ARSON ACCELERANT ANALYSIS BY ATTENUATED TOTAL REFLECTANCE SPECTROSCOPY by Stephen P. Ray Submitted in Partial Fulfillment ofthe Requirements for the Degree of Master ofScience in the Chemistry Program SCHOOL OF GRADUATE STUDIES YOUNGSTOWN STATE UNIVERSITY AUGUST 1998 Arson Accelerant Analysis by Attenuated Total Reflectance Spectroscopy Stephen P. Ray I hereby release this dissertation to the public. I understand that this dissertation will be housed at the Circulation Desk ofthe university library and will be available for public access. I also authorize the University or other individuals to make copies ofthis dissertation as needed for scholarly research. Signature: Approvals: 7/';}/lY Date iii Abstract Arson is a serious problem both locally and nationally. In 1994, the National Fire Protection Association reported over 100,000 arson fires with over $1.2 billion in damages and 550 deaths. Locally, in 1996, Youngstown had 326 arson fires with almost $4.0 million in damages and five deaths. Often these fires are accelerated by flammable materials. Organic, petroleum-based, non-water soluble solvents, such as gasoline, can be detected at very low concentrations after an intense fire by several well developed techniques. However, water soluble solvents, such as methyl (wood), ethyl (grain), and isopropyl (rubbing) alcohol have proven to be difficult to analyze. Not only does the water used to extinguish the fire wash away the accelerant by convection, it also disperses it by dissolution. Additionally, most techniques used for analysis require organic solvents to dissolve the materials to be analyzed. Since the accelerants to be studied are in water, they would need to be extracted with organic solvents reducing their concentration further. Attenuated Total Reflectance Spectroscopy (ATR), a relatively sensitive and selective technique, was used to perform analysis on these water-soluble accelerants. Concentration gradient experiments, time controlled burn experiments, and field-controlled burns were performed with detectability in more than 75% ofthe samples. Most tests were on carpet samples, but some tests were on cloth samples as well. Using ATR has proven to be an ideal way to handle those situations where a water soluble accelerant needs to be detected. iv Acknowledgements I would like to thank the following people for their help and support throughout this project. First and for most, I would like to thank my family for their constant encouragement during the rough times ofthis project and for the kick in the rear when I needed it. I would also like to thank Dr. Daryl W. Mincey, my research advisor, who also help to encourage me throughout the project and for his insight and suggestions without which I would still be on the old drawing board. I would like to thank Dr. Larry Curtin, my assistant pyromaniac and thesis committee member, and Dr. Sherri Lovelace-Cameron, thesis committee member. I wish to thank my friends, especially Ryan and Jimmy, who have always been there to listen to my when I just needed to ramble on - sorry guys. I wish to thank Youngstown State University for the resource to perform this project. I would also like to thank Home Carpet in Boardman, Ohio for donating the carpet samples that were used throughout this project. Finally, I would like to thank the Youngstown Arson Bureau, mainly Lt. Bob Sharp and Sharon Sawyer for all their help in coordinating the controlled bums in the field. It was a much needed part ofmy research and I couldn't have done it without you. Table ofContents Title Page .i S · P .. Ignature age 11 Abstract .iii Acknowledgements .iv Table ofcontents v List ofSymbols and Abbreviations vii List ofFigures .ix List ofTables xii Chapter I: Introduction 1 Statistics ofArson l Arson Investigation ., ., 2 IR Theory 3 Quantum Mechanics ofInfrared Spectroscopy .4 Interpretation ofIR Spectra 6 Evanescent Wave 7 Reflective and Refractive Properties ofLight With Water 10 Conditions For Internal Reflection Spectroscopy .,12 Problems With Internal Reflection Spectroscopy 17 Advantages OfAttenuated Total Reflectance Spectroscopy 18 The Infrared Spectrometer 19 v Chapter 2: Project Information 21 Chapter 3: Methods and Materials 23 Materials 23 Methods 23 FTS 40 Components 23 Absorbing Materials 26 Chapter 4: Procedure and Results 29 Preparing Concentration Gradient Curve Solutions 29 Obtaining Spectra 29 Initial Burning OfCarpet Samples .30 Burning OfCarpet Samples - Second Stage 31 Burning OfSamples - Third Stage ,.31 Results 34 Third Stage Burns 35 Chapter Five: Discussion and Results 101 Third Stage Burns 102 References 105 vi Symbol ATR IR I m rr mm IRS Ur Il m n v R a FT-IR GC-MS v:v vii List ofSymbols and Abbreviations Definition Attenuated Total Reflectance Infrared Moment ofInertia molecule axis ofrotation millimeters Internal Reflection Spectroscopy angle ofincidence critical angle micrometers refractive index propagation velocity reflectivity incident identity reflected intensity Frustrated Total Reflection Internal Reflection Element adsorption parameter Fourier Transform-Infrared Gas Chromatography-Mass Spectrometry volume to volume viii mL milliliters nm nanometer L/min Liters per minute pSI Pounds per square inch °C Degree Celsius kHz kilohertz sec. seconds hrs. hours mIll. minute(s) ix List ofFigures Michelson Interferometer .27 Methanol standard solutions - spectra .36-39 Ethanol standard solutions - spectra .40-43 Isopropanol standard solutions - spectra 44-47 Acetone standard solutions - spectra .48-51 Methanol: absorbance vs. % v:v , 53 Ethanol: absorbance vs. % v:v 54-55 Isopropanol: absorbance vs. % v:v 56-58 Acetone: absorbance vs. % v:v 59-60 Methanol burn 2/12 - 25 sec 61 Methanol burn 2/12 - 90 sec '" 62 Ethanol burn 2/19 - 30 sec 63 Ethanol burn 2/19 - 90 sec 64 Ethanol burn 2/19 - 150 sec 65 Isopropanol burn 2/19 - 30 sec 66 Figure Page Exponential decay ofelectric field 9 Effect ofthe angle ofincidence 11 Propagation oflight through a medium 13 Principle ofrefraction 14 FTS-40 optical schematic diagram 24 ATR cell reservoir. 25 1.1 1.2 1.3 1.4 3.1 3.2 3.3 4.1-4.4 4.5-4.8 4.9-4.12 4.13-4.16 4.17 4.18-4.19 4.20-4.22 4.23-4.24 4.25 4.26 4.27 4.28 4.29 4.30 4.31 Isopropanol burn 2/19- 90 sec , 67 4.32 Isopropanol burn 2/19 - 150 sec 68 4.33 Methanol: absorbance vs. burn time 70 4.34 Methanol: % recovery vs. burn time 71 4.35 Ethanol: absorbance vs. burn time (1045 em-i) 72 4.36 Ethanol: % recovery vs. burn time (1045 cm-\ 73 4.37 Ethanol: absorbance vs. burn time (1086 em-I) 74 4.38 Ethanol: % recovery vs. burn time (1086 cm-\ 75 4.39 Isopropanol: absorbance vs. burn time (1126 cm- I 76 4.40 Isopropanol: % recovery vs. burn time (1126 cm- I ) 77 4.41 Isopropanol: absorbance vs. burn time (1164 cm- I ) 78 4.42 Isopropanol: % recovery vs. burn time (1164 cm- I ) 79 4.43 Ethanol burn - second stage - 30 sec 80 4.44 Ethanol burn - second stage - 90 sec 81 4.45 Ethanol burn - second stage - 240 sec 82 4.46 Ethanol burn - second stage: absorbance vs. burn time 84 4.47 Sample 2 - 4/29 85 4.48 Sample 3 - 4/29 86 4.49 Sample 4 - 4/29 87 4.50 Sample 5 - 4/29 88 4.51 Sample 8 - 4/29 89 4.52 Sample 10 - 4/29 90 4.53 Sample 11 - 4/29 91 x 4.54 Sample 12 - 4/29 92 4.55 Sample 15 - 4/29 93 4.56 Sample 16 - 4/29 94 4.57 Sample 17 - 4/29 95 4.58 Sample 18 - 4/29 96 4.59 Sample 19 - 4/29 97 4.60 Sample 20 - 4/29 98 4.61 Sample 21 - 4/29 99 4.62 Sample 22 - 4/29 100 xi List of Tables Table Page 4.1 List ofparameters 29 4.2 Concentration gradient data 52 4.3 Stage one bums data 69 4.4 Ethanol bums - Second Trial data 83 xii 1 Chapter One Introduction The characterization ofwater soluble accelerants used in arson fires is difficult to accomplish by commonly used techniques in a forensic lab. Ofother possible techniques, the one used in this study was Attenuated Total Reflectance (ATR) Infrared Spectroscopy. ATR occurs when radiation passes through a prism made up ofa substance with a high refractive index relative to aqueous solutions and reaches the interface ofthe prism where the sample is in contact with it. The radiation penetrates the sample to the depth ofa few micrometers and then suffers total reflectance. An accelerant is any substance that is used to accelerate (and sometimes direct) the spread ofa fire. The accelerants ofinterest in this study were those that are water-soluble and readily available to an arsonist. They include simple alcohols such as methanol (wood alcohol), ethanol (grain alcohol), isopropanol alcohol (rubbing alcohol) and acetone. Statistics ofArson Arson is the intentional setting ofa fire. Arson is a crime that kills people and destroys property and life. In 1994, The National Fire Protection Association reported over 100,000 arson-related fires, with over $1.2 billion and 550 lives lost. 1 Locally, in 1996, Youngstown had 326 arson fires that resulted in $3,805,805 in losses and 5 deaths. 2 The arson fires were not only to residential structures but also commercial structures and vehicles. In the period from 1990-1997, Youngstown has averaged 326 arson fires per 2 year with $2,238,797 in damages and 4.7 deaths 2 . This is a serious problem that will not get better until there is not anyway for an arsonist to get away. Arson Investigation An arson investigation usually begins immediately after the fire is brought under control. The fire battalion chiefmakes a preliminary cause and origin determination. If arson is suspected, an investigator is sent to the scene. The fire scene is photographed from all angles. The investigator will then do a cause and origin determination to reassert the battalion chiefs assessment. Samples are then taken from selected areas throughout the scene, with emphasis at the site oforigin. The evidence is recorded and sent to the crime laboratory for accelerant analysis. While the evidence is being tested, the investigation centers on motive, means, and opportunity. Arson is one ofthe most difficult crimes to prosecute because there is usually little direct evidence. An arson case depends mainly upon the chemical analysis ofphysical evidence collected at the scene. The basic goal ofchemical analysis offire debris is to establish whether materials are present in the remnants ofthe fire that could have been used to start and/or accelerate the fire. Even a highly volatile fuel, such as gasoline, is detectable after an intense fire. According to Bertsch, in controlled bums, these fuels are still detectable after 90% evaporation, i.e. 90% ofits mass is gone. Kerosene maintains many ofits characteristic chromatographic features following 90% evaporation. 1 3 IR Theory Infrared (IR) Spectroscopy deals with the interaction ofinfrared radiation with matter. The infrared region ofthe electromagnetic spectrum lies between the visible and microwave regions. This region is divided into three categories: Region Near-infrared Mid-infrared Far-infrared Range (cm- I ) 13,300-4,000 4,000-400 400-10 Infrared spectroscopy theory states that a molecule may absorb infrared radiation ofthe appropriate frequency to excite it from one vibrational or rotational level to another. When a beam ofinfrared energy, covering a broad frequency range, passes through a sample, the energy is absorbed at certain frequencies by the sample. In polyatomic molecules, many kinds ofvibration are possible, which gives rise to an infrared spectrum composed ofmany relatively sharp absorption bands. The frequencies ofthe molecular vibrations, and consequently the wavenumber ofthe absorption bands appearing in the vibrational spectra, are determined by the mass ofthe atoms and the forces acting between them. 3 A graph ofenergy absorbed versus frequency can then be plotted. This is the absorption spectrum ofthe sample. The spectrum is characteristic ofthe particular molecule and its molecular motions. 4 Most molecules absorb infrared radiation with the exceptions being homonuclear diatomics. 5 The quantized energy uptake ofsubstances is connected with the internal motions ofthe molecule. The change in energy ofthe molecule is composed ofenergy changes of three kinds, associated with electronic motion, rotation, and vibration. 6 Absorption 4 occurring in the region ofinfrared radiation is caused by a change in the vibrational and rotational state ofa molecule. Vibrational and rotational motions where absorption can be detected only occurs for polyatomic molecules. The fundamental postulates upon which chemists base the utility ofinfrared spectroscopy in determining structural data can be stated as follows: 1. Organic substances exhibit characteristic group frequencies in the infrared regIOn. 2. The absorption spectrum ofa given substance is generally specific for that and only that substance. 3. The absorption spectrum ofmixtures is generally additive, i.e. the sum ofthe individual spectra ofthe components. 4. The intensity ofan absorption band is related to the concentration ofthe substance that absorbs the incident radiation. [7] Quantum Mechanics ofInfrared Spectroscopy According to quantum mechanics, adsorption ofradiation induces a change in the rotational state ofa molecule ifthe dipole moment ofthe molecule is also changing, or, in the case ofrotation combined with vibration, ifa dipole moment transition occurs during vibration. 8 For polyatomic molecules, the structure ofthe rotational spectrum is determined by the distribution ofthe moment ofinertia ofthe molecule. The moment of inertia, I, is given by the masses ofthe atoms forming the molecule, mi , and their distances from the axis ofrotation, ri . (eqn. 1) [10] 5 Investigating the value of! along directions starting from the center ofgravity, it is found that in three mutually perpendicular directions the value ofI exhibits a maximum or minimum. In vibratory motion, atoms undergo vibrations about their equilibrium position in a molecule while the center ofgravity remains in the same position. To discuss the vibrational modes ofpolyatomic molecules, a knowledge ofnormal modes and ofthe symmetry characteristics ofthe molecule is needed. For normal vibrations, each atom forming the molecule undergoes near-harmonic vibration ofidentical frequency in identical or opposite phase along the straight line passing through the equilibrium position. ii The vibration ofthe molecules results from the superposition of the normal vibrations. According to Svehla, normal vibrations can be divided into two groups. In the first group, the displacement ofthe atom occurs in the direction ofthe valence bond, so that bond distances will increase and decrease periodically. The bond angles change only ifit is necessary to ensure that the center ofgravity remains the same. These vibrations are referred to as stretching vibrations. The other type is known as deformation vibrations. Here, the bond angles change periodically and the bond distance will change only ifnecessary to ensure the center ofgravity remains the same .11 Infrared spectra can be taken on solids, liquids and gases. Generally, work is done at room temperature and atmospheric pressure. When measuring the IR spectra ofliquids, it is very important that the cell windows and the solvent used transmit in the infrared region selected for analysis. For quantitative examination, the spectra ofsome substances must be taken using in solution. The interaction between solute and solvent, the so-called solvent effect, increases with increasing complexity and polarity ofthe solvent molecule. i2 Whenever possible, aqueous solutions should be avoided. Water dissolves most materials 6 used to make cell windows, and moreover it has strong absorption bands, so that it can only be used below a cell thickness of0.02 mm. 25 The spectra ofaqueous solutions can be taken without putting them in a cell, by forming a capillary film on a support frame in the presence ofa surfactant. Owing to the strong absorption ofwater, the energy ofthe reference beam path must also be reduced in most cases. It is best to examine substances soluble only in water in reflected light or in the form ofan emulsion. The transmission spectra ofattenuated total reflectance and H 2 0 give reasonable results. The advantages of ATR will be discussed later. Interpretation ofIR Spectra Interpretation ofan IR spectrum is done by looking at the characteristic bond and group frequencies. Characteristic bond and group frequencies depend on the atoms which form the molecule, on the nature ofthe chemical bonds between the atoms, and on the structural characteristics ofthe molecule. From the positions ofthe bands appearing in the spectrum and from their intensities or relative intensities, conclusions can be drawn regarding the types ofbonds, atomic groups and their arrangement within the molecule. 13 Some ofthe physical and chemical effects that considerably affect the position, shape, and intensity ofthe bands appearing in the spectrum are: 1. Mass effects, due to a change in the mass vibrating with the group and to isotope exchange 2. Steric effects, among which diastereomerism, rotational isomerism, ring stretching, steric crowding and restriction, bond angle distortion, collinearity, and coplanarity should be mentioned. 7 3. Effect ofthe electron affinity ofsubstituents, leading to hyperconjugation, inductive and mesomeric effects. 4. Tautomerization effects, e.g. giving rise to keto-enol forms. [14] Evanescent Wave The history ofInternal Reflection Spectroscopy (IRS) began in the early 1600's when Newton observed an evanescent field in a lower index ofrefraction medium in contact with a higher index ofrefraction medium in which a propagating wave ofradiation undergoes total internal reflection. According to Mirabella, the established method ofthe time was a transmission method in which the critical angle ofrefraction was located with the sample placed as a thin film between the hypotenuse face oftwo isosceles prisms of glass. At angles smaller than the critical angle, light was transmitted through both prisms. At angles at or above the critical angle, total internal reflection occurred at the prism sample interface ofthe first prism and no light was transmitted through the second prism. For most angles ofincidence above the critical angle, the reflection spectra resemble transmission spectra fairly closely; however, for angles ofincidence just below the critical angle the spectra may resemble the mirror image ofthe dispersion in the index of refraction. In order to discuss the theory ofIRS, a description ofthe properties ofthe evanescent field is needed. Some ofthe basic properties ofthe evanescent wave are as follows: 1. The field intensity in the rarer medium is nonzero and there is an instantaneous normal component ofenergy flow into this medium whose time average is 8 zero. Thus, there is no loss ofenergy and the propagating radiation in the denser medium is totally internally reflected, except at those frequencies in the rarer medium where absorption occurs. 2. The evanescent field in the rarer medium is a nontransverse wave and has components in all spatial orientations. 3. The evanescent field is confined to the vicinity ofthe surface ofthe rarer medium and decreases in intensity with distances into this medium normal to the surface. 4. There is a net energy flow parallel to the surface resulting in a displacement of the incident and reflected waves. [15] The evanescent field decays exponentially in the rarer medium as a result ofthe presence in an optically denser medium (one having a high refractive index) ofa standing wave established at a totally reflecting interface. The exponential decay ofthe electric field amplitude (ElE o ) and the intensity (ElE o )2 as a function ofdepth into the surface Z is shown by Figure 1.1. 16 It shows that most ofthe absorbance information comes from the first 3 or 4~mofdepth into the surface. The fact that the evanescent wave is nontransverse, and has vector components in all spatial orientations, is significant because it permits these vector components to interact with dipoles in all orientations. The properties ofthe evanescent field in the rarer medium also depend on the thickness ofthat medium. The interaction ofthe evanescent field with bulk materials and thin films can be expressed in terms ofan effective thickness. The effective thickness is a measure ofthe strength ofcoupling to a sample and is useful in comparing internal reflection spectra to transmission spectra. 17 Two distinct cases can be defined. The first is the semi-bulk case which is when the electric field amplitude falls to a very low value ~m~m G.g ::It~ DR n, G IIl C l'igure 1 • 1 'Ratio of inrensity and arnplilUdc of electric vector at depth Z to Ihat ~tthe surface versus the depth Z. The values used to calculate the curves wereIl~ = 1.50. III "" 2.38, ilnd 0 = 45" at a wavelenglh of 10 IJ.m. 9 ::It~ ~tIl~ ::It~ 10 within the thickness ofthe rarer medium. Second, the thin film case is when the electric field amplitude remains essentially constant over the thickness ofthe rarer medium. Reflective and Refractive Properties ofLight with Water Before discussing the specifics ofIRS, some ofthe basic principles governing the interaction oflight with water should be established. More specifically, reflective and refractive properties oflight propagating through media having distinctive optical properties are ofinterest because they will dictate not only the amount ofenergy propagating through but also the direction ofpropagation. 18 Light is an electromagnetic wave. When electromagnetic radiation strikes an interface between media oftwo different refractive indices, both refraction and reflection can occur. When the two media are in contact with each other, the path ofthe light will be distorted, depending on the angle of incidence (see Figure1.2). 19 According to Harrick, light is transmitted at a 90° angle of incidence and partially reflected at 0.1 < e c or totally reflected at 0.1 > e c . When the angle ofincidence, 0.1, is greater than the critical angle,e c , the light is totally reflected, and this forms the basis for internal reflection spectroscopy. Under the circumstances of0.1 > e c , the amount ofelectromagnetic energy being put in the sample by the evanescent wave exceeds the amount coming out, and the evanescent wave will be attenuated.20 In spectroscopy, we start with a system in some stationary state, expose it to light, and observe whether or not the system has made a transition to another stationary state. 'r-ammiUed Refracted It 1 1 III Total reflection Fi gure 1 .2 Tffect ofaI,gle oj'incirlence un direction~fl)ml)(I!.!ml(Jll~fl)ml)(I!.!ml(Jll 12 Conditions For Internal Reflection Spectroscopy A realistic set ofconditions in IRS include a medium oflow absorbance and a finite diameter irradiating beam. Further, an absorbing rarer medium requires considerations ofdispersion ofthe refractive index across an absorbing medium. 21 The refractive index is a measure ofthe medium's interaction with radiation. The variation of the refractive index with wavelength ofradiation is referred to as dispersion. Dispersion curves usually show a gradual increase in the refractive index with decreasing wavelength. Iflight propagates through a medium with refractive index nl and enters a medium with refractive index n2 (Figure 1.3), the light path will change, and the extent ofrefraction is given by (eqn.2) [22] where 0.1 and 0.2 are the angles ofincidence and refraction, respectively, and VI and V2 are propagation velocities in media 1 and 2, respectively. The principle ofrefraction, known as Snell's law, is illustrated by Figure 1.4 and is given by R = IIJIo (eqn. 3) [23]. The same results can be obtained by starting with Maxwell's equations which also clearly show the essential presence ofthe evanescent wave. From this mathematical treatment, the dependence ofreflection on angle ofincidence and polarization is obtained for non-absorbing and absorbing media. The reflection phenomenon is then treated from a physical viewpoint which is valid for low absorptions. The latter treatment gives a clear insight into the interaction mechanisms ofthe evanescent wave with the absorbing rarer medium. 'I'~ansmiued Refracted It 11 HI Total reflection Fi gure 1 . 2 Fffi!ct ofaI.gll' oj'incidl'nce on direction (!{jJroj){lpmiOll 'I'~ansmiued'I'~ansmiued 12 Conditions For Internal Reflection Spectroscopy A realistic set ofconditions in IRS include a medium oflow absorbance and a finite diameter irradiating beam. Further, an absorbing rarer medium requires considerations ofdispersion ofthe refractive index across an absorbing medium. 21 The refractive index is a measure ofthe medium's interaction with radiation. The variation of the refractive index with wavelength ofradiation is referred to as dispersion. Dispersion curves usually show a gradual increase in the refractive index with decreasing wavelength. Iflight propagates through a medium with refractive index nl and enters a medium with refractive index n2 (Figure 1.3), the light path will change, and the extent ofrefraction is given by (eqn.2) [22] where (li and (l2 are the angles ofincidence and refraction, respectively, and VI and V2 are propagation velocities in media 1 and 2, respectively. The principle ofrefraction, known as Snell's law, is illustrated by Figure 1.4 and is given by (eqn.3) [23]. The same results can be obtained by starting with Maxwell's equations which also clearly show the essential presence ofthe evanescent wave. From this mathematical treatment, the dependence ofreflection on angle ofincidence and polarization is obtained for non-absorbing and absorbing media. The reflection phenomenon is then treated from a physical viewpoint which is valid for low absorptions. The latter treatment gives a clear insight into the interaction mechanisms ofthe evanescent wave with the absorbing rarer medium. 13 I:Ct~~1AJ .._~~~---­ 2.: 5«W' ple.. I I I Ie Figure 1.3:~chematicdiagram of the perpendicularly polarized light beam Ev impinging on the interface between two semi-infinite media and split into two beams: one transmitted and one reflected. Ct~~1AJ .-IIII!~~---­ ~chematic Ct~~1AJ .-IIII!~~---­ 14 Fi gure 1. 4 Sne 11 I sLaw 15 For internal reflection, the electric field amplitude at the surface is nearly zero. For total internal reflection, there is still a sinusoidal variation ofthe electric field amplitude with distance from the surface in the denser medium; however, by selecting the angle of incidence, it is possible to obtain large electric field amplitudes. 24 It is important to note that electric fields exist in all spatial directions at the reflecting interfaces. This is one reason for the difference between IRS and transmission spectra, i.e. dipoles will absorb energy in internal reflection regardless oftheir orientation, whereas for transmission, dipoles oriented parallel to the direction ofpropagation cannot absorb energy.2S Newton's experiments showed, and it follows from Maxwell's equations, that an electromagnetic disturbance exists in the rarer medium beyond the reflecting interface for total internal reflection. This disturbance is unusual since it exhibits the frequency ofthe incoming wave, but it is an evanescent wave whose electric field amplitude falls off exponentially with distance from the surface. The depth ofpenetration, is defined as the distance required for the electric field amplitude to fall to e- 1 ofits value at the surface. To demonstrate the existence ofthe evanescent wave and to measure its depth ofpenetration, it is necessary to disturb it either by absorbing its energy or by redirecting it. 26 Light escaping a medium very near the critical angle may be coupled back into this medium and reappear at a distance along the surface, considerably removed from the point ofescape.27 There are two distinctly different methods ofcoupling the evanescent wave and extracting energy from it and thereby making reflection less than total. In one coupling mechanism some or all ofthe energy is redirected and there is no energy loss, whereas in the other, energy is absorbed and there is a loss. The former is known.as frustrated total reflection (FTR) while the latter is known as attenuated total reflectance 16 (ATR).28 Attenuated total reflection is due to an absorbing coupling mechanism whereby the reflectivity for total internal reflection can be continuously adjusted between some value greater than 0 and 100% by placing an absorbing medium in contact with the reflecting surface. 29 Attenuated total reflection is observed when the angle ofincidence is set and remains above the critical angle and the wavelength is swept through an absorption band. When electromagnetic radiation strikes an interface between media 1 and 2 that have different refractive indices, reflection also occurs. The fraction ofradiation that is reflected becomes larger with increasing differences in refractive index. 22 The case ofan absorbing rarer medium can be treated in terms ofthe intensity loss per reflection. If1 0 is the incident intensity and I R is the reflected intensity, then the reflectivity, R, is given by: (eqn.3) [23] For total reflection, I = 1 0 and R = 1. The dispersion in the index ofrefraction in the vicinity ofmolecular absorption bands affect the nature ofthe optical spectrum for transmission and reflection. For internal reflection spectroscopy, where the angles of incidence employed may be between a value somewhat below the critical angle to "grazing incidence", the dispersion in the refractive index also plays an important role in the nature ofthe spectrum ofbulk materials, but the degree ofits influence depends on the angle of incidence. For thin films, the effect ofthe dispersion in the refractive index is quantitatively similar to that observed in transmission spectra. 30 Physical insight into the nature ofthe spectrum can be obtained from a consideration ofthe absorption parameter equations as the refractive index in the rarer medium changes due to the dispersion near molecular resonances. Internal reflection spectra do not represent a single optical 17 constant but are related to both the refractive index and absorption coefficients. 31 The dispersion in the refractive index affects the spectra. Furthermore, the effect ofthis dispersion is dependent on the angle ofincidence near the critical angle for measurements on bulk materials. Total internal reflection occurs when light traveling in an optically dense medium impinges on an interface with a less dense medium. Problems With Internal Reflection Spectroscopy The primary problem in obtaining high quality qualitative spectra or for performing quantitative IRS work is the difficulty in obtaining good and reproducible contact ofthe sample to the internal reflection element (IRE). The IRE is the transparent optical element used in internal reflection spectroscopy for establishing the conditions necessary to obtain internal reflection spectra ofmaterials.32 The ability to obtain an internal reflection spectrum and information from the spectrum is determined by a number ofcharacteristics ofthe IRE. Choices must be made regarding the working angle or range ofangles of incidence, number ofreflections, aperture, number ofpasses, surface preparation, and material from which it is made. 33 These characteristics can be determined with some knowledge ofthe sample to be studied, the interaction mechanisms, and a few simple considerations regarding the reflection element. In general, it is advantageous to use an internal reflection element oflow reflective index, and an angle near the critical angle because a higher contrast spectrum is obtained in this manner.31 The index ofrefraction changes with wavelength regardless ofthe material. Therefore, the degree ofrefraction at the entrance face ofthe IRE, hence the internal angle ofincidence, e, on the sampling surface is wavelength dependent for oblique incidence. Since it is desirable to maintain the 18 same angle ofincidence over the entire wavelength range under investigation, refraction should be eliminated. This is the case when light enters and exits the IRE via the aperture at normal incidence. The aperture ofthe IRE is thus defined as that portion ofthe beveled area which can be utilized to conduct the light into the IRE at the desired angle of incidence e. An absorbing medium may strongly affect the reflectivity for internal reflection, particularly in the vicinity ofthe critical angle. It should be noted that when the absorbing medium is absorbing, the critical angle loses its significance; i.e. there is no longer a sharp critical angle for the absorbing case and the reflectivity curves become less steep in this region. 34 The absorption loss is quite large near the critical angle, is greater for parallel polarization than it is for perpendicular polarization, and decreases with increasing angle ofincidence for both polarizations. It is interesting to note the dependence ofthe reflectivity on absorption coefficient for internal reflection at angles exceeding the critical angle. The absorption parameter, a, is defined as the reflection loss per reflection and is greater near the critical angle than it is for larger angles and is also greater for 11 polarization than for -i-polarization. 34 Advantages ofAttenuated Total Reflection Spectroscopy The primary advantage ofusing internal reflection spectroscopy is that water soluble substances that are usually very difficult to look at using transmission methods can be studied. Another ofthe advantages is that the IRE's used, most likely Ge or ZnSe cells, are not water soluble whereas in transmission methods NaCI cells are used which are 19 water soluble. Although AgCl could be used in transmission, transmission thin layer cells are difficult to construct. In attenuated total reflectance spectroscopy, the sample is placed against the internal reflection crystal. Depending on the angle ofincidence, the IR beam undergoes multiple internal reflections before it exits the crystal. A more detailed description, including figures, is in Chapter Three. The refractive indices ofthe sample and the crystal are the critical factors that determine what the obtained spectra will look like. ATR band intensity is proportional to concentration in accordance with Beer's Law. 35 This is the only difference between IR absorption spectra and ATR spectra. The Infrared Spectrometer The spectrophotometer has four fundamental components: 1. A stable source ofenergy emitting continuous radiation. 2. An interferometer. 3. A detection system capable ofmeasuring the amount ofabsorption from the sample. 4. A recorder [7]. The main component ofan FT-IR spectrometer is the Michelson interferometer. It contains a fixed mirror, a movable mirror, and a beamsplitter. The beamsplitter transmits halfthe incident radiation to a moving mirror and reflects the other halfto the fixed mirror. The two beams are reflected back to the beamsplitter, where they recombine. As the moving mirror is moved away from the beamsplitter, an optical path difference is generated. This is known as retardation. The optical retardation, in a Michelson 20 interferometer, is equal to twice the difference between the distance ofthe fixed mirror and the distance ofthe moving mirror from the beamsplitter. The two beams travel different distances before recombining. The frequency ofthe retardation and the position ofthe moving mirror are the basis for the pattern ofconstructive and destructive interferences that are generated. Thus, the intensity ofthe radiation varies in a complicated pattern, as a function ofmirror movement. The output beam is a result of modulation by the interferometer. The modulated output beam is directed through the sample compartment to the detector. At the detector, it generates a continuous electrical signal called an interferogram. A beam from the He-Ne laser is also passed through the interferometer to its own detector, which generates a reference signal. This enables the spectrometer electronics to sample the interferogram at precise intervals. The computer converts the interferogram into a single-beam spectrum via a Fourier transform. The FTS-40 (BIO-RAD Cambridge, MA) first collects the spectrum ofthe source (background spectrum) and stores it. The single beam spectrum ofthe source, modified by the absorption due to the sample, is collected and compared to the background spectrum. The difference between these gives the desired absorption spectrum. 21 Chapter Two Project Information Arson fires are a major problem for the fire investigator. Often, these fires are accelerated by flammable materials. Organic, petroleum based, non-water soluble solvents, such as gasoline, can suprisingly be detected at very low concentrations by several well-developed analytical techniques such as GCIMS. However, water soluble solvents, such as methanol, ethanol, and isopropanol, have proven to be difficult to analyze. Not only does the water used to extinguish the fire wash away the accelerant by convection, it also limits detectability ofthe accelerant by dissolution. Dilution standards were run on each ofthe accelerants tested. These standards were 5.0% v/v accelerant:water, 2.0%, 1.0%, and 0.5%. These solutions gave an idea of how little accelerant is needed to be in the sample and still be able to be detected. The first stage ofcontrolled burnings were done on carpet samples. The carpet samples were discontinued floor samples that were donated by Home Carpet in Boardman, Ohio. A known amount ofaccelerant (1 OmL) was poured on the carpet, burned for various times ranging from 30 sec. to 180 sec. , and extinguished with a known amount ofwater (lOOmL). The water was collected from the carpet and run-offthat went into a collection trough. An IR spectrum was then obtained for the water samples. In each trial, the accelerant being tested was detectable. The second stage ofburning was also done on carpet samples. The burning was the same as mentioned above. In this stage, The samples were left out overnight and then collected into evidence cans (Tom's Automotive Hubbard,OH). They were tested a 22 week later for the presence ofan accelerant. This stage was then repeated to ensure that the results were reproducible. The third stage ofburning was done in surroundings that were similar to those that would actually be encountered in the field. In coordination with the Youngstown Fire Department and the Arson Bureau, controlled bums were performed at a vacant house (320 E. Lucius Ave. Youngstown, OH) that was scheduled for demolition. Twenty-one samples were taken. However, only sixteen were tested because a few ofthe samples were too bulky to be tested. The samples were collected using the techniques that are consistent with those used by the Youngstown Arson Bureau. The samples were placed in evidence cans (Tom's Automotive Hubbard, OH) which are the same used by the Arson Bureau. The lids were hammered on and a piece oftape was placed over the lid to ensure that the cans were not opened at any time before analysis. The samples were tested for the presence of the selected accelerant. For the interpretation ofthe results, only the fingerprint region (1800-750 cm- I ) was used. The results ofthese experiments will be significant because there is currently no method that is able to accurately detect water soluble accelerants. 23 Chapter Three Materials and Methods Materials Analytical grade methanol, ethanol, and isopropyl alcohol (Fisher Scientific, Fair Lawn, NJ) were used as accelerants. Carpet samples were discontinued floor samples (donated by Home Carpet Boardman, Ohio). A BIO-RAD FTS 40 spectrophotometer (BIO-RAD Cambridge, MA) was used to perform the analysis (see Figure 3.1). The ATR cell used is a Harrick Scientific Corp. ZnSe nine reflection prism liquid cell with a 45° incident angle (see Figure 3.2). Methods Attenuated Total Reflectance Spectroscopy is a very effective way to perform analysis on aqueous samples. A background can be obtained using water and then subtracted out ofthe sample spectra. There is very little time required to prepare samples for analysis. Samples can be collected and run without having to do the organic solvent extraction step that is necessary with other techniques, such as GC-MS. There is good contact between the sample and the crystal which gives quality spectra that are highly reproducible. A detailed list ofparameters used for analysis is listed as Table 4.1. FTS 40 Components The FTS 40 spectrophotometer (BIO-RAD Cambridge, MA) is fitted with purge shutters, which are on both sides ofthe sample compartment. When the knobs are out, 24 -------- .----...... SAMPLE. C.OMPARTMENT -----. \ \ oINi£R~ER.OMti£R. Al .------- .- ~--n- ,-- - I _ 0 BE BOA-R,O ~EL.E:C"IClONICS I I I I I I I I I I I: I . ! I~ I I !-- - --- - - -i:;~~;:::;;;::;;;:~;:~=- CE:iEC,iOR. t Figure 3.1: FTS-40 Optical Schematics INi£R~ER.OMti£R. ~--n- ~ ~ :;~~;:::;;;::;;;:~;:~=- INi£R~ER.OMti£R. ~--n- ~ ~ :;~~;:::;;;::;;;:~;:~=- ~I__-r lHTRODUC11OH FTi"f1NG ~_---~eSERYOIR 25 Figure 3.2: ATR Cell Reservoir ~I ~_---~eSERYOIR ~I ~_---~eSERYOIR 26 the shutters isolate the sample compartment preventing water and carbon dioxide from entering the interferometer compartment. The sample compartment has a hinged cover to isolate it from the environment. The FTS 40 (BIO-RAD Cambridge, MA) uses a He-Ne laser operating in the visible region at a wavelength of632.8 nm. It has an aperture wheel which has four holes ofdifferent diameter. Each hole restricts the beam differently so that the desired light intensity is reached. In this study, the open aperture setting was used to allow the unrestricted beam to pass through the crystal. The FTS 40 (BIO-RAD Cambridge, MA) uses a MCT high-sensitivity narrow band detector that it cooled with liquid N 2 . The IR source requires a supply ofcooling water. A minimum flow of 1 Llmin, with an inlet pressure no higher than 25 psi and temperature no higher than 20°C are required. The FTS-40 (BIO-RAD Cambridge, MA) has a Michelson interferometer (Figure 3.3). It contains a fixed mirror, a moveable mirror, and a beamsplitter. The beamsplitter sends halfofthe incident radiation to the moving mirror, and the other halfto the fixed mirror. The two beams are reflected back to the beamsplitter and recombined. The resolution is able to be varied with the higher resolution giving more detailed spectra. Absorbing Materials Accelerants are usually poured on carpets, couches, chairs, mattresses, hardwood floors, etc. The physical and chemical characteristics ofthese absorbing materials play an important role in the accelerant behavior during the fire. Some ofthe important absorbing material characteristics include: 1. Absorb flammable liquids. Carpets, chairs, mattresses, and hardwood floors are examples of 27 Detector \ Beamsplitter Moving Mirror .' nhc::=·::··::;···:..::J?7~~1~:"~'curceof ......... U ---.. Radiation fixed Mirror Figure 3.3: Michelson Interferometer ::J?7~~1~:"~'curce::J?7~~1~:"~'curce 28 good absorbing materials since liquids poured on them soak in. 2. Retain absorbed liquids. Materials that possess little ability to absorb, will have the accelerant pass through to something that absorbs and retains them. 3. Contaminate the absorbed liquid. The most commonly used accelerants are composed of hydrocarbons. The hydrocarbons will dissolve and mix with many synthetic materials such as carpet rubber backing. The mixture will interfere with patterns produced by IR analysis. 4. React to heat and fire. The exposed surface ofmany absorbing materials will melt and/or char upon exposure to heat and flame. When a liquid is poured on a porous or a semiporous material, a certain amount of the accelerant will soak into that material. The amount depends on the factors discussed above. All ofthese factors were taken into account when deciding which absorbing materials to use. However, a majority ofthe time accelerants are poured onto carpeting. With this in mind, it was decided that carpet samples would be the best absorbing material to use for the project. It has the ability to absorb and retain a large quantity ofliquid. It also will melt and bubble when burned to form air bubble pockets that will retain the accelerant so that it is easier to detect in analysis. 29 Chapter Four Procedures and Results A. Preparing Concentration Gradient Curve Solutions Initially, a 200 mL 5.0% methano1:95.0% water v:v solution was prepared. From this solution, a 2.0%, 1.0%, and 0.5% v:v solution was prepared. Then, a 200 mL 5.0% ethanol:95.0% water v:v solution was prepared. From this solution, a 2.0%, 1.0%, and 0.5% solution was prepared. Then, a 200 mL 5.0% isopropyl alcoho1:95.0% water v:v solution was prepared. From this solution, a 2.0%, 1.0%, and 0.5% solution was prepared. Obtaining Spectra The procedure for obtaining spectra remained the same throughout this project. Any exceptions will be mentioned when necessary. The procedure will be discussed thoroughly in this section. The parameters used to obtain spectra were as follows: Table 4.1: List ofParameters Speed Filter UDR Resolution Aperture Sensitivity IR source Detector Beamsplitter Apodization Number ofscans 5kHz 4.5 kHz 2 8 open 1 ceramiC Mid-IRDTGS KBr triangular 64 30 The cell reservoir was filled via a needle and syringe. A background spectrum of deionized water was obtained each day that analysis was to be performed. The cell reservoir was flushed twice with the desired solution before being filled for analysis. The cell reservoir was then placed in the sample compartment. The instrument was purged for I min. before analysis was initiated. B. Initial Burning of Carpet Samples The carpet samples were trimmed to approximately 4" by 6". To each sample, 10 mL ofmethanol was poured over the carpet and ignited. The burn times were varied (30 sec., 60 sec., 90 sec., 120 sec., and 180 sec.) so that % recovery versus time could be determined. The samples were extinguished with 100 mL ofwater. The water that stayed in the carpet was collected, as was the water than ran-offin to the overflow trough. A sample was then burned with 10 mL m-xylene so that a burn comparison could be made. Also, a 10% v:v methanol:water solution was prepared that would represent the maximum recovery possible. The spectra were obtained using the procedure previously discussed. Another set ofcarpet samples were then burned using ethanol employing the same procedure as with methanol with one exception: a 150 sec. bum time was included. A third set ofcarpet samples were burned with isopropyl alcohol using the same procedure as ethanol. 31 C. Burning of Carpet Samples - Second Stage In this stage, more realistic conditions are trying to be obtained. The procedure for burning the samples is the same as in section C. However, the collection procedure was changed. Instead ofcollecting the water samples immediately, the carpet was placed on aluminum foil overnight. After 24 hrs, the carpet samples were placed in evidence cans and stored for one week. When it was time for analysis, an extraction step had to be performed. The extraction step consisted ofpouring 15 mL ofwater on the carpet sample and squeezing the water out. The spectra were obtained using the procedure previously described. Burning of Samples - Third Stage This stage ofburning was the most important. In this stage, the conditions are as realistic as we can get without actually setting an arson fire. The bums were done at a vacant house (320 E. Lucius Ave. Youngstown, OR) in coordination with the Youngstown Fire Department and the Arson Bureau. The accelerants used were methanol, ethanol, isopropyl alcohol, rubbing alcohol and two unknowns prepared by Dr. Curtin. The first bum was in the upper left bedroom. An area was cleared ofall debris. A pile ofclothes were placed in the center ofthe cleared area. Approximately 300 mL of methanol was poured over the pile and ignited. The pile was burned for 25-30 sec. before being extinguished by firefighters with an minimum amount ofwater. Samples collected were sample 1- sweater, sample 2 - socks, sample 3 - carpet, and sample 4 - carpet. 32 The second bum was also in the upper left bedroom. It was done the same way as the first bum with the accelerant being unknown A. The samples collected were sample 5 - free standing water, sample 6 - clothes, sample 7 - cloth, and sample 8 - carpet. The third bum was in the upper right bedroom. The pile burned included clothes and part ofa curtain. Approximately 300 mL ofcommercially available rubbing (isopropyl) alcohol was poured on the pile and ignited. It was burned for 1 min. before being extinguishing with an minimum amount ofwater. The samples collected were sample 10 - cloth, sample 11 - curtains, and sample 12 - carpet. The fourth bum was in the upper right bedroom. The area was cleared and a pile ofclothes were placed in the center. Approximately 300 mL ofunknown B was poured on the pile and ignited. It was burned for 1 min. before being extinguished with an minimum amount ofwater. The samples collected were sample 13 - clothes and sample 14 - clothes. The fifth bum was done on the front porch. Approximately 300 mL ofunknown B were poured on a couch cushion and ignited. It was burned for 90 sec. before being extinguished. The samples collected were samples 15,16, and 17 - cushion material and padding. Analysis ofthese samples required an extraction step consisting ofpouring 30 mL water on the sample and squeezing it out. The final bum was done in the third bedroom upstairs. The area was cleared of all debris. Approximately 300 mL ofan ethanol and isopropyl alcohol mixture was poured on the carpet and ignited. It was burned for 120 sec. before being extinguished. The samples collected were sample 18 - free standing water and samples 19-22 carpeting. 33 The spectra were obtained using the methods and parameters previously described with one exception: the number ofscans was 128. The number ofscans was changed to 128 instead of64 so that the spectra would have less noise. 34 Results The results from the concentration gradient curves are the first presented. For methanol, there were peaks at 781 cm-\ 824 cm-\ 1017 em-t, 1113 em-I, 1472 em-I, 1506 em-I, 1558 em-I, and 1686 em-I. For ethanol, there were peaks at 877 em-I, 1045 em-t, 1086 em-t, 1418 cm-\ and 1454 em-I. For isopropyl alcohol, there were peaks at 761 cm 1,945 em-I, 1107 em-I, 1126 em-I, 1164 em-t, and 1558 em-I. For acetone, there were peaks at 770 cm-\ 789 em-I, 1094 cm-\ 1239 em-t, 1371 em-I, 1423 em-I, and 1699 em-t. The spectra have been included (see Figures 4.1-4.16). The wavenumbers in bold type are the characteristic peaks that were used to distinguish between the accelerants. For methanol, the peak at 1017 cm- I represents the C-O stretch. For ethanol, the peaks at 1045 cm- I and 1086 cm- I represents the C-O stretch. For isopropanol, the peaks at 1107 cm-\ 1126 em-I, and 1164 cm- I represent a C-O stretch. For acetone, the peak at 1699 cm- I represents the ketone C=O stretch. The data for these characteristic peaks of each solution have been tabulated (see Table 4.2) and plotted (see Figures 4.17-4.24). Representative spectra from section C have been included (see Figures 4.25-4.32). The results ofsection C have been tabulated (see Table 4.3) and plotted (see Figures 4.33-4.42). Only the characteristic peaks from the concentration gradient curves were looked at. Representative spectra have been included from section D (see Figures 4.43 4.45). The results ofthis section were tabulated (see Table 4.4) and plotted (see Figure 4.46). 35 Results - Third Stage Burns Sample 1 was too bulky and was unable to be analyzed. Samples 2-4 show peaks at 1017 cm- I and 1112 cm- I (see Figure 4.47-4.49). Sample 5, accelerant unknown A, shows a peak at 1016 cm- I and at 1112 cm- I (see Figure 4.50) Sample 8, accelerant unknown A, also shows a peak at 1017 cm- I and 1112 cm- I (see Figure 4.51). Samples 6-7 were too bulky to be analyzed. Sample 10, accelerant rubbing (isopropyl) alcohol, had peaks at 1385 em-I, 1165 cm-t, and 1128 em-I. Sample 11 showed peaks at 1698 cm- I and 1129 em-I. Sample 12 had no peaks that corresponded to isopropyl alcohol (see Figures 4.52-4.54). Samples 15, accelerant unknown B, showed peaks at 790 em-I, 1016 em-I, 1101 em-I, 1375 cm-t, and 1453 em-I. Sample 16 had peaks at 748 cm-t, 1107 em-I, and 1630 em-I. Sample 17 had peaks at 790 cm-t, 846 cm-t, and 1631 cm- I (see Figures 4.55-4.57). Sample 18 consisted ofseveral peaks. The main peaks were at 1045 em-I, 1087 em-I, 1126 cm- I and 1164 em-I. Samples 19-22 shows the same peaks as sample 18 (see Figures 4.58-4.62). Figure 4.1: 5.00/0 v/v methanol/water .2 0--1--~-'~'--""'- 10001500 2500 2000 Wavenumber (cm-1 ) 30003500 .25 -.05-- .15 Q) () c co .1..a L- a (/) ..a « .05- LoU (j\ 0-- ~ C Figure 4.4: 0.50/0 v/v methanol/water .12 .1- .08 ~ c: ro -e o (/) ~ .06 .04- .02- O-i /,,,._/JY'''' . prJf -.02-Lf~~~ ___~- -._.-~4)11r.,!,,",",_~I~ \/1--~_./J~.- 3500 3000 2000 Wavenumber (cm-1 ) 1500 1000 W \D ~ ----------- ----~-------~----I~---------------l----~~~--T----.----~~I--. -- ------~___r------- 2500 ---~-------~----I~---------------l----~~~--T----~~I-----~___r------- Figure 4.5: 5.00/0 v/v ethanol/water ~--_..--'~------'-'-'~-~._-----_._--------,- .... - - --_._--------- ---_..._-_._._-_._..._-_..._--_._.,~-,_._----_._.._._-'-,','- .25- .2- .15- ~ c co € .1- o (/) ~ .05 O-j }J\ j'-..,.J\rV'--.A.~ .-----~·~~-~~'\,~//f/-\\,./ -v/~j .~ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 ~ o --__ _._ __ ._ .._-~.__.- -----------,-_. ._--- --------------,-_--_._.,~-,_._----_.__._-._,',._•.......•_-------_._.__ ..- Q) () .Q \- 0 .Q « _-~._--_._.,~-,_._----_._ Figure 4.6: 2.0% v/v ethanol/water .2 .15 ~ c:: ro ..0 .1 '- o f/) ~ .05 o+""",~/~/~,~,~,\,}r{'c",,-----' " --,--"",,~r-~ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 01::>0 ..... ~ ----------1--------,----- - - ---- -----,------- -----T-------------l --- --- ----1---- ~ Figure 4.7: 1.00/0 v/v ethanol/water -.02- l______ --- -, ----------------,----------~---- --------,----- -----T------- .14 .08 .. ,-_.,~~./ / V-Ot.- ---- / ""\ Mlir--~/ /),~"J\)," .1- o .04 .06- .12- .02- ~ c: ro ..c l- o C/) ~ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1 ~ tv ------- ... -1----------1 ... ------T- - ·---1-- - ·-T------- - ... 1 1000 Figure 4.8: 0.50/0 v/v ethanol/water .12 .1 .08 Q) .06 u c co .0 ~ .04 0 f/) .0 « .02 0 -.02- -.04- //r"", !fr/' )'~-~} ~_,~A,.-.---_.'-"~~._~_.~f- 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 ol::> W ~ ~ o o 44 ,0 LO 1 0 I (") I i 0 .......... '0"' 10 I IN § I ':' I~ , ::J '0 C i 0 Q) 11.0 > N as $ 10 LO I~ I ,0 10 rO I~ 1.0 o I o \ / \ , , ) \ \ \ I 1.0 o .. , • I .. I as I ~I '0 ' .c o u ii ~ e a. o fn - ~ ote Q .,; •• 0) "fi- e :::Sl .e» I u:1 '--------,----------,----------r--------'--~~ aoueqJosqv ~ I~ ~I ~ '--------,----------,----------r--------'--~~ ~ ~I ~ '--------,----------,----------r--------'--~~ Figure 4.10: 2.0% v/v isopropyl alcohol/water .1 ~ c ro .0 ~ o C/) .0 « .05~ O~/~~flr\",) v~vj/-~ _f~J-~"/ -.05 ~~~___,.~A~ _~f'vr___J/ 3500 3000 2500 Wavenumber (cm-1 ) 1500 1000 >l:>o U1 ~ ro ~ C/) .05~ - ----------1------ -- ------1----------~---------~---------~--~-r-------------- --- r- ----------- 2000 .05~ ~ c co .n L- a U) .n « .1 .05- o -.05 Figure 4. 11: 1.00/0 v/v isopropyl alcohol/water /~JVA"jl'~"fYr/--_/~--..-._4"J(ORf'v AJ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 ~ 0'\ ~~ Figure 4.12: 0.50/0 v/v isopropyl alcohol/water .12 .1 .08- .06 Q) u c ro .04 .0 I- 0 en .0 .02 « 0 -.02-- -.04 -.06 .~."MI~V/~- / rtJ' .-~-j~ ~~V // /~""'\r/Afc/ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 II:» -....J 500 Figure 4.13: 5.0 % v:v acetone/water .15- .1 ~ c: .e .05 ~ ~ o -.05 f~~J '1 ,j,~_~A~''_~~ ",j~ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 ,j:>. 00 .1 ~ c: o rJ) ~ ---~-,-~-_.. I ---~I---------------r-~_··-~-~--T-~----,---- .- ~ ---~-,-~-_-~I---------------r-~_··-~-~--T-~-- Figure 4.14: 2.0 % v:v acetone/water .1~ l\ .08---i~ I .06- ~ c: <'Cl 04 .0 I- 0 C/) .0 « .02 o -.02 --p, " \/ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 ~ \D .1 .08 Q) c.> .04 I --1---- -- ---- -------1-----1 -------~---- -r ----~---- Figure 4.15: 1.0 % v:v acetone/water .03 .02- -.03 .01 Q) u c co .c L- a en .c \ ,p\ « V~t~llA -.01 -.02 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 U1 o }~ " --------~~~I----------- - ------T -------------I--------I~------- r----1---~~~I--------------------I--------I~-- .04 .03 .02···· .01 Figure 4.16: 0.5 % v:v acetone/water ~ c 0 ~ L.- a ~-.01 ~ -.02 -.03 -.04 AJj -.05- ..---~.------r---..... ··---·---i-----·---······--·T-----------T-- -------------1------------·-----I-~---·--··- 3500 3000 2500 2000 1500 1000 Wavenumber (cm-1 ) U1 ~ .- .. _._._.._._.~-----------_.-~'----, -- ------------------------. --- --~-------.'.'.'. _.._---_._--------------------- -- -..- _..••.. ---,------------- ~ CI:l .0 ~ ~ .....---~.------r----------·-----I-~---·--··- _._.~-----------_.-~'---~------- ~ ~ ---~.------r--- Table 4.2: Data from concentration gradient solutions 52 % methanol\water 0.5 1 2 5 % ethanol\water 5 2 1 0.5 % isopropyl\water 0.5 1 2 5 % acetone\water 0.5 1 2 5 1016cm-1 0.0354 0.0614 0.105 0.2361 1086 cm-1 1045 cm-1 0.0705 0.185 0.043 0.088 0.038 0.0595 0.0249 0.0336 1164 cm-1 1126 cm-1 11 07 cm-1 0.0238 0.0308 0.0306 0.0273 0.0378 0.0341 0.0379 0.0555 0.0464 0.0528 0.0896 0.0647 1699 cm-1 1239 cm-1 0.00854 0.00135 0.0275 0.00909 0.058 0.03149 0.1496 0.0936 Figure 4.17: Methanol Standards 1016 cm-1 0.25 T'------------------------------- i y = 0.0442x + 0.0155 R 2 =0.9996 0.2t--------------------~~~~-------~~I 0.15 I~ ~I Q) U l: III € o III .c III 0.1 I .-- ,. I 0.05 I ,r I Ln W 65432 o I " ii' o % V:V methanol:water -r--------------------------------------------....., +------------------------------------,~---------------1 +----------------------~!!:.....----------------------l 0.1 -I-~f1!:---------------------------------1 +----------::;;l,tIL-------------------------------------------I O-!-------"""T'"-------,...-------....,....---------r-------........-------1 v:v 0.2 0.18 0.16 0.14 0.12 m 0 ~ m ~ 0.1~ 0 0 ~ m 0.00 0.00 O.~ O.~ Figure 4.18: Ethanol standards -1045cm-1 y =0.0328x + 0.0219 R 2 =0.9965 ~ / / / / / / / r m o o 2 3 % v/v ethanol:water 4 5 6 lJ1 *'" Gl [:; III .c L- a III .c III 0.08 0.06 0.04 0.02 ethanal:water 0.08 0.07 0.06 0.05 CI) (J C nl -e 0.04 0 III .c nl 0.03 0.02 0.01 Figure 4.19: Ethanol standards - 1086cm-1 y =0.0094x + 0.0242 R 2 =0.9676 ~ ~ ~ ~ ~ , • o o 2 3 % v/v ethanol:water 4 5 6 lT1 lT1 .c .. -- ~ ~ ~ ~ - ~ ~ ~ ~ Figure 4.20: Isopropyl Alcohol Standards 1107 cm-1 y = 0.0076x + 0.0279 R 2 =0.9783 0.06rl---------------------------:::;;ii~~=-------~~1 0.07 f'--~=------------------------------. , 0.051~1 ~0.04 I~ c~ ~I L o III .c " -.05 f,,\__/_ ~~ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1 1000 '" w ~ -----1---- --------I---------r -----------------~----------------T----- --------1--------- 1500 --------------~----------------T Figure 4.28: 10mL eth./100mL water· 90sec.• retrieved from carpet .2- .15 ~ c ro .0 '- o en .0 « .1 .05 0- -.05-j~V~\ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 ~,j J 1000 "" "'" ~ --1-------------------~I~~----~-r------ ----T----- ---~-~T---------- --T------- -----~-~T------ Figure 4.29: 10mL eth./100mL water -150sec. - retrieved from carpet .04- .02 ~ c co -e o (/) .0 « 0- -.02 _.04_,-~r -.06 ~/'A ~/' ~~-~.----~ 11 ///~-// J ~///// -_.~~ ~yJ-~/' 3500 3000 2500 2000 Wavenumber (cm-1 ) 1 1000 0'\ U1 ~ ----------I--~-~---I-~--I------I--------T-- 1500 --1------ --- '" '" 100015002500 2000 Wavenumber (cm-1 ) 3000 / 3500 .06 Figure 4.30: 10mL prpyl/100mL water· 30sec.· retrieved from carpet ----_._._._._-_.~------------------------------------------------_.,------,".------------------------------------------_..- .08-' .04 -.06- ~ .02 c l'(l .0 I- 0 0 en .0 « -.02 -.04 -----.-.-.-.---.~-------------------------------------------------..------.... -------------------------------------------.. ...-.--- .08- ~ l'(l ,~ ~, -----T---------- - --------1------- --------------T---- ---- ---I -------T---- ---- --1----- ~ ~, Figure 4.31: 10mL prpyl/100mL water· 90sec.· retrieved from carpet .06 .04 .02- ~ c co -e 0 o (/J !t -.02 -.04--v -.06 3500 3000 // /\.1// i////~ 2500 2000 Wavenumber (cm-1 ) 1500 1000 0'\ -...J ~ .0 "- 0 .0 <: -.04-~J ·---r---------T--------- -- - ------r----------r----- -- -- - --------,--------- - -----T- ~ ~J .06 .04- Figure 4.32: 10mL prpyll100mL water· 150sec.• retrieved from carpet .02- (]) u c ro ..c l- n 0 C/) ..c « -.02 -.04 -.06 ~_~_~,J /~-~ // //// ~ //-'//~/ // V 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 ~ 00 0 I- ---------------~~---~----r-----.. -- ... --·~--T--~-~~-~-··-------·------T---~---------T-~-----·~--T--~-~~-~-··-------·------T---~---------T-~---- 69 Table 4.3 Stage One Burns Methanol 2/12/98 1016 cm-1 1016 cm-1 Burn time absorbance % recovery o 0.434 100 25 0.4561 105 60 0.3082 71 90 0.2612 60 120 0.4063 94 180 0.0877 20 Ethanol 2/19/98 1016 cm-1 1016 cm-1 Burn time absorbance % recovery o 0.434 100 25 0.4561 105 60 0.3082 71 90 0.2612 60 180 0.0877 20 1085 cm-1 1085 cm-1 1045 cm-1 1045 cm-1 Burn time absorbance % recovery absorbance % recovery o 0.0948 100 0.3112 100 30 0.0856 90 0.3111 100 60 0.0636 67 0.2455 79 90 0.0646 68 0.2392 77 120 0.0456 48 0.165 53 150 0.0161 17 0.0769 25 180 0.0263 28 0.0915 29 Isopropanol 2/19/98 1164cm-1 1164cm-1 1126cm-1 1126cm-1 1107cm-1 1107cm-1 Burn Time absorbance % recovery absorbance % recovery absorbance % recovery o 0.0739 100 0.1303 100 0.0813 100 30 0.0826 112 0.1423 109 0.0877 108 60 0.0656 89 0.1107 85 0.0669 82 90 0.0541 73 0.0811 62 0.0504 62 120 0.0528 71 0.0753 58 150 0.0374 51 0.0429 33 Unable to detect 180 0.0377 51 0.0416 32 0.5 0.45 0.4 0.35 0.3 Q) 0 s:::: a::s of! 0.25 0 III .c a::s 0.2 0.15 0.1 0.05 Figure 4.33: Methanol Burn 2/12 ...... • ~ y = -0.0021x + 0.4573 R 2 = 0.9569 ~ ~ ~ ~ ~ ~ "'- o o 20 40 60 80 100 burn time (seconds) 120 140 160 180 200 -...J o .c L- ~~ Figure 4.34: Methanol Burn 2/12 120 I I y =-0.4811x + 105.36 R 2 = 0.9574 100 • .......... I MI~I 20r----------------------------------~::~~-----J ~I 40 I ............... I ~ \I) > 8 60 I l!! ':!e. o ........... I -....J .... 200180160140120100 burn time (seconds) 80604020 o , iii i , o -,.....--------------------------------------------, --"'"""'~--------------------------------------------I 80 +-------------""lolo;;;;;:"------------------------------------I +--------------------------------""""""=,------------------1 ~ 60+--------------------~'=_----------------------~ 0 60 u e ~ 0 40 20 I I -...J W 200180160140120100 time (seconds) 80604020 oIiiiii i I o 120,.---------------------------------------------, +----=-.o;;;;:---+---------------------------------------l ~ ~ +----------------------------------------------1 O-!-----,------,-----,------...,.-----,-------,.----.,..----...,-----.,-------l ~ ~ Figure 4.37: Ethanol: Absorbance(1085.7cm- 1 ) vs. time 0.12 i i y =-0.0004x + 0.0955 R 2 =0.9216 0.1 I I 0.08 Ql U C ra of 0.06 0 CIl .c « 0.04 0.02 I """"""-= I • -...J ~ 200180160140120100 Time (seconds) 80604020 o I iff iii i I o 0.12,..--------------------------------------------..., +-----------------------------------------------1 .c ... +---------------------------------------"""'0...,.--------1 0+----......,..----,...-----,--------,.----.,.------,------r-------,....------,r--------4 Figure 4.38: Ethanol: %recovery(1085.7cm- 1 )vs. time 120 I I y =-0.4536x + 100.54 R 2 = 0.9209 100....... I 40 I~I MI~I ~ ~ 8 60 I E';!.. ........-.. I • 20 I ............. I • -..] U1 200180160140120100 Time (seconds) 80604020 o , iii iii I o 120,..---------------------------------------------, 100 ....;:---------------------------------------------1 80 +-----------"'''''''''"=:------------------------------------1 +------------------------------"""'0;;::---------------1 ~ ~ o 60+-----------------------"""'"O;;;::-------------------------j ';!.. +--------------------------------------------=""""";;o------j O+-----.,..----,.------,--------.,r------,.....-----r----.,.------,------r-----j Figure 4.39: Isopropanol Burn 2/19 1126 cm-1 0.16 i i y =-0.0006x + 0.1428 R 2 =0.9274 0.14 I.............. - I 0.12 I .............. I 0.1 G> () C III -e 0.08 0 Ul .c III 0.06 0.04 I .............. • I 0.02 I I -.J 0'1 200180160140120100 burn time (seconds) 80604020 o , Iii , o 0.16,....--------------------------------------------..., +"'''00;;;;:,-------==------------------------------------------1 +-------------'''"''''''';;;::---------------------------------------1 .c ... 0.04 +------------------------------------------"'........=--"--------1 +---------------------------------------------1 0+----"""T"""------r-----r__---~---"""T"""---___._----r__---...,.._---__r_-----1 Figure 4.40: Isopropanol Burn 2/19 1126 cm-1 120 i i y =-0.456x + 109.46 R 2 = 0.9273 • 100.~I 80 ~ • Q) > 60 I 0 0 Q) ... ~ 0 40 • 20 -..J -..J 200180160140120100 burn time (seconds) 80604020 o , iii iii , o 120,--------------------------------------------...., 100 _-----""""0;;;;;::----------------------------------------1 ~ 60 ~ O+-------.-----.,------,-------,-----r-------,.----...,.----......,..----.,....-----I ~ 0.09 0.08 0.07 0.06 B 0.05 c lIS .c ... 0 ~0.04 lIS 0.03 0.02 0.01 Figure 4.41: Isopropanol Burn 2/19 1164 cm-1 • ------- y =-0.0003x + 0.0804 R 2 =0.8931 ------- ~. ------- • ~ o o 20 40 60 80 100 burn time (seconds) 120 140 160 180 200 .....:J co C1I (J Ul .c ~. ~ ~. ~ Figure 4.42: Isopropanol Burn 2/19 1164 cm-1 120 I I • y =-0.3417x + 108.89 R 2 =0.8934 100 •~I 80 ~ Gl > 0 60 • 0 Gl ... ~ I 0 40 20 I I -....I \.0 200180160140120100 burn time (seconds) 80604020 o , iiI o 120,---------------------------------------------, _----~"'"=----------------------------------------l ~ ~ +-------------------------------------------------l O+----.....,.-------r-----.-----~---.....,....---___r----.,..._---_r_---___r_---~ ~ .05 Figure 4.43: 15mL water added to 30sec. burn sample from 3-2-98 ~ c ro ..c L.- a C/) ..c « o -.05- -.15~ /",/-'/"" /" " 'V,-y- ,,),-\ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 ex> o ~ -.15~ --1- -~I~--~---~----I-~----~~--~-~~-~r----~r-·----------I----------~----~I~--~--~----I-~----~~--~-~~-~r---~--------I----------~--- Figure 4.44: 15mL water added to 90sec. burn sample from 3-2-98 .02-r-----------~--~----------------~-~-~---------------------------------, o -.02- -.04 ~6 c -.0 ..... .02-------------~--~----------------~-~-~------------------------------ 0 Q) <..> -.06 0 (/) -.08 .n ~'" r""\ "'--c,,-// \ --- -------------I~--~---~--T---------I-~---~--r~-------- r- 1 ------ -- -- -----------I~--~---~--T---------I-~---~--r~----- -.05 ~ c co .0 L- a (J) .0 « Figure 4.45: 15mL water added to 240sec. bum sample from 3·2·98 o ~~~., /_ / ,eJ/\" J''I1 /~~_/', .. /J V/'V ./ " ;J/-- '\J ///- -,1-1 / ~,./' '" ,--'I~",;- _/J \ "'h P \ If -.15-- -.2- 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 (Xl N ~~~.. //\, / Jy/rJ"v ../ //~--- --- I . .--~-~ '\.J ;J/- ..-/ / Q) ./ U // C _./ co ; ..L.. -.1- 0 (J) ""'.//~~ ,",..,/ If "\ I \ -.15-- \ V (cm-1) (Xl .--~-~ Table 4.4: Ethanol Burns - Second~Stage Ethanol 3/2198 1045 cm-1 1045 cm-1 Burn Time absorbance % recovery o 0.3312 30 0.0153 60 0.0115 90 0.0059 120 0.0078 150 180 210 -0.0062 240 -0.0099 83 ~ Figure 4.46: Ethanol Burn - second Stage 3/2/98 1045 cm-1 0.02 I I y = -0.0001x + 0.0188 R 2 =0.9691 0.015 +- '" --1 0.01 -0.01 -f • ---I B 0.005 c ! .... 0 Ul .c 0tV t 50 100 150 ~OO 2503~0 -0.005 -0.015 I ! (Xl ~ burn time (seconds) ~------------------------------------------, 3 0200 ---·-----~---·--T--·--'·-·--·-~----"-·--·---·--···-------1------------------- • ------------'~---------------------------------------- o--------------,------------~--,--------------------------,---- 0.01- -0.005 ----------- ---------------------------------------------------------------------------------- • B I -0.015 .1-.. ----1 --------------,------------~--,--------------------------,---- Figure 4.47: sample 2 fron4/29· water that was in can .15 .1- ~ c ro -e o (/) ~.05- o -.05- '\ \~ ~.~.. ~" -~'-'~.'''"'' rV~.... f\ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 ex> lJ1 ~ -1--- -I---------~----------I----------T-----------T------- ~ .14 .12- .1- Figure 4.48: sample 3 fron4/29· water that was in can· carpet .08- ~ ffi .06 -e o C/) ~ .02 -.04- ~..~ -"\ vv~,,/p 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 ()j 0'\ a> () c ro 0 .0 Figure 4.49: sample 4 fron4/29· water that was in can· carpet .25-r---~------------.-------------------- --- - ----------- --------------- .2 .15 ~ C ttl ..0 '- o ~.1- « .05 o 3500 3000 2000 Wavenumber (cm-1 ) 1500 1000 co --.J .25----~-------- ---------------------------------------------- ~ --------T----- --- - ------T--------I----- -- --------T-- -------------------r- 2500 ----T------- ---~----- ~ "" 10001500 ~ \ """ '~ " ~ ~"-"'''''' ''\~v'~ V·V/ AA 2500 2000 Wavenumber (cm-1 ) 3000 sample 5 from 4/29 • free standing water· unknown AFigure 4.50: 1--- .12 .1- .08- Q) .06- t) c ro .0 .04- L- a fI) .0 « .02- 0- -.02 -.04 3500 Q) Q) 4/29· .1~ .08~ .02~ .1~ .08~ .02~ .12- .1'·· .08 ~ ffi .06 -e o ~.04 « .02 0- -.02- -.04 ....~ ~'~" ~ -'-.... ~ '--, ~ -"'-'"'-~ \r~~vj 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 (Xl 1.0 Figure 4.51: sample 8 from 4/29 • carpet· unknown A _______________"._.___~.,_ m ..~·_•..~__·__, ._.__~_._..__.. _.__. .. . • ._. ••_•._.__• .14 .1-·· ~ c co C/) ~ o· ···-··---··-··-1--·---··-····--1-···-··-·· . -······-T···----·-----r-----r-----··-r-···· ~·_ Figure 4.52: sample 10 from 4/29· water extracted· part a of cloth -.02 I -.04---j -.06 I ! /j \ fvv' AI -.08- ",,/J'/'JivJ ~ c:::: co -.1- .0 I- 0 C/) .0 « -.12- -.14 -.16 -.18- 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 1.0 o -.04 Q) () ------------~I~~~~----l---~-----~I-----------------I---------------------l-~----~----r------------- Figure 4.53: sample 11 from 4/29· water in can· curtains or-~-------~~----------~------~------------~----...-.-..---~-~---------------- -.02- -.04 -.06 Q) () c -.08 ctl .0 L.- a (J) -.1 .0 « -.12 -.16- -. 18-c--,---~------,---~------- -1- - ----- ---r--- l/V~'"'\/P} 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 \D ..... ~~"-~.---------_._-----.----...--~----------".-_.....,----------------.-~---...,...--.-".. ------- ..--.-----------------...--- ... -0'---'-"---'" f-rJ /v{'"\\/v! -.18----- -1------- --- ----1-------- -- ---r-----------r--- ------~----------r------- ~~"-~.---------_._------~----------"----------------.-~--- -----~----------r------- Figure 4.54: sample 12 from 4/29 • water that was in can· carpet ~ c co .0 L- a UJ .0 « .15 .1- .05 0- -.05 \\ ~, ,Jv'\I,A\~f~, 3500 3000 2000 Wavenumber (cm-1 ) 1500 1000 ~ tv ~ ~~~---~---~~--I----~~-~-~~~~T~~-----------r-----~--~~-----T---------T--------~-~~~l~- 2500 ~~~---~---~~--I----~~-~-~~~~T~~-----------r-----~--~~-----T---------T--------~-~~~l~- Figure 4.55: sample 15 from 4/29· water extracted· foam of cushion .05 o -.05---1 / \ 11 I V\('JV V V"' '\r, Q) c..> c co .0 -.1 I- 0 (/) .0 « -.15 1\ -.2---i ~\ -.25 L~_ 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 ~ w -.05- -.2 ~\ -----~... ..'--I~----'-"-T..---.....------I-~~~~---·-·~r·---- ······T--·· ---1'---' ~\ -----~..'--I~----'-"-T------I-~~~~---·-·~r·---- Figure 4.56: sample 16 from 4/29 • water extracted· foam of cushion -.05 -.1- ~ c ro -e o U) ~-.15- -.2 -.25 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 1.0 ,j:>. .....~--.-------~.~_~..~.~~~~--------~-~_.~.~~.~~~~~.~._ ..~_.--__~-_··_~----_··_·~···I ~ ~-.15· « ~ Figure 4.57: sample 17 from 4/29· water extracted· fabric of cushion -.02 -.04- -.06 ~ c ~-.08 L.- a ~ « -.1-- -.12 -.14- -.16- ~.J/ ~"/v'V 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 I.D U1 o ------- ------------------------------"---------------"---"--------------"--"-------------"- ------------ ~ co £l C/) ~ --,--"-""- ----1-" "- -----r----------------i- -----,-------- ----r----"-" ~ Figure 4.58: sample 18 from 4/29-water that was in can-free standing H20 100015002000 Wavenumber (cm-1 ) 30003500 .1 .15- ~I / c: .05m .n L- a en .n « 0 .1 HJDJ,I] ill l,! \ I \ j \r \ -.05 \D 0'\ ~ ffi o ---------,---------- -- -- -T-----------1-------- -----1------ ---- ----T-------- --------r-------- 2500 ~ Figure 4.59: sample 19 from 4/29-water that was in can-carpet .4~ ·3~ ~ c ro .0.2~ L- a en .0 « .1 o . )\t -. 1--L.-------- -..---'----~-----------~-f-----------------------r--------1--~--·--.----~---------..--.------.~--_--.----..-.J 3500 3000 2500 2000 Wavenumber (cm-1 ) 1500 1000 '-0 -...,J .3~ ~ .2~ 1-~~-~-~~--- --~~~-I~----~-~----~---T~---~-------~--T----------r~---- - ---- ----1-----------------1------- .2~ 1-~~-~-~~--- -~~~-I~----~-~---~---T~---~-------~--T----------r~- Figure 4.60: sample 20 from 4/29-water that was in can-carpet .3 .25 .2 B c: .15 co .0 r... 0 C/) ~ .11~ 05 3500 3000 2000 Wavenumber (cm-1 ) 1500 1000 ~ CXl c: co r... C/) .0 « .1 0 .05 0 ~~~-----~-~------T~~----------~~~~r--~-~----r··~~.~~--~--~----I---~-_.----~~---l-----~-----~._.~~-r~-------- 2500 ~~~-----~-~------T~~----------~~~~r--~~----r··~~.~~-~--~----I---~---~~---l-----~-----~_.~~~-------- Figure 4.61: sample 21 from 4/29-water that was in can-carpet .4 .3- ~ .2 c: m .0 I- 0 (/) .0L~-/~ « .1 0- -.1 3500 3000 2500 2000 Wavenumber (cm-1) 1500 1000 1.0 1.0 --------,. ----_._-_..__._------- --. - ----------~.. __._._------------_._._"._._--_._---_.__.--_..._------------ -.-----.- .-, . _.._--------- ~ c: I- 0 (/) ,,~ ~~"" .1 -----------T----- - --------r-----------------i----------T--- -- -----------,------------- ----1-- - ---------~ ~ Figure 4.62: sample 22 from 4/29-water that was in can-carpet rI .35-- .3- .25- Q) .2 u c co .0 .... .15-0 C/) .0 I /""''"' «-~ .1 .05 0- -.05 }~, - ---1---------- --1---------------~-------------r------------ - r------------ r------- 3500 3000 2500 2000 1500 1000 Wavenumber (cm-1 ) .... o o --""~ ------~----- 101 Chapter Five Discussion and Conclusions The concentration gradient curve was done so that an estimate ofdetection limits could be made and an approximate concentration ofunknown alcohols could be determined. The determination ofthe presence ofaccelerant were based on those. The percent recoveries were calculated based on the 9 % v:v solution which gives the maximum recoverable since 10 mL ofaccelerant and 100 mL ofwater were used. On the concentration gradient graphs, a linear regression line is shown with the line equation and the R 2 value. The R 2 values range from 0.9676 for ethanol at 1086 cm- 1 to 0.9996 for methanol. The graphs also show that the strength ofabsorption is directly related to the concentration. As can be seen in Table 4.3, the accelerant is still able to be detected after as long as 240 sec. ofburning. For each ofthese burns, the percent recovery was calculated. For the graphs corresponding to Table 4.3, a linear regression line is shown with the line equation and the R 2 value. The R 2 values range from 0.8931 for isopropanol at 1164 cm- 1 to 0.9569 for methanol. The R 2 values were not expected to be linear. There are several factors, such as intense heat, that have an effect on the accelerant. The R 2 value is given only to show that bum time is not the only factor acting upon the accelerant while burning. From the line equation, the detection limit can be calculated. For each accelerant, the limit would be around 240 seconds. Further inquiry in this project should look at extending burn times past this time to determine the exact limit. In Table 4.3, a second list ofvalues was included for methanol. After examining the 120 second point, it 102 was detennined to be an erroneous value and therefore dropped. The graph that is shown reflects this point not being included. It changed the R 2 value from .6497 to .9569. In these burns, only a few milliliters ofaccelerant were used. The typical arsonist will use a considerable amount ofaccelerant which would only make it easier to detect. Third Stage Burns The results ofthese burns were qualitative. The purpose behind these burns was to detennine ifthe accelerant could be detected. The peaks found in samples 2-4, 1017 cm- 1 and 1112 cm-t, correspond to those of the 5.0% v:v methanol spectrum. The peaks that were found in samples 5 and 8, 1017 cm- 1 and 1112 cm-t, corresponded to those ofmethanol. Therefore, unknown A was concluded to be methanol. However, when a 5.0% v:v solution ofunknown A was run, it shows the presence ofboth methanol and ethanol. Unknown A was 4: 1 methanol:ethanol. There is the loss ofthe ethanol peaks in the samples. Although the exact composition ofthe unknown accelerant was unable to be detennined, the presence ofan accelerant was still detected. Further inquiry into this project might look at the interaction ofmethanol and ethanol, both before and after being used as an accelerant. Samples 10 has two peaks, 1165 cm- 1 and 1128 cm-t, that correspond to isopropanol. Sample 11 has one peak, 1129 cm- 1 , that corresponds to isopropanol. Samples 15 had peaks at 1016 cm- 1 , 1375 cm-t, and 1453 cm- 1 that best corresponded to both methanol and acetone. Samples 16 and 17 did not have peaks that corresponded to any ofthe accelerants. Therefore, unknown B was concluded to be made 103 up ofacetone and methanol. Further analysis proved this conclusion correct. The fact that only one sample had peaks that corresponded shows the importance oftaking several samples from each area. Samples 18-22 each showed four peaks, 1045 cm-t, 1086 cm- I , 1126 cm- I , and 1164 cm- 1 that correspond to ethanol and isopropanol. There did not appear to be any effects ofinteraction between these accelerants which were encountered analyzing unknown A. The results ofthe vacant house bums are very promising. However, further inquiry into this project should deal with a larger scale bum. When the fires were extinguished, a minimal amount ofwater was used. Because ofthis, the detection limits were not tested. The Youngstown Arson Bureau is willing to help with any further bums that need to be done in this project. Time limitations could not allow for a full scale bum to be done in time to be included in this project. For the samples that were too bulky to be analyzed, a distillation method should be looked at. The evidence cans are sealed air tight. The cans could be heated to the boiling point ofthese low molecular weight alcohols and puncturing the lid ofthe can with the distillation apparatus to allow for the collection ofthe fraction coming off This can be diluted with water to give the desired volume to use in the cell reservoir. In this study, Attenuated Total Reflectance Spectroscopy was used as a way of analyzing arson accelerants that are water soluble. According to Bertsch in his article, few laboratories routinely test for accelerants that are not petroleum based and there is little evidence that current analytical methods are capable ofdealing with water-soluble accelerants. ATR has proven to be a method that is well suitable for the detection of 104 water-soluble accelerants. There is little time needed for sample preparation. The concentration gradient curve for each accelerant show that the detection limit is lower than this study has tried. Further inquiry into this project should look at finding the lowest concentration still detectable not only by the concentration gradient method but also by burning. The area ofarson investigation is constantly changing with the introduction of new technology. Most new methods still require a lot ofhuman time whether in sample preparation or analysis. Further inquiry into this project should look at expanding the methods proven here in to other materials, such as cloth or cushions. As technology makes it easier to detect petroleum based accelerants, the arsonist will find new ways to beat the system. Attenuated Total Reflectance Spectroscopy is a relatively cheap, easy method that is capable ofdealing with accelerants that most new methods are unable to deal with. 105 References 1. Bertsch, Wolfgang Anal. Chern. 1996,68/17, 541A. 2. Youngstown Fire Department Annual Report 1996. 3. Conley, RT. Infrared Spectroscopy; Allyn and Bacon: Boston, 1974, pg. 7-8. 4. BIO-RAD, Win-IR Instruction Manual pg.35. 5. Mirabella, F.M., Jf., Ed. Internal Reflection Spectroscopy: Theory and Applications; Dekker: New York, 1993, pg. 2. 6. Ingle, James D. and Crouch, Stanley R Spectrochemical Analysis; Prentice Hall: New Jersey, 1988. 7. Ibid, page 433. 8. Ibid, page 434. 9. Svehla; Ed. Analytical Infrared Spectroscopy; Elsevier Scientific Publishing, New York, 1976, page 15. 10. Slater, C.D. et al Infrared Spectroscopy; Willard Grant Press: Boston, 1974, page 15. 11. Svehla; ed. Op. Cit. Page 18. 12. Ibid page 23. 13. Ibid page 35. 14. Ibid page 39. 15. Ibid page 181. 16. Ibid page 211. 17. Ibid page 217. 106 18. Mirabella, F.M., Jf., Ed. Gp. Cit. page 2. 19. Ibid page 18-19. 20. Ibid page 45. 21. Harrick, N.J., Internal Reflection Spectroscopy; Harrick Scientific Corp.: New York, 1967, page 41. 22. Urban, Marek W., Attenuated Total Reflectance Spectroscopy ofPolymers; American Chemical Society: Washington, D.C., 1996, page 6. 23. Ibid page 12. 24. Ibid page 35. 25. Harrick, N.J.; Mirabella, F.M., Jf. Internal Reflection Spectroscopy: Review and Supplement; Harrick Scientific Corp.: New York, 1985, page 5. 26. Urban, Marek W. op. cit. page 9. 27. Harrick, N.J.; Mirabella, F.M., Jf. op. cit. page 9. 28. Harrick, N.J. op. cit. page 27. 29. Ibid page 30. 30. Ibid page 32. 31. Ibid page 35. 32. Ibid page 35-36. 33. Ibid page 44. 34. Ibid page 83. 35. Ibid page 68. 36. Harrick, N.J.; Mirabella, F.M., Jr.; op. cit. page 38. 37. Harrick, N.J.; op. cit. page 89. 38. Ibid page 20. 39. Slater, C.D. et a1 op. cit. page 18. 107