Anal. Chem. 1980, 52,371-373
dertaken without t h e aid of high level languages.
LITERATURE CITED (1) R. E. Dessy, P. J. Van Vuuren, and J. A. Titus, Anal. Chem., 46, 917
(1974). (2) R. E. Dessy, J. A. Titus, and P. J. Van Vuuren, Anal. Chem., 46 1055A
37 1
(5) "Computers in Chemistry and Instrumentation", J. S . Matheson, H. B. Mark. Jr., and H. C. MacDonaM. Eds., Marcel Dekker, New York, 1972. (6) Princeton Applied Research Corporation, Princeton, N.J. (7) W. R. Matson, E. Zink, and R. Vitukevitch, Am. Lab., 10, 59 (1977). (8) K. F. Drake, R. P. Van Duyne, and A. M. Bond, J . Electroanal. Chem., 89 231 (1978).
1197Al
(3) R.-E.'Dessy, Science, 192, 5 1 1 (1976). (4) W. G. Sherwood, D. F. Untereker. and S . Bruckenstein, Anal. Chem., 47,84 (1975)and references cited therein.
RE-cE1VED for
review March 26, 1979* Accepted October 26,
1979.
Portable Fluorometric Monitor for Detection of Surface Contamination by Polynuclear Aromatic Compounds Daniel D. Schuresko Chemical Technology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830
Uptake of polynuclear aromatic hydrocarbon (PNA) compounds via contact with contaminated surfaces in t h e workplace has been identified as a major mode of worker exposure ( I ) . T h e need for instrumentation t o monitor P N A buildup on surfaces has also been recognized ( I , 2). Such a n instrument must be capable of (a) functioning during actual plant operation in varied working environments; (b) detecting material spilled on different work surfaces, including machinery, plumbing, construction materials, and on personnel and clothing; and (c) being easily and reliably operated by all plant personnel. T h e fluorescence spotter described in this paper ( U S . P a t e n t applied for) has these capabilities. This instrument, which consists of a hand-held optics unit connected via a n umbilical cable to a n electronics console, enables remote monitoring of work area surfaces at distances 53 m. A portable, rechargable, battery-powered electronics module, which will replace t h e electronics console, is also being developed. It is anticipated that portable fluorescence spotters will be used extensively t o detect surface contamination in changing areas, lunch rooms, control rooms, and other "clean" areas in coal conversion facilities, and to monitor skin contamination of coal plant workers. T h e currently used method of detecting spills is to turn off the ambient lighting and scan the suspected area with a black light. I n contrast, t h e newly developed spotter (a) can be operated outdoors in direct sunlight or indoors in t h e presence of strong background illumination; (b) can provide a quantitative measure of t h e amount of fluorescent material; (c) will discriminate between t h e fluorescence of organic materials and some inorganic compounds based on their fluorescence lifetimes; and (d) does not present a vision hazard t o personnel.
EXPERIMENTAL The spotter induces and detects the fluorescence of PNAs that characteristically absorb light in the 350- to 440-nm region of the spectrum and fluoresce with high efficiency in the blue-green region of the spectrum. Multiring heteroatom aromatic compounds, including acridines, are also detected, although they generally absorb light and fluoresce at longer wavelengths than do their pure hydrocarbon counterparts. The optics unit, shown in Figure 1, consists of two subunits: (a) a UV illuminator, which produces a beam of amplitudemodulated light; and (b) a fluorescence detector, which detects 380- to 600-nm fluorescence induced by the UV light. The operation of the spotter is depicted in Figure 2, as are the signal traces corresponding to the radiated UV intensity (trace A) and 0003-2700/80/0352-037 1$0 1 .OO/O
to the fluorescence (trace B) that is produced when a spill has been "sighted". The radiated UV beam is amplitude-modulated at 1kHz by a mechanical chopper; thus the induced fluorescence is also modulated at 1 kHz. This 1-kHz fluorescence signal is superimposed on a much stronger signal from reflected background illumination of the viewed surface. The electronic frequency spectrum of ac-powered fluorescent or incandescent room lighting consists of a 120-Hz frequency component and its harmonics superimposed on white noise (dc-powered lamps and sunlight produce only white-noise background); hence, it is possible to separate the fluorescence signal from the background signal by electronic filtering (trace C). Demodulating and low-pass filtering of this signal (trace D) effectively averages each 'iz-ms pulse and allows detection of fluorescence only 3'% as intense as the background illumination in the optical wavelength band of interest. A schematic diagram of the prototype spotter optics unit is shown in Figure 3. The illuminating beam, produced by a high-pressure mercury arc lamp, is modulated by an electromagnetic, tuning-fork chopper. A dichromatic beam splitter reflects the illuminating beam to the telephoto output lens. The longer wavelength fluorescence from the sample is collected by the telephoto lens, passed through the dichromatic splitter and emission filters, and detected by the photomultiplier tube. The 4% beam splitter and photodiode assembly provides a reference signal that is used to offset the background fluorescence of the dichromatic splitter and to correct fluorescence measurements for lamp output drift. Optical filters in both units are mounted in interchangeable dovetailed filter slides, allowing for filter selection appropriate to the class of compounds being monitored while the unit is in use. Using a simple resistor network, a photomultiplier signal component, which is caused by the fluorescence of the dichromatic beam-splitter and telephoto lens, is cancelled by subtracting a preset fraction of the photodiode signal from the photomultiplier signal. The difference signal is then fed into an Ithaco Model Dynatrac 3 lock-in amplifier (Ithaca N.Y.) which filters and demodulates at 1 kHz. The lock-in reference signal is provided by the chopper oscillator circuit. The lock-in output drives a voltage-to-frequency converter coupled to a small speaker, which provides an audio signal whose frequency increases with the measured fluorescence intensity. In addition, the lock-in output is measured on a front-panel meter.
RESULTS AND DISCUSSION T h e spotter has been laboratory tested with pure compounds and with several coal and oil shale products and wastes. When operated in a laboratory area illuminated by fluorescent lighting, the spotter can typically detect 0.2 pg of perylene (in dilute solution in cyclohexane); in darkened areas, it is sensitive to 0.001 pg of perylene Considering that 10-pg
G 1980 American Chemical Society
372
ANALYTICAL CHEMISTRY, VOL. 52, NO. 2, FEBRUARY 1980 .
Table I . Specific Fluorescence of Coal Liquefaction and Oil Shale Products
sample
specific fluorescence: unitslmg
COED syncrude hydrotreated coal distillate product distillate centrifuged shale oil O R N L hydrocarbonization oil (HC-12) SRC-I1 fuel oil blend SRC-I recycle solvent ( r a w ) SRC-I wash solvent SRC-I light organic liquid (raw) SRC-I process solvent SRC-I light oil
FLUORESCENCE RECEIVER
4.6 0.9 0.1 3.1 b
Table 11. Specific Fluorescence of Coal Liquefaction Wastes
sample
uv
-
nn-
7.7
a One unit of fluorescence is defined as the equivalent fluorescence of 1 yg of perylene in dilute solution in cyclohexane. Not detectable.
Figure 1. Fluorescence spotter optics unit
I
6.3 3.4 11.3 4.8 0.2
INTENSITY
A
DETECTOR SIGNAL
B
SRC-I bio-unit feed SRC-I bio-unit effluent SRC-I bio-unit sludge SRC-I plant effluent SRC-I surge tank sludge (diluted 400:lin cyclohexane) SRC-I surge reservoir water SRC-I graver effluent
specific fluorescence: unitsimg
0.7x 10-3 0.4 x lo-’ 1.9x 10-3 b 12.9 x 10-3
0.7x 1 0 - 3 0.6 x 10-3
a One unit of fluorescence is defined as the equivalent fluorescence of 1 y g of perylene in dilute solution in cvclohexane. Not detectable.
t
1
PROTOTYPE FLUORESCENCE MONITOR Figure 2. Principle of operation of spotter ;EMISSION F1-TFP PMT,
2 CHROMATIC BEAW
STOP
Figure 3. Schematic diagram of spotter
doses of several PNAs are cytotoxic and will cause induction of the mixed function oxidase enzyme system in mammalian cells ( 3 , 4 ) ,it is evident that the high sensitivity of the spotter is appropriate to its intended task. Tables I and I1 list the spotter-measured specific fluorescence of several synthetic fuel products and wastes. The oil samples were prepared by placing 1- to 10-mg droplets of each oil onto microscope slides and covering each with a cover slip. This technique yielded 1-to 20-pm-thick films of known area and thickness whose optical absorbance at 365 nm, the principal excitation wavelength, was less than 0.3-0.d. units. T h e undiluted waste samples were placed into 3.6-cm3vials for measurement, with the exception of the surge-tank sludge
which was diluted 400:l in cyclohexane. The variation in the specific fluorescence of the SRC-I recycle solvent, wash solvent, and light oil (Table I) parallels preliminary determinations of their 1,2-benzopyrene (BP) content (5). This is noteworthy because consideration is being given t o t h e use of B P as the proxy compound for PNA determination in coal and oil-shale derived liquids (6). Milligram-level films of Oak Ridge National Laboratory (ORNL) hydrocarbonization oil and COED syncrude on both floor tile and metallic surfaces have also been examined in a less quantitative fashion. The film fluorescence was at least ten times stronger than the substrate fluorescence, which made the “spots” readily distinguishable. Preliminary field testing work with t h e spotter, carried out in t h e ORNL hydrocarbonization facility, detected milligram-level coal tar contamination on the plumbing and floor. Coal liquefaction products may contain certain organic compounds, heavy-metal inorganic materials, and particulate matter, all of which quench the fluorescence of PNAs and thus interfere with fluorometric monitoring techniques. Two such classes of aromatic compounds abundant in coal conversion products are simple phenols and acridines. We have tested for phenolic interference by adding large excesses of phenol to solutions containing perylene. Phenol did not interfere with our fluorometric P N A determination. However, heteroatom aromatic compounds such as acridines, which are abundant in oils having a high asphaltene content and whose absorption maxima overlap the emission maxima for PNAs, may partially quench P N A fluorescence. This quenching may partially explain the decreased sensitivity of the spotter to ORNL hydrocarbonization oil and to SRC-I light organic liquid (Table I). Both materials undergo a large increase in specific
Anal. Chem. 1980, 52, 373-375
fluorescence when dissolved in cyclohexane. T h e spotter is capable of discriminating between classes of organic contaminants, such as PNAs and acridine dyes, by selecting optimal excitation and emission wavelengths with optical filters. T h e spotter can also discriminate the fluorescence of inorganic materials such as uranyl nitrate from the fluorescence of organic compounds, although both materials have similar optical properties due to the long luminescence lifetime of uranyl nitrate. The delayed fluorescence of uranyl nitrate results in a nonzero phase shift in its fluorescence signal that can be measured electronically with the lock-in amplifier. The spotter is sensitive to 1-mg quantities of uranyl nitrate and could conceivably be used to locate spills or leakage in nuclear-fuel and waste-handling facilities. The phase-shift measurement capability of the spotter should also be useful for discriminating P N A fluorescence from background fluorescence caused by inorganic pigments commonly used in paints. This fluorescence spotter is suitable for the work-area surface monitoring required by the coal conversion and oil shale industries. Its attributes of high sensitivity, operability in various environments, quantitativeness, and remote detection capability make it an important component of the environmental and industrial hygiene monitoring hardware.
373
Actual field testing of the spotter in selected coal conversion facilities is planned, and transfer of this technology to potential commercial manufacturers will be pursued a t the conclusion of the field testing.
LITERATURE CITED (1) Report of the Wwking Group on Assessing Industrial Hygiene Monitoring Needs for the Coal Conversion and Oil Shale Industries, Otto White, Chairman, Brookhaven National Laboratory, August 1978. (2) Schuresko, D. D., Jones, G. "A Portable Fluorometric Monitor to Detect PNA Contamination of Work Area Surfaces", presented at the Symposium on Assessing the Industrial Hygiene Monitoring Needs of the Coal Conversion and Oil Shale Industries, Brookhaven National Laboratory, November 1978; Proceedings, BNL 51002, July 1979. (3) Schuresko, D. D., et al. unpublished results. (4) Nebert, D. W., Gelboin, H. V . J . Biol. Chern. 243(23), 6242 (1968). (5) Kubota, H., et al. Analytical Chemistry Division, Oak Ridge National Laboratory, private communication. (6) Gammage. R. B. "Preliminary Thoughts on Proxy PNA Compounds in the Vapor and Solid Phase", presented at the Symposium on Assessing the Industral Hygiene Monitoring Needs of the Coal Conversion and Oil Shale Industries, Brookhaven National Laboratory, November 1978; Proceedings, BNL 51002, July 1979.
RECEIVED for review July 11,1979. Accepted October 16,1979. Research sponsored by the Office of Health and Environmental Research and the Office of Fossil Energy, U S . Department of Energy under contract W-7405-eng-26 with the Union Carbide Corporation.
I n Situ Synthesis of Aromatic Iodine Compounds for Gas Chromatographic Retention Measurements Veluppillai Paramasigamani and Walter A. Aue * Chemistry Department, Dalhousie University, Halifax, Nova Scotia, Canada
The measurement of retention in gas-liquid chromatography is carried out for many reasons: qualitative analysis of complex mixtures, characterization of liquid phases, physicochemical studies of intermolecular forces, determination of AG values for the Martin equation, etc. For such studies, especially those of a systematic nature involving larger numbers of solutes, it is often difficult or very expensive to procure all of the necessary standards. While synthesis by conventional methods is usually possible, the time involved is prohibitive. It would therefore be convenient, wherever possible, to use in situ synthesis, Le., to generate the desired compound in the gas chromatograph from a suitable, chromatographically separated, hence chromatographically pure, precursor. We have made use of this approach in a recent study that required a larger number of iodinated aromatics. Aromatic iodine compounds, especially when they involve various other substituents and particular isomer configurations, are often not available commerically. On the other hand, one can frequently purchase compounds in which the suggested position of iodine is occupied by chlorine, bromine, nitro, or other reactive groups. In most of these cases, a straightforward substituent exchange would be difficult to achieve and time-consuming by conventional organic synthesis. Yet, in an intercolumnar, iodine-doped argon plasma, the desired conversion can take place in the time it takes a gas chromatography (GC) peak to flow through. The plasma is contained in a Penning ionization reactor, Le., essentially a simple, leak-tight argon detector. This reactor and the necessary two-column gas chromatographic system have been described ( I ) . For introduction of 0003-2700/80/0352-0373$0 1.OO/O
iodine, a small side-flow of argon was doped a t ambient temperature by passage through a tube filled with iodine crystals, then added to the first GC column effluent just before the Penning reactor. To establish the validity of this approach, a variety of compounds were processed for which authentic iodine analogues were at hand. Iodine compounds are not the only ones formed, but the relative simplicity and predictability of product patterns usually allows a reasonable correlation. I t goes without saying, of course, t h a t any identification based solely on retention data is subject to the well-known limitations of gas chromatography in this regard. The prediction of the product pattern is, in most cases, straightforward. A simple example may serve to illustrate this: the degradation of 1,3,5-trichlorobenzene under nondoped and doped conditions. Without iodine, 1,3-dichlorobenzene is formed as the primary and chlorobenzene as the secondary product (Figure 1). Under doped conditions, iodine competes with hydrogen for the positions &here chlorine has left, yielding 3,5-dichloroiodobenzene as the major and 1,3-dichlorobenzene as the minor primary product (Figure 2). The isomeric configuration is preserved, Le., iodine does not attack other positions; at least to the extent that no major peak was ever found to originate from such an attack on a larger number of investigated substrates. This parallels the experience obtained earlier with a similar but nondoped system used for isomeric structure determination (2). A variety of substituents can be exchanged against iodine, the easiest being -C1, -Br, -NOz, and -CO-R. For instance, m-iodotoluene can be obtained from m-chlorotoluene, mIC 1980 American Chemical Society
