxide wire were enriched by several parts per mil (Table 11). Furthermore, stable carbon isotope analyses of the COz recovered subsequent to the reheating of tubes that were pinched in half revealed that the gas in the top portions of the tubes was depleted in carbon-13 relative to NBS 22 (-31.69 f 0.41%) and the gas in the bottom portions of the tubes that remained in contact with the copper oxide was enriched in carbon-13 relative to NBS 22 (-28.66 & 0.04%0). Assuming that the
1998
Anal. Chem. 1989, 67, 1998-2000
observed stable carbon isotope depletion for NBS 22 that occurs with prolonged storage is a consequence of the interaction of the gaseous products of combustion with the copper oxide wire, it is not surprising that an increase in the size (34 cm) of the combustion tubes diminished the isotope depletion effect that was observed for the shorter (22 cm) combustion tubes (Table 11). The synthesis of copper carbonate minerals such as malachite results in an isotope fractionation such that the carbonate carbon becomes enriched in carbon-13 relative to the COz starting material (7). Also, the fact that copper carbonate minerals such as malachite and azurite decompose at 200 and 220 "C, respectively, well below the combustion temperature employed in this study (i.e., 550 "C), lends credence to the hypothesis that (1) the formation of a copper carbonate phase is responsible for the observed isotope depletion of C02 with prolonged storage of the NBS 22 samples subsequent to combustion and (2) reheating of older NBS 22 samples prior to stable isotope analysis results in the destruction of this copper carbonate mineral phase and the release of this I3Cenriched component back into the COz reservoir. X-ray diffraction analysis of the preheated (850 "C) copper oxide wire indicated that the dominant mineral present was, as expected, tenorite (CuO). X-ray analysis of the wire subsequent to the combustion and storage of NBS 22 samples was apparently not sensitive enough to detect the copper carbonate phase. It was observed, however, that a significant percentage of the tenorite had been converted to cuprite (Cu,O), which is a t an appropriate oxidation state for the formation of copper carbonate (Cu2C03). In summary, the sealed tube combustion method is rapidly becoming the method of choice for the conversion of organic
matter to C02 for stable carbon isotope analysis. Unlike those who use dynamic combustion, scientists who employ static (sealed tube) combustion are typically combusting 30 or more individual samples at a time. As indicated above, the storage of samples for prolonged time intervals (Le., several weeks) subsequent to combustion will result in 613C values that are depleted by up to 1-3%0 relative to what would have been observed had the samples been analyzed within several days subsequent to combustion. For situations that arise (such as instrument down time) when combusted samples must be stored prior to analysis, it is possible to avoid this problem by isolating the COz gas from the copper oxide wire. A more practical solution, however, is to simply reheat the tubes immediately prior to analysis.
ACKNOWLEDGMENT We thank D. Powell (OU) for performing the X-ray diffraction analyses and Z. Sofer (Amoco Production Co.) for his critical review of this manuscript.
LITERATURE CITED Stuermer, D. H.; Peters, K. E.; Kaplan, I. R. Geochim. Cosmchim.
Acta 1978, 4 2 , 989-997. Sofer, 2. Anal. Chem. 1980, 5 2 , 1389-1391. Schoell, M.; Faber, E.; Coleman, M. L. Org. Geochem. 1983, 5 , 3-6. DesMarais, D. J.; Hayes, J. M. Anal. Chem. 1976, 4 8 , 1651-1652. ( 5 ) Bonilla, J. V.; Engel, M. H. Org. Geochem. 1986, 10, 181-190. (6) Hut,G. Report to Director General, International Atomic Energy Agen-
cy: Vienna, 1987, 42 pp.
(7) Smith, A. W. Archammetry 1978, 20, 123-133.
RECEIVED for review January 24,1989. Accepted June 1,1989. We acknowledge the National Science Foundation, Division of Earth Sciences (Grant No. EAR-8352055) for partial support of this research.
Foam Countercurrent Chromatography of Bacitracin with Nitrogen and Additive-Free Water Hisao Oka,' Ken-ichi Harada? Makoto Suzuki? Hiroyuki N a k a ~ a w aand , ~ Yoichiro Ito* Laboratory of Technical Development, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland 20892 Foam separation has long been used for separation of a broad spectrum of samples ranging from metal ions to mineral particles (1). The separation is based on a unique parameter of foaming capacity and/or foam affinity of samples in an aqueous solution and bears a great potential for application to biological samples. Utility of the method in research laboratories, however, has been extremely limited mainly due to a lack of efficient instruments to fully exploit versatility of the method. Recently, an innovation of the foam separation technology has been achieved by development of foam countercurrent chromatography (CCC), which utilizes a true countercurrent movement between foam and its mother liquid through a long, narrow coiled tube by the aid of a particular mode of planetary motion generated by a coil planet centrifuge (2). Introduction of a sample mixture into the coil results in separation of the Visiting Scientist from the Aichi Prefectural Institute of Public Health Nagoya, Ja an. Permanent adgesa: Faculty of Pharmacy, Meijo University, Tempaku, Nagoya 468, Japan. 3Permanentaddress: The Institute of Public Health, Tokyo 108, Japan. 0003-2700/89/0361-1998$01.50/0
sample components: Foam-active materials quickly move with the foaming stream and are collected through the foam outlet while the rest are carried with the liquid stream in the liquid outlet. The method has been applied to various test samples including ionic compounds collected with suitable surfactants and surface-active proteins separated in a phosphate buffer solution to prevent denaturation of the molecule (2-4). This paper describes a successful application of the foam CCC technology to chromatographic fractionation of bacitracin complex using nitrogen gas and distilled water entirely free of surfactant or other additives. Bacitracin (BC) is a basic, cyclic peptide antibiotic commonly used as a feed additive of livestock all over the world. It consists of more than 20 components, each with different antimicrobial activities, and chemical structures of these components other than BCs-A and -F are still unknown. Because of its strong foaming activity, BC is an ideal test sample to demonstrate capability and usefulness of the present method.
EXPERIMENTAL SECTION The apparatus used for the present study is a multilayer coil planet centrifuge,which produces a synchronous planetary motion of the gear-driven column holder and has been described pre0 1989 American Chemical Society