832
Anal. Chem. 1982,5 4 , 832-833
complexes with cyanide and thus interfere with the new procedure for cyanide. The results in Table I11 indicate no interference from common anions such as chloride, nitrate, sulfate, perchlorate, chromate, and small amounts of sulfite. Thiosulfate and thiocyanate react readily with iodine and cause serious interference. An actual chromatogram of a sample containing both cyanide and chromate is shown in Figure 2. The initial positive peak is called the “pseudopeak” and is always encountered in single-column ion chromatography because of differences in the conductances of the sample and the eluent. The negative peak immediately following the pseudopeak is caused by the buffer, sodium acetate. The next positive peak is iodide; the final positive peak is chromate. The sensitivity of the method is improved by reacting cyanide to produce iodide, because the chromatographic detection of iodide is several times more sensitive than it would be for the cyanide ion directly. LITERATURE CITED
and 20
(1) Chambers, W. E.; et al. “Treatise on Analytical Chemistry Part 11: Analytical Chemlstry of Inorganic and Organic Compounds”; Kolthoff. I. M., Elving, P. J., Eds.; Wiley: New York, 1978; Vol. 10. (2) Beran, P.; Bruckensteln, S. Anal. Chern. 1880, 52. 1183-11813. (3) Gjerde, D. T.; Fritz, J. S.; Schmuckler, G. J . Chrornatogr. 1878, 786, 509-519 (4) Jarosz, M.; Fritz, J. S., unpublished work. (5) “Handbook for Analytical Quallty Control in Water and Wastewater Laboratorles”; Analytlcal Quallty Control Laboratory, National Environmental Research Center: Clnclnnatl, OH, 1972.
Interference Study. The effects of various metal ions and foreign anions on the results for cyanide were studied. The results for metal ions reported in Table I1 show essentially no interference from nickel(II), magnesium(II), cadmium(II), or zinc(I1). Mercury(I1) and cobalt(I1) form more gtable
RECEIVED for review November 16,1981. Accepted January 18,1982. Operated for the US. Department of Energy by Iowa State University under Contract No. W-7405-Eng-82. This research was supported by the Director of Energy Research, Office of Basic Energy Sciences.
I
I
I
I
0
3
6
9
Minutes
Figure 2. Chromatogram for a sample containing 4 ppm
CN-
ppm CrOd2-.
Integration of a Gas Flow Proportional Counter into a Commercial Gas Chromatagraph/Mass Spectrometer/Data System D. L. Doerfler, G. T. Emmons, and I . M. Campbell” Department of Biological Sciences, University of Pittsburgh, Parran Hall, 130 DeSoto Street, Pittsburgh, Pennsylvania 1526 1
Coupling a gas flow proportional conter (PC) to a commercial gas chromatograph/mass spectrometer (GC/MS) is a straightforward undertaking (1-3). Development of software to collect and process data from the radiogas chromatograph/mass spectrometer so formed is also straightforward (4,5).Integration of a data channel derived from a PC into a data system that was provided with a commercial GC/MS respresents a more formidable challenge. This report describes how this challenge can be met. EXPERIMENTAL SECTION Hardware. A Packard Model 894 gas flow PC and an Hew1etbPackard (HP) Model 5985B GC/quadrupole-MS/data system were used. The former converts a portion of the effluent of a GC column to COzand H20 (copper oxide furnace), reduces the H20 to H2 (iron chip furnace, used if tritium is involved) or removes it from the gas flow (magnesium perchlorate or sodium sulfate), and countv the processed gases in a proportional tube following addition of a quench gas (e.g., propane). The computer of the HP 5985B data system is an HP 21MX-E, configured with 32k words of core, all standard features plus a fast Fortran processor and four unused 1 / 0 slots. Plumbing the PC into the GC/MS. The gas flow in the GC module of the GC/MS was broken at the effluent end of the 0003-2700/82/0354-0832$0 1.25/O
column and a fiied ratio splitter was inserted (Packard Instrument Co., part 5060723,3:1). The splitter resides in an aluminum block of the same dimensions as the flame detector housing and is positioned where that housing normally resides. The temperature sensing and heating circuitry provided for the flame detector is used to heat the splitter. The majority of the gas flow (75%) is fed to the copper oxide furnace of the PC by means of a heated 1/16 in. nickel line while the minority (25%) is led to the molecule separator of the MS. The drying tube (we have worked so far exclusively with 14C)and PC counting tube/shield assembly sits atop the GC. The PC electronics have been repackaged to fit on the GC/MS main frame. Constructionof a Pulse Counter/Computer Interface. For each nuclear disintegration that the Model 894 PC detects, a 5 V/3 ps pulse can be picked up at its scaler terminal. In our interface this pulse is narrowed in width to 600 ns (with a 74C221 CMOS IC) and is fed to a 16-bit synchronous pulse counter (four 74C161 CMOS IC’s) that increments on a positive edge. The maximum permissible count rate of the pulse counter is in excess of 1MHz, a figure that comfortably encompasses the maximum count rate of the PC (20 kHz). The running sum of all counts collected in an RGC/MS run (maximum range, 0-177 7778,065 53510)is stored in the interface; the current value is displayed on a 6-digit LED display on the interface cabinet. The interface communicates with the HP MX21-E by means of 16-bit parallel 0 1982 American Chemical Society
$83
Anal. Chem. 1982, 54, 833-835 .....-,..-
Figure 1. Traces produced from an RGC/MS run of the "ene-ol" standard (top to bottom): Integral of radioactivity; derlvatlve of radioactivity; m / z 69 selected Ion current profile; total ion current proflle.
1 / 0 board (HP part 12566B-002,inserted in 1/0slot 168of the 21MX-E) that, on computer command, transmits the current reading of the running sum of counts to the computer as a 16-bit word. The running sum can be set to zero under program control. To avoid the possiblity of attempting to read the running s u m while it is being augmented, the pulse counter is stopped briefly (1.5 MS) prior to being read by the computer. Although it is conceivable that some pulses will be lost during this pause, such losses are not likely to be great. At low count rates, loss likelihood will be small whille at high count rates the losses will represent only a small proportion of the total counts. Modificationci to tho HP Software. The program AQUIRE is the heart of the HP data acquisition and processing software of the Model 5985B. It scans the dc voltage to the quadrupole rods, reads the ion multiplier ouput, and establishes for each scan a file of m/z values against the logarithm of the appropriate ion intensity (a parameter derived directly from the ion multiplier hardware). These log ion intensities are subsequently converted to absolute form. We have modified AQUIRE such that (a) at the beginning of each RGC/MS run, the running sum in the counter
interface is set to zero, (b) at the beginning of each MS scan in that run, the running sum is read into the m/z:intensity file at a position corresponding to the intensity of m/z = 4.0, (c) the difference between the current value of the running sum and that from the previous scan is read into the m/z:intensity file at a position corresponding to the intensity of m/z = 5.0, and (d) the process of intensity linearization, that is one of the last operations executed by AQUIRE, is restricted to intensities above m/z = 5.0. The modified version of AQUIRE and the schematic for the counter/interface are available from the authors. RESULTS AND DISCUSSION Figure 1shows a set of traces derived from a single run of our "ene-ol" standard (2). This standard contains approximately 1pg/pL of the n-alk-1-enes of carbon number 12, 14, 15, 17, 18, and 19 and the n-alk-1-01s of carbon number 12 through 19; the CI4,Ci6, and Cia alcohols are labeled with 14C (ca. 440 dpm/pL). The run was made on a 6 f t X 4 mm glass column containing 3"10 OV-17 on 80-100 mesh support (temperature program: 150 "C to 300 "C at 10 "C/min); the mass spectrometer was set to scan from 50 to 300 m / z every 3 s. The panels of the figure contain respectively the time courses of (top to bottom): the counter's running sum (called as the m/z 4.0 ion profile); the derivative of that running sum (called as the m / z 5.0 ion profile); the m / z 69 icn profile; and the total ion profile. As shown, standard unmodified HP software (ANSWER) can be used to integrate peaks in total ion, selected ion, and radioactivity profiles. The radioactivity readings obtained by integration are relative; absolute values are obtained by multiplication by the factor:gas flow rate through the counting tube/tube volume. Clearly the RGC/MS unit we have constructed is capable of providing simultaneously information on the amount (integration of total or selected ion current profiles), radioisotopic content, and identity (GC Rt value and mass spectrum) of those solutes in a mixture that are volatile in the GC senne or can be rendered 510. The importance of this capability in biochemistry has already been noted (6) and will be developed further elsewhere. LITERATURE CITED (1) Hobbs, D. C. Am. Soc. Mass Spectrom.. Annual Meeting, Abstract 851, 1970. (2) Nulton, C. P.; Nawaral, J. D.; Campbell, I. M.; Grotzinger, E. W. Anal. Biochem. 1976, 75. 219. (3) Braun, W. H.; Madrid, E. 0.;Karbowskl, R. J. Anal. Chem. 1976, 48, 2284. (4) Campbell, I . M.; Doerfler, D. L.; Donahey, S. A.; Kadlec, R.; McGandy, E. L.; Naworal, J. P.; Nulton, C. P.; Venza-Raczka, M.;Wimberly, F. Anal. Chem. 1077, 49, 1726. (5) Doerfler, D. L.; Rosenblum, E. R.;Malloy, J. M.;Naworal, J. D.; McManus, 1. R.; Campbell, I. M. Biomed. Mass Spectrom. 1980, 7 , 259. (6) Campbell, I. M. Anal. Chem. 1979, 57, 1012 A.
RECEIVED for review November 18, 1981. Accepted January 11,1982. This work was supported in part by the U.S. Public Health Service, NIH GM 25592.
Pressure Activated Shutdown for Liquid Chromatography Systems Peter Y. T. Lln and C. LeRoy Blank" Department of Cht~mistry,Unlverslty of Oklahoma, 620 Parrlngton Oval, Norman, Oklahoma 7301 19
Clogged frits, blocked tubing, and other malfunctioning components in liiquid chromatographs can lead to excessive and, sometimes, damaging high pressures. Recent advances in liquid chromatography have caused such undesirable events to become even inore important to practitioners in the field. The pulse dampener recently introduced by Nikelly and
Ventura (I), e.g., is now finding widespread utilization in the removal of pump pulsation and lowering of detection limits. But, this device loses a considerable fraction of its damping capability if exposed to excessive pressures (currently >3000 psi). The higher resolution afforded by columns packed with particles having very small diameters has also been advan-
0003-2700/82/0354-0833$01.25/00 1982 American Chemical Society
