Solving crimes with 3-D fluorescence spectroscopy - ACS Publications

Jul 1, 1985 - Solving crimes with 3-D fluorescence spectroscopy. Jay A. Siegel. Anal. Chem. , 1985, 57 (8), pp 934A–940A. DOI: 10.1021/ac00285a002...
0 downloads 0 Views 8MB Size
I

I

1 I

S34A

The Santa Barbara channel off the coast of California is choked with oil tankers waiting to unload their cargoes of crude oil for refining. Suddenly the water becomes fouled with crude leaking from one oithe vessels. But which one? An old car pulls into the driveway of an expensive home. The driver of the car commits a burglary in the home and leaves the scene. but his car leaves a deposit of motor oil that has leaked from the crankcase. Can the oil spill be matched to the oil of the car? An arsonist is hired by the owner of a failing business to torch the building. He shows up at the building in the middle of the night and tosses a Molotov cocktail, a bottle containing gasoline and a cloth wick that is ignited, through the window. After the fire is put out, an investigator finds the broken bottle still sontaining a small amount of gasoline. Did the gasoline come from a five-gallon can of gasoline found in the suspe~Vscar? These incidenta have certain aspects in common. They occur frequencly in the h i e d States today, and they involve a petroleum-based hydrocarbon product as a critical piece of evidence. These cases require comparisons of a hydrocarbon whose source is known (such as the crude oil in the tankers, the engine oil in the crankcase, and the gasoline in the fivegallon can) with one whose source is not known and which is found at the scene of the incident (the oil slick on the water, the motor oil spill on the driveway, and the remains of the Molotov cocktail). The goal of the chemical analysis in each case is to de-

ANALYTCAL CHEMISTRY, VOL. 57. NO. 8, JULY 1985

termine to a standard of reasonable scientific certainty whether the known samples and the scene samples could have had a common source. The eimilarities among various types and brands of hydrocarbon products create formidable analytical problems that may call for unconventional methodologies.

Pn3sentmethudsdanalysls The conventional methods used to analyze petroleum products in forensic cases are usually some variation of gas-liquid chromatography (GLC). The more volatile products such as gasoline are u a d y analyzed by headspace techniques, perhaps after concentration by a purge-and-trap method. The less volatile oils and lubricants are solvent extracted andlor directly injected. Although GLC can easily determine what type of petroleum product is present, i.e.,gasoline VB. motor oil vs. petroleum jelly, it is not discriminating enough to determine reliably what brand or type of a particular hydrocarbon product is present. Thus, GLC cannot, by itself, be used to determine if two samples could have had a common source. Although capillary column GLC has improved resolution significantly and uses smaller samples, it still cannot overcome the fundamental weakness of this method. -F l specboscopy h hydmcarbon analysis In recent times, the possibility of using fluorescence spectroscopy in the analysis of petroleum products has received increased attention. Virtually ~)003-27W/85/0357-934ASO~ ,5010 @ 1985 American Chemical Society

Jay A. Siege1 Michigan Slate University School of Criminal Justice East Lansing, Mich. 48824-1118

all of the fuels and lubricants derived from petroleum exhibit significant fluorescence because of the presence of various types of polynuclear aromatic hydrocarbons (PAHs). The simple excitation fluorescence spectrum of a gasoline or motor oil sample contains three distinct regions of fluorescence (Figure 1).The relative amounts of the various PAHs in these products are instrumental in determining their particular physical and chemical properties and would be expected to vary within the same type of products made by different manufacturers and perhaps within different lots from the same manufacturer. Thus, the potential exists to use the fluorescence of a particular sample as a unique marker for that product.

This, in theory, would make it possible to determine whether or not two different samples could have had a common source. To be able to use fluorescence as a tool in this type of comparative analysis, however, it is necessary to obtain something more than the conventional excitationemission-synchronous fluorescence spectra generated from a fluorescence experiment. These “simple” spectra reveal the fluorescence of a sample within only one area and do not provide nearly enough data to distinguish between two closely related samples. This type of fluorescence allows us to study only one of the three major areas of fluorescence of the petroleum products of interest a t a time. For this reason, we decided to evaluate 3-D

Flgure 1. Excitation fluorescence spectrum of Quaker State Super Blend Motcf Oil

fluorescence to determine if it could provide enough spectral information to distinguish between two closely related samples to the standard of reasonable scientific certainty. 3-D fW lThree-dimensional fluorescence spectroscopy involves the collection of data encompassing the total fluorescence of a sample. This involves obtaining not one, but a whole series of either excitation or emission spectra in which the nonscanning monocbromator is stepped up by a fixed interval through a whole range of wavelengths while the scanning monochromator scans a fixed wavelength range. Or,in the case of synchronous fluorescence, a series of spectra is obtained in which the difference in wavelength between the two monochromators is increased systematically. The instrument used in this study permits the collection of up to 30 such spectra in a series. The spectra are stored on a floppy disk and then plotted on a Cartesian system with the wavelengths of the scanning monochromator, the nonscanning monochromator (or the wavelength difference in synchronous fluorescence), and the fluorescence intensity as the x-, y-, and z-axes, respectively. The plotting program can also plot the spectra starting with either the highest or lowest wavelength of the nonscanning monochromator (y-axis). This can generate two plots that view the fluorescence profile from either the “front” (Figure Za) or “back” (Figure 2b). If 3-D plots are taken of the excitation, emission, and synchronous fluorescence, one can obtain a total of

ANALYTICAL CHEMISTRY, VOL. 57. NO. 8. JULY 1985

-

835 A

Craig 8. Warren and John P. Walradt, €difors International Flavors and Fragrances Invaluable new tool for all Scientists using computers in their research. Provides a practical approach to changes in inlormation management, product development, and analytical chemistry. Looks at coupling computers with analytical instrumentsto create new analytical techniques. Examines the future of robotics for routine lab operations. Deals with the flavor and lragrancefield but is useful to anyone involved in the development of new molecules or products based on molecules. CONTENTS Application of Microcomputer Technology in Flavor Research: Sensory Evaluation MAECIS: Computer System for Handling and Analysis of Flavor and Fragrance Molecules Computer-Based Molecular Design of Artificial Flavoring Agents Mathematical Approaches for Ouantitative Design of Odorants and Tastants Use of Microcomputers lor Product Optimization Economicsol Laboratory Information Management Systems Instantaneous Analysis of Fragrances, Flavors, and Other VaporPhase Chemicals * New IR Reflectance Analysis Multivariate and GC Techniques in Flavor Research * Analytical Flavor Data: Enhancementwith Computer Techniques Experience in Use of Robotics in Analytical Research Lab

-

-

Based an a symposium sponsored by the Division of Agricultural and Food Chemistry of the American Chemical Society

ACS Symposium Series No.261 176 pages(1984)Clalhbound

LC 84-14607

US 8CanadaS29.95

Order from:

ISBN 06412-0661-1 ExpOR $35.95

American Chemical Society Distribution Mlice Dept. 10 1155SineenthSt., N.W. Washington, DC 20036 orCALL TOLL FREE 800-424-6747 and use your VISA, Mastercard. or American EXoreM credll card.

Flgure 2. (a)Front and (b) back view of the 3-0excitation fluorescence spectrum of Quaker State Super Blend Motor Oil six distinct views of the fluorescence profile.

Gasoiin8 samples In certain incendiary fires, some of the gasoline used to accelerate the fire may be recovered unburned as in the case of the Molotov cocktail that is recovered partially intact. This constitutes the crime scene evidence. If a suspect is apprehended and has in his possession a container that contains some gasoline that is suspected to be the source of the gasoline from the crime scene, then the technique of 3-D fluorescence can be used to determine if the two gasoline samples could have had a common source. In such cases the two gasoline samples to he compared are dissolved in spectro-grade hexane and diluted to a final concentration of 25 ppm. All of the spectra are obtained on a PerkinElmer Model Lambda 5 spectrofluorimeter with a Model 3600 data sta. tion controlling the generation and storage of the spectra. Perkin-Elmer's PLOT program is used to generate the 3-D plots. Prescans are first performed on each sample, and then simple excitation and emission spectra are obtained to identify the important areas of fluorescence and to make a

936 A * ANALYTICAL CHEMISTRY, VOL. 57, NO. 8. JULY 1985

preliminary determination of the likelihood that the two samples could have had a common source. If the simple spectra are markedly different a t this point, we can immediately conclude that the samples had different sources. If the spectra are similar then 3-D analysis is performed. Examination of 3-D spectral plots of many gasoline samples in our laboratory has shown that the 3-D emission plot is more effective a t differentiating between two samples of gasoline than are the excitation or synchronous excitation plots. Therefore, 3-D emission plots are obtained for each sample. The architecture of the disk operating system and the capacity of the disks permit storage of three 3-D spectra stacked plots on one disk. The 3-D emission plots of the two samples of gasoline are put on the same disk, which leaves room on that disk for the resultant plot obtained by computer comparison of the two plots. Hard copies of the front and back views of the emission plots are obtained and visually compared to determine if they are obviously different, although experience dictates that if the simple emission spectra are very similar, the stacked plots will be also. Because these stacked plots are quite

..becaui

!our standards are i tough as,

Your water purity standards must be met with precision every single day. Othewise your productivity and profits can be drastically reduced. At Continental, we have the expertise and technology to design water purification equipment to meet any standard 01 your industry. In (act. Continental has pioneered the new technology to do so. As a result, we've designed systems for literally thousands of industries across the country meeting production needs from pharmaceutical and biotechnology manufacture to microchip rinsing. But Continental's strength doesn't lie only in technology.

_-______-----I I

I I 1 - Ultrafiltration I Microliltration I I I PhOW I state 28" I -------

It's the resourcefulness and ingenuity of our design and service professionals that really make a diflerence. The Continental team is topnotch, ready to get right to the heart of your problem, assess your needs, and give you practical, economical solutions. And we guarantee we'll be around alter the equipment is in place to ensure your satisfaction. Add to this the fact that continental has factory-trained service specialists in more than 40 locations, and you'll see why more industries are calling Continental for water purification systems. It's the one clear choice.

continental"

Plewe Send more information regarding your Capabilities and producIs I am interested in:

I I I I I I

C Reverse Osmosis

c Deionization Name

F'm

C'h

_-_____

Continental Water Systems Corporation 5405 Bandera Rd. #117. San Antonio, TX 78238 CIRCLE 37 ON READER SERVICE CARD

L+A-A

Water Systems Corporation P.O. Box 40369 San Antonio, Texas 78229

1-800-426-3426 Texas Residents Call (512)

680-2851

Telex 75-7331

Worilmderih

??.whyWutd

Chemical Research Faculties: An International Directory

For the first time -a new internationaldirectory from the ACS. Contains all the information you need on chemical research and researchersat universities around the world. Offers data on who’s doing what in chemistry, chemical engineering, biochemistry, and pharmaceuticallmedicinal chemistry. Gives you information on over 735 academic departments worldwide and lists more than 8,900 faculty mernbers and their areas of specialization. Also provides invaluable data on graduate programsat universities in more than 60 countries as well as names, addresses, and important information on 63 chemical societies around the world. Cross-indexed three ways for easy use. An indispensable referencefor industry and academia alike. 640 pages(1984)Slothbound complex, however, visual comparison LC 83-22323 ISBN 0-84120817-4 may not he sufficient to make a definiUS aCanadaSt29.95 E X ~ ~ R S I ~ ~ .tive S ~determination. Thus we wrote a

I

Order from: American Chemical Society DistributionOffice Dept. 60 1155 Sixteenth St., N.W. Washington, DC 20036 or CALL TOLL FREE 800-424-6747 and use your MasterCard, VISA, or American Expresscredit card.

program that would permit computerized comparison of two plots by suhtracting one from the other, spectrum hy spectrum, and then prepare a stacked plot of the resultant spectra. If the 3-D emission plots of the crime scene gasoline sample and the sample of known origin do have a common source then the resultant spectra oh-

1 938A

ANALYTICAL CHEMISTRY, VOL. 57, NO. 8, JULY 1985

tained by subtracting one plot from the other should have essentially zero intensity a t any point in the plot. If, on the other hand, the two samples being compared have different sources, then the subtracted plot should have some regions of net fluorescence. This program also accounts for “negative fluorescence” obtained from subtraction by adding a factor equal to the highest negative intensity . tothe whole p k t .

One of Lady Ashby's guests has become a corpse, having fallen against the mantelpiece. Spectra of the meager clues found at the scene appear above. The clues are: an empty pillbox with the initials "MM" engraved on it, a sliver of currency, and a 14 micron fiber from beneath a fingernail of the deceased. Was the deceased drugged? Robbed? Or assaulted? The suspects are: (A) Mr. Clive Woolsey, a carpet salesman from Perth; (B) Miss Mildred Mockingbird, a very nervous pharmacist from Miami; and (C) J.D. Buckpasser, a retired rugby player who has just won a great deal of money playing bridge. All the suspects say they were nowhere near the scene of the

crime, and the last ferry to the mainland leaves at six p.m. A fast, accurate solution is needed. The mystery can be solved by the Model 1800 FT-IR from Perkin-Elmer. Its great speed and accuracy make the Model 1800 ideal for super sleuthing. These clues make difficult samples. But the very high signal-to-noise ratio of the Model 1800 (better than 7000:l RMS with an MCT detector) lets you run low transmission samples, or use energy-robbing accessories and still get extremely accurate results. Have you put it together? The pillbox does belong to Miss Mockingbird, but the dust in it was only caffeine.The currency drew attention to Mr. Buckpasser, but his winnings

were all in Pounds Sterling - and the ink on the currency was from the U.S.Treasury. This leaves Mr. Woolsey, who, despite his name, sells polyester carpets. The 14 micron fiber is polyester, and Mr. Woolsey has a great deal to explain. You, too, can be a brilliant sleuth with the Model 1800. For complete details, contact your Perkin-Elmer representative. For literature, please call 1-800-762-4000. Perkin-Elmer Corp., Main Ave. (MS-12), Norwalk, CT06856 U S A Tel: (203) 762-1000.Telex 965-954. Bodenseewerk Perkin-Elmer & Co., GmbH, Postfach 1120, 7770 Ueberlingen, Federal Republic of Gerrnany.Tel:(07551) 811 Perkin-ElmerLtd.. Post Office Lane Beaconsfield, Bucks HP9 I Q A , England. Tel: Beaconsfield (049 46) 6161,

-E LM

The science and computer company. Where solutions come first. CIRCLE 165 ON READER SERVICE CARD

ANALYTICAL CHEMISTRY, VOL. 57, NO. 8, JULY 1985

939A

Intensity

CONTENTS

Medical School influences on the Ficton of Arthur Conan Doyle Strong Poison: Chem:stry in the

1 Judicial Process

Based on a symposium sponsored by

I

the Divisron of Hlsforyof chemtsiry the Amenean Chemtcal Society

I

1 3 5 ~ ea fi9w) u o t w n d LC 8371907 ISBN 0-841 US 8 Canada $19.95 ExPo

135pges(lW)Paper LC8311907 ISBN 044125785.2

US &Canada$14.85

ExWrI$17.95

Order from: American Chemical Society

1

DlstrlbutlonOffice Oept. 54 1155 Sixteenth St., N.W. Washington, DC 21x136 or CALL TOLL FREE 800-424-6747 and use your VISA or MasterCard.

940,.

ANALYTICAL CHEMISTRY, VOL. i

Figure 4. Threedimensional stacked plots of (a)unknown and (b) known gasoline samples of common algln and (c)the result of Subtracting the two stacked plots Figures 3a and 3b are the 3-D emission spectral plots of an unknown sample of gasoline and a sample of known origin, respectively. The resultant stacked plot in Figure 3c indicates that the samples had different sources, which is in fact the ease. Likewise, Figures 4a and 4b are emission stacked plots of different known and unknown gasoline samples. The flatplane appearance of Figure 4c, the result of subtracting the two stacked NO. 8.

JULY

1985

plots, suggests chat the two samples have a common source, which was the case in this instance. The results of these examples of 3-D fluorescence as well as other experimental work suggest that this technique has great potential as a test that allows the analyst to determine, with reasonable certainty, if two petroleum product samples in a criminal or civil case could or could not have had a common source.