<%BANNER%>

Investigation of Select Energetic Materials by Differential Reflection Spectrometry

Permanent Link: http://ufdc.ufl.edu/UFE0021294/00001

Material Information

Title: Investigation of Select Energetic Materials by Differential Reflection Spectrometry
Physical Description: 1 online resource (208 p.)
Language: english
Creator: Fuller, Ann Marie
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The presence of explosive or energetic materials is prevalent in today?s world. Terrorists continue to target buildings and mass transit systems with explosive devices. The detection of these energetic materials is necessary to insure national security and welfare. Detection techniques such as X-ray scanners, Raman spectroscopy, Terahertz spectroscopy and ion mobility spectrometry are in current use or development; however, none of these are appropriate for all necessary applications. These techniques include. The present document provides an overview of the current detection techniques and describes a new technique for detecting energetic materials called differential reflection spectrometry (DRS). DRS essentially measures the optical absorption of energetic materials. The use of DRS has led to the discovery of previously unreported optical characteristics for some energetic compounds that are unique to the individual material. These optical characteristics consist of absorption shoulders between 270 and 420 nm, e.g. near 420 nm for 2, 4, 6 trinitrotoluene (TNT). In the presented research, the origin of the differential reflection spectra obtained was investigated using several techniques including UV-Visible spectrophotometry (transmission and reflection) and computer molecular modeling. Experimental DRS spectra of TNT, hexahydro-1,3,5 trinitro-1,3,5 triazine (RDX), octahydro 1,3,5,7-tetranitro- 1,3,5,6 tetrazocine (HMX), pentaerythritol tetranitrate (PETN), and 2, 4, 6, n-tetranitro-n-methylaniline (Tetryl) were taken and analyzed. From the experimental results and verification by molecular modeling, it was found that the absorption features observed in the redder region of the UV range (270-420 nm) are due to molecular orbital transitions in the nitro (NO2) groups of the measured explosives. These transitions only occur in specific conditions, such as high concentration solutions and solids, where the normally forbidden transitions are allowed. The unique optical characteristics of the energetic materials presented in this dissertation are observed only in the solid or relatively high concentrated states suggesting the interaction of several molecules. Therefore these absorption features are proposed to be due to a charge transfer self-complex. This phenomenon can be interpreted in the same manner as the accumulation of atoms and be modeled using quantum mechanics.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Ann Marie Fuller.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Hummel, R. E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2008-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021294:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021294/00001

Material Information

Title: Investigation of Select Energetic Materials by Differential Reflection Spectrometry
Physical Description: 1 online resource (208 p.)
Language: english
Creator: Fuller, Ann Marie
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The presence of explosive or energetic materials is prevalent in today?s world. Terrorists continue to target buildings and mass transit systems with explosive devices. The detection of these energetic materials is necessary to insure national security and welfare. Detection techniques such as X-ray scanners, Raman spectroscopy, Terahertz spectroscopy and ion mobility spectrometry are in current use or development; however, none of these are appropriate for all necessary applications. These techniques include. The present document provides an overview of the current detection techniques and describes a new technique for detecting energetic materials called differential reflection spectrometry (DRS). DRS essentially measures the optical absorption of energetic materials. The use of DRS has led to the discovery of previously unreported optical characteristics for some energetic compounds that are unique to the individual material. These optical characteristics consist of absorption shoulders between 270 and 420 nm, e.g. near 420 nm for 2, 4, 6 trinitrotoluene (TNT). In the presented research, the origin of the differential reflection spectra obtained was investigated using several techniques including UV-Visible spectrophotometry (transmission and reflection) and computer molecular modeling. Experimental DRS spectra of TNT, hexahydro-1,3,5 trinitro-1,3,5 triazine (RDX), octahydro 1,3,5,7-tetranitro- 1,3,5,6 tetrazocine (HMX), pentaerythritol tetranitrate (PETN), and 2, 4, 6, n-tetranitro-n-methylaniline (Tetryl) were taken and analyzed. From the experimental results and verification by molecular modeling, it was found that the absorption features observed in the redder region of the UV range (270-420 nm) are due to molecular orbital transitions in the nitro (NO2) groups of the measured explosives. These transitions only occur in specific conditions, such as high concentration solutions and solids, where the normally forbidden transitions are allowed. The unique optical characteristics of the energetic materials presented in this dissertation are observed only in the solid or relatively high concentrated states suggesting the interaction of several molecules. Therefore these absorption features are proposed to be due to a charge transfer self-complex. This phenomenon can be interpreted in the same manner as the accumulation of atoms and be modeled using quantum mechanics.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Ann Marie Fuller.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Hummel, R. E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2008-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021294:00001


This item has the following downloads:


Full Text

PAGE 1

1 INVESTIGATION OF SELECT ENERGETIC MATERIALS BY DIFFERENTIAL REFLECTION SPECTROMETRY By ANNA MARIE FULLER A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREME NTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Anna Marie Fuller

PAGE 3

3 T o my family and friends; we all got through this together and I could not have done it without you. I further dedicate this document to my biggest supporter Stephen Tedeschi who walked with me throughout my research and often carried me.

PAGE 4

4 ACKNOWLEDGMENTS I gratefully acknowledge my advising professor Dr. Rolf E. Hummel. Through his guidance, he has helped me grow into the best scientist I can be and has inspired me to never stop learning. I also thank my supervisory committee Dr. Paul Hol loway, Dr. Wolfgang Sigmund, Dr. Anthony Brennan, and Dr. Alex Angerhofer for their intellectual conversations throughout my research. Dr. Alex Angerhofer allowed me continual access to his UV Visible instruments as well as many meaningful conversations. I also acknowledge research group members past and present especially Thierry Dubroca, Claus Schoellhorn, and Maxime Lemaitre for their valuable contributions to my research. Bryce Devine, Dr. Hai Ping Cheng, and her student Sabri Alkis assisted in the D FT modeling for the reported research and their contribution is acknowledged. I acknowledge Dr. Kevin Powers and the Particle Engineering Research Center for allowing me use of their molecular modeling program. Dr. Jodie Johnson at the mass spectrometry lab in chemistry performed the chemical analysis of samples in the research as well as explained the results to me for which I am thankful. In addition, I thank my friends and fellow students for their support. And finally I acknowledge the enormous love and support given by my family. I would have never made it without them.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 8 LIST OF FIGURES ................................ ................................ ................................ ......................... 9 ABSTRACT ................................ ................................ ................................ ................................ ... 17 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 19 1.1 Introduction to the Presented Research ................................ ................................ ............ 19 1.1.1 Current Detection Techniques ................................ ................................ ................ 19 1.1.2 Novel Detection Method: Differential Reflection Spectrometry ........................... 20 1.2 Proposed Detection Techniques for Energetic Materials ................................ ................. 21 1.2.1 General Techniques ................................ ................................ ................................ 21 1.2.2 Optical Techniques ................................ ................................ ................................ 24 1.3 Concentration of the Presented Research ................................ ................................ ......... 28 1.4 Summary ................................ ................................ ................................ ........................... 29 2 BACKGROUND INFORMATION FOR ENERGETIC MATERIALS ............................... 31 2.1 Characteristics of Energetic Materials ................................ ................................ .............. 32 2.2 Chemistry of Energetic Materials ................................ ................................ ..................... 33 2.2.1 TNT ................................ ................................ ................................ ........................ 36 2.2.2 RDX ................................ ................................ ................................ ........................ 39 2.2.3 HMX ................................ ................................ ................................ ....................... 41 2.2.4 Tetryl ................................ ................................ ................................ ...................... 43 2.2.5 PETN ................................ ................................ ................................ ...................... 45 2.2.6 Nitroglyc erine ................................ ................................ ................................ ......... 46 2.3 Optical Properties of Energetic Materials ................................ ................................ ......... 47 2.3.1 Ultraviolet and Visible Range (180 750nm) ................................ .......................... 47 2.3.2 Infrared Properties (0.750 ................................ ................................ ...... 50 2.3.2.1 Fourier Transform Infrared Spectroscopy ................................ .................... 51 2.3.2.2 Raman Spectroscopy ................................ ................................ .................... 52 2.4 Summary ................................ ................................ ................................ ........................... 53 3 ELECTRONIC THEORY OF ENERGETIC MATERIALS ................................ ................. 54 3.1 Electronic Structure of Explosives ................................ ................................ ................... 54 3.2 Photon Interactions ................................ ................................ ................................ ........... 56 3.3 Summary ................................ ................................ ................................ ........................... 58

PAGE 6

6 4 EXPERIMENTAL PROCEDURE ................................ ................................ ......................... 59 4.1 Ultraviolet and Visible Techniques ................................ ................................ .................. 59 4.1.1 Differential Reflection Spectroscopy (DRS) ................................ .......................... 59 4.1.1.1 Instrument ................................ ................................ ................................ ..... 59 4.1.1.2 Theory and Interpretation ................................ ................................ ............. 61 4.1.1.3 Line Shape analysis ................................ ................................ ...................... 62 4.1.1.4 Sample Preparation and Operating Parameters ................................ ............ 64 4.1.2 Ultraviolet Visible Spectrophotometery ................................ ................................ 66 4.1.2.1 Transmission UV Visible spectrophotometry ................................ .............. 66 4.1.2.2 Reflectance UV Visible spectrophotometry ................................ ................ 67 4.1.2.3 Sample Preparation ................................ ................................ ...................... 67 4.1.3 Infrared Spectroscopy ................................ ................................ ............................. 68 4.2 Chemical and Crystalline Structure Analysis ................................ ................................ ... 69 4.2.1 High Performance Liquid Chromatography and Mass Spectrometry (HPLC/MS) ................................ ................................ ................................ .................. 69 4.2.2 Differential Scanning Calorimetry (DSC) ................................ .............................. 69 4.2.3 Optical Microscopy ................................ ................................ ................................ 70 4.2.4 Optical Profi lometry ................................ ................................ ............................... 70 4.2.5 Particle Size analysis ................................ ................................ .............................. 71 4.3 Computational Chemistry ................................ ................................ ................................ 71 5 EXPERIMENTAL RESULTS ................................ ................................ ............................... 74 5.1 Differential Reflectograms of Energetic Materials ................................ ........................... 74 5.2 Characterization of TNT Purity, Crystal Structure, and Morphology .............................. 75 5.3 Behavior of Energetic and Other Materials with a Ring Structure ................................ ... 76 5.4 Mixtures of Sev eral Energetic Materials ................................ ................................ .......... 77 5.5 Effect of Rotating or Tilting the Sample ................................ ................................ .......... 77 5.6 Dependence on the Number of Nitro Groups Present ................................ ...................... 78 5.7 Comparison of Differential Reflectograms with Results from Other Optical Techniques ................................ ................................ ................................ .......................... 78 5.8 Experiments Leading to Possib le Answers for the Observed Transitions ........................ 81 5.8.1 Experimental Modifications of UV Visible Transmission Measurements ............ 81 5.8.2 Exp osure of TNT to UV Light ................................ ................................ ............... 83 5.9 Line Shape Analysis of Differential Reflectograms ................................ ......................... 86 6 COMPUTATIONAL CHEMISTRY RESULTS ................................ ................................ 126 6.1 Single Isolated Molecule of TNT ................................ ................................ ................... 126 6.2 One Isolated TNT Molecule from Monoclinic and Orthorhombic Crystals .................. 127 6.3 Two Isolated TNT Molecules from the Monoclinic and Orthorhombic Crystals .......... 128 6.4 Several (more than two) Molecules Involving Monoclinic and O rthorhombic Crystals ................................ ................................ ................................ .............................. 129 6.5 Entire Molecular Crystals of TNT: Monoclinic and Orthorhombic ............................... 129 6.6 Changing the Intermolecular D istance between Pairs of TNT Molecules ..................... 130 6.7 Modifying the TNT Unit Cell to Yield Defects in the Crystal ................................ ....... 133

PAGE 7

7 6.8 Other Energet ic Materials ................................ ................................ ............................... 135 6.8.1 RDX ................................ ................................ ................................ ...................... 135 6.8.2 HMX ................................ ................................ ................................ ..................... 136 6.8.3 Tetryl ................................ ................................ ................................ .................... 137 6.8.4 PETN ................................ ................................ ................................ .................... 138 6.8.5 Trinitrotoluene Derivatives: DNT and MNT ................................ ....................... 138 6.9 Non High Explosive Materials (Components of the TNT Molecule) ............................ 140 7 FURTHER DISCUSSION ................................ ................................ ................................ .... 182 7.1 Review of TNT Absorption ................................ ................................ ............................ 182 7.2 Origin of the Long Wavelength Feature ................................ ................................ ......... 183 7.2.1 Solid State Effect? ................................ ................................ ................................ 184 7.2.2 Charge Transfer Self Complex ................................ ................................ ............. 185 7.2.3 Contribution of the NO 2 Groups ................................ ................................ ........... 186 7.3 Summary ................................ ................................ ................................ ......................... 189 8 SUMMARY AND CONCLUSIONS ................................ ................................ ................... 190 9 FUTURE WORK ................................ ................................ ................................ .................. 193 9.1 Optical Characterization ................................ ................................ ................................ 193 9.2 Photolysis of Energetic Materials ................................ ................................ ................... 194 9.3 Computational Chemistry ................................ ................................ ............................... 194 APPENDIX A EXTENDED BACKGROUND OF MOLECULAR ORBITAL THEORY ........................ 196 A.1 Electrons ................................ ................................ ................................ ........................ 196 A.2 Atomic orbitals ................................ ................................ ................................ .............. 196 A.3 Molecular orbitals ................................ ................................ ................................ .......... 198 LIST OF REFERENCES ................................ ................................ ................................ ............. 201 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 208

PAGE 8

8 LIST OF TABLES Table page 1 1 Summary of current and proposed energetic material detection techniques and their capabilities. ................................ ................................ ................................ ........................ 30 2 1 The density and nitrogen percentage of several common energetic materials including those studied 60 ................................ ................................ ................................ ... 34 2 2 The explosive velocities and powers of several common energetic materi als 63 ............... 36 2 3 Unit cell parameters of crystalline TNT at room temperature 65 ................................ ....... 38 2 4 Torsion angles of the NO 2 groups in several different conformations of the TNT molecule. ................................ ................................ ................................ ............................ 39 2 5 Unit cell parameters for the RDX molecular crystal at room temperature 71 .................... 40 2 6 Unit cell parameters for the beta HMX molecular crystal at room temperature 74 ............ 43 2 7 Unit cell parameters for the Tetryl molecular crystal at room temperature 75 ................... 44 2 8 Unit cell parameters of the PETN molecular crystal at room temperature 76 .................... 45 2 9 Ultraviolet absorption maxima for TNT, RDX, HMX, and PETN as rep orted in the literature. ................................ ................................ ................................ ............................ 50 5 1 Polarities of the solvent used in UV Visible transmission spectrophotometry. .............. 125 5 2 Electron tr ansitions of various energetic materials determined by the graphical line shape analysis of the differential reflectograms. ................................ .............................. 125

PAGE 9

9 LIST OF FIGURES Figure page 2 1 Potential energy diagram demonstrating the activation energy (E a ) and reaction energy (E r ) of a molecule. ................................ ................................ ................................ .. 33 2 2 Molecular compositional structures of several common energetic materials. ................... 34 2 3 Characterization of the explosives investigated by chemical and explosive properties 5 ................................ ................................ ................................ ......................... 35 2 4 The molecular structure and molecular crystals of TNT present at room tempe rature. .... 37 2 5 Trinitrotoluene m olecules in conformation A and B from the monoclinic molecular crystal superimposed on one another.. ................................ ................................ ............... 39 2 7 The RDX molecular structure and crystal structure stable at room temperature. ............ 40 2 8 HMX that is the stable crystal structure at room temperature.. ................................ ................................ .............. 42 2 9 The molecular structure and crystal structure of Tetryl in the stable phase at room temperature. ................................ ................................ ................................ .................... 44 2 10 The molecular structure and molecular crystal structure of PETN. ................................ 46 3 1 Molecular orbital energy diagram of mononitrobenzene (MNB). ................................ ..... 55 4 1 Differential reflection spectrometer used in the research. ................................ ............... 60 4 2 Curves representing the function of the photon energy (s) and an optical parameter 97 ........................ 63 5 1 Differential reflectogram of a typical solid sample of dehydrated TNT on carbon tape. ................................ ................................ ................................ ................................ .... 88 5 2 Differential reflectograms of common organic and inorganic materials on carbon tape compared to TNT ................................ ................................ ................................ ....... 89 5 3 Differential reflectograms of vegetation from several p lants compared to that of TNT. ................................ ................................ ................................ ................................ .. 90 5 4 Differential reflectograms of several common energetic materials on carbon tape. ......... 91 5 5 Differential reflectograms of TNT received from several sources including pure standards and military compositions on carbon tape. ................................ ........................ 92 5 6 Differential reflectograms of several commercial explosive materials on carbon tape ..... 92

PAGE 10

10 5 7 Differential reflectograms of TNT on carbon tape, recrystallized independently from solution with different solvents. ................................ ................................ ......................... 93 5 8 Differential reflectograms of TNT on several different substrates ................................ .... 94 5 9 Chemical analysis results of high pressure liquid chromatography mass spectra f or TNT from a standards company and a military explosive.. ................................ ............... 95 5 10 Differential scanning calorimetry spectrum of TNT, recrystallized from an acetonitrile solution. ................................ ................................ ................................ ........... 96 5 11 Spectrum from x ray spectrometry of solid TNT recry stallized from an acetonitrile solution ................................ ................................ ................................ ............................... 96 5 12 Optical micrographs of the measured energetic m aterials at 20X magnification. ............. 97 5 13 Optical profilometry using false color to indicate height from the base plane (pale yellow in color) of a solid TNT sample recrystallized from an acetonitrile solution. ....... 98 5 14 Different ial reflectograms of several materials with cyclic or ring molecular structures on carbon tape. ................................ ................................ ................................ .. 98 5 15 Molecular structures of the materials with cyclic molecular structures shown in Figure 5 12. ................................ ................................ ................................ ........................ 99 5 16 Differential reflectograms of PETN mixed with TNT (A) and C 4 mixed with TNT (B) on carbon tape. ................................ ................................ ................................ ............. 99 5 17 Differenti al reflectograms of a solid TNT sample, on carbon tape, at different rotation angles with respect to the incident light beam. ................................ ................... 100 5 18 Schematic representation of the tilting of a sample resul ting reflection in the direction of the dotted arrows. ................................ ................................ ......................... 100 5 19 Differential reflectograms of TNT, on carbon tape, at different tilt angles with respect to the incident light. ................................ ................................ ............................. 101 5 20 Differential reflectograms of solid samples of 2, 4 DNT, 2, 6 DNT, and 1, 3 DNB on carbon tape. ................................ ................................ ................................ ...................... 101 5 21 The wavelength of the inflecti on point of the 400 nm feature for TNT, 2, 4 DNT, 2, 6 DNT, and 1, 3 DNB ................................ ................................ ................................ ......... 102 5 22 Reflection spectrum of solid TNT on the reflection stage of a Perkin Elmer UV Visible spectrophotometer. ................................ ................................ .............................. 102 5 23 Molar absorbtivity of TNT measured in transmission utilizing an acetonitrile solution with at a concentration of 3.1 g/L taken with four different spectrophotometers. ........... 103

PAGE 11

11 5 24 Absorbance spectra of varying concentrations of TNT in acetonitrile solution measured by a Varian UV visible spectrophotometer. ................................ .................... 104 5 25 Mol ar absorbtivity spectra of various energetic materials in solution of equal parts of ethanol and acetonitrile at concentrations of 1 g/L. ................................ ......................... 105 5 26 Molar absorbtivity spectra with varying conce ntration of the sample in the solution for (A) RDX, (B) HMX, (C) PETN, and (D) Tetryl. ................................ ....................... 106 5 27 Absorbance spectra of separate solutions of 2, 4 DNT, 2, 6 DNT, and 1, 3 DNB in high 1 g/L (A) a nd low 0.1 g/L (B) concentrations with equal parts ethanol and acetonitrile. ................................ ................................ ................................ ....................... 107 5 28 Absorbance spectra of separate solutions of 2 MNT, 3 MNT, 4 MNT, and nitrobenzene with equal parts of etha nol and acetonitril e at a concentration of 1 g/L. .. 107 5 29 Wavelength of the inflection point for the 400 nm feature taken with an UV Visible transmission spectophotometer of solutions of TNT 2, 4 DNT, 2, 6 DNT, 1, 3 DNB, 2 MNT, 3 MNT, 4 MNT, and nitrobenzene with their respective solvents at a concentration of 1 g/L. ................................ ................................ ................................ ..... 108 5 30 Molar absorbtivity for varying concentrations of TNT in acetonitrile. ........................... 109 5 31 Absorbance intensity versus the concentration of TNT in solution with acetonitrile at several wavelengths. ................................ ................................ ................................ ........ 110 5 32 Molar absorbtivity of TNT in solution with several different solvents at the same concentration. ................................ ................................ ................................ ................... 112 5 33 Wavelength of the inflection point of the 420 nm feature for TNT in solu tion with several solvents versus the solvent polarity. ................................ ................................ .... 112 5 34 Differential reflectograms of solid TNT with increasing time of exposure to the monochromated 200 nm (A) and 250 nm (B) light o f a high pressure Xe lamp. ............ 113 5 35 Differential reflectometry of a solid TNT sample exposed to the sun at increasing time intervals (A) short term and (B) long term.. ................................ ............................ 113 5 36 Spectrum of sunlight behind and in front of a lab window. ................................ ............. 114 5 37 Differential reflectograms of solid TNT with increasing time of exposure to the 325 nm line of a defocused HeCd laser. ................................ ................................ ................. 115 5 38 Ratio of the intensity of the normalized differential reflectometry at 380 nm to 450 nm of TNT versus the time of exposure to the 325 nm HeC d laser line. ........................ 116 5 39 Differential reflectograms of solid TNT that has been irradiated for 16 minutes with the 325 nm HeCd laser line followed by anneal ing for various times at 45C. .............. 117

PAGE 12

12 5 40 Ratio of the normalized differential reflection intensity at 380 nm and 450 nm of TNT that has been irradiated with the 325 nm line of the HeCd laser for 16 minutes .... 117 5 41 Differential reflectograms of solid TNT that has not been exposed to a UV laser versus annealing time at 45C. ................................ ................................ ........................ 118 5 42 Differential reflectogra ms of solid TNT with increasing time of exposure to the 457 nm (A) and 488 nm (B) lines of an Ar ion laser. ................................ ............................. 118 5 43 Ratio of the normalized differential reflection at 380 and 450 nm of solid TNT versus increasing time of exposure to the 488 nm line of an Ar ion laser. ...................... 119 5 44 Fourier transform infrared spectrometry of solid TNT on stainless steel with and without 16 minutes of ex posure to the 325 nm line of a HeCd laser. .............................. 120 5 45 Chemical analysis results by high pressure liquid chromatography and mass spectrometry. ................................ ................................ ................................ .................... 120 5 46 Absorption transitions of TNT based on a graphical line shape analysis of the respective differential reflectogram, that is, of Figure 5 1. ................................ ............. 121 5 47 Graphical line shape a nalysis of the respective differential reflectogram of RDX (Figure 5 4) revealing the absorption transitions. ................................ ............................ 121 5 48 Absorption transitions of PETN based on graphical line shape analysis of the respective differential reflectogram, that is, of Figure 5 4. ................................ ............. 122 5 49 Graphical line shape analysis of the respective differential reflectogram of HMX (Figure 5 4) giving the absorption t ransitions. ................................ ................................ 122 5 50 Absorption transitions of Tetryl based on graphical line shape analysis of the respective differential reflectogram, that is, of Figure 5 4. ................................ ............. 123 5 51 The absorption transitions of 2, 6 DNT found using graphical line shape analysis of the respective differential reflectogram that is Figure 5 4. ................................ .............. 123 5 52 Graphical line shape analysis of the respective differential reflectogram of 2, 4 DNT (Figure 5 4) revealing the absorption transitions. ................................ ............................ 124 5 53 Absorption transitions of 1, 3 DNB based on graphical line shape analysis of the respective differential reflectogram, that is, of Figure 5 4. ................................ ............. 124 6 1 Calculated molecular orbital transitions by ZINDO of one isolated TNT molecule. ...... 141 6 2 Calculated molecular orbital transitions by ZINDO of one TNT molecule based on monoclinic molecular crystal geometry and optimized using the PM3 method. ............. 142

PAGE 13

13 6 3 A modeled UV Visible spectrum based on the calculated molecular orbital transitions by ZINDO for one TNT molecule based on the monoclinic molecular crystal geometry optimized with the PM3 method. ................................ ......................... 143 6 4 One isolated TNT molecule based on the monoclinic molecular crystal geometry and geometry optimized using the PM3 method with molecular orbitals that participate in absorption transitions highlighted. ................................ ................................ ................... 144 6 5 Molecular orbital transitions calculated by ZINDO for one isolated TNT molecule based on the orthorhombic molecular crystal geometry and optimized using the PM3 method. ................................ ................................ ................................ ............................. 145 6 6 Calculated molecular orbital transitions by time dependent density functional theory (TDDFT) for one isolated TNT molecule with the geometry optimized. ........................ 146 6 7 The molecular orbital energies versus orbital occupancy for one isolated TNT molecule calculated by ZINDO and density functional theory (DFT). ........................... 147 6 8 Molecular orbital tr ansitions calculated by ZINDO for one isolated TNT molecule having the A conformation in the monoclinic molecular crystal. ................................ .... 147 6 9 The molecular orbital transitions for one isolated TNT molecu le in the A conformation of the orthorhombic molecular crystal. ................................ ..................... 148 6 10 Modeled spectrum of one isolated TNT molecule with the A conformation of the monoclinic molecular crystal based on the molecular orbital transitions calculated by ZINDO.. ................................ ................................ ................................ ........................... 149 6 11 Molecular orbitals that participate in the MO transitions calculated for one isolated TNT molecule with the A conformation of the m onoclinic molecular crystal ................ 150 6 12 Calculated molecular orbital transitions by ZINDO of one isolated TNT molecule having the B conformation of the monoclinic molecular crystal. ................................ .... 151 6 13 Molecular orbital transitions of one isolated TNT molecule in the B conformation of the orthorhombic molecular crystal calculated by ZINDO. ................................ ............. 151 6 14 Calculated molecular orbital transitions of two TNT molecules in both A and B conformations of the monoclinic molecular crystal as they would appear in the crystal as a pair in the same plane. ................................ ................................ ................... 152 6 15 Molecular orbital transitions of two TNT molecules based on the A and B conformations of the monoclinic molecular crystal and geometry optimized using the PM3 method calculated by ZINDO. ................................ ................................ ................ 153 6 16 Calculated molecular orbital transitions by ZINDO of two TNT molecules in the same plan having the A and B conformations of the orthorhombic molecular crystal as they would appear in the crystal structure. ................................ ................................ .. 154

PAGE 14

14 6 17 Molecular orbital transitions calculated by ZINDO of two TNT molecules based on the A and B conformations of the orthorhombic molecular crystal and geometry optimized using the PM3 method. ................................ ................................ ................... 154 6 18 The molecular orbital transitions of two TNT molecules based on the A and B conformations of the monoclinic crystal calculated by TDDFT. ................................ .... 155 6 19 The energy of the molecular orbitals versus occupancy of one and two TNT molecule systems calculated separately by ZINDO and DFT ................................ ......... 155 6 20 Summary of the calculated molecular orbi tal transitions by ZINDO for increasing numbers of monoclinic conformation A TNT molecules in a system. ............................ 156 6 21 Wavelength of the longest wavelength (reddest) calculated molecular orbital tra nsition by ZINDO for increasing numbers of TNT molecules in the B conformation of the monoclinic molecular crystal ................................ .......................... 157 6 22 Summary of the molecular orbital calculations by ZINDO of increasing numbers of TNT molecules in the A conformation of the orthorhombic crystal ............................... 158 6 23 The longest wavelength (reddest) MO transition calculated by ZINDO with increasing number of TNT molecules in conformation B of the orthorhombic crystal. .. 159 6 24 Calculated molecular orbital transitions of a unit cell (8 molecules) of the monoclinic molecular crystal of TNT by ZINDO. ................................ ................................ ............. 160 6 25 Representations of the unit cell of the monoclinic molecular crystal of TNT after geometry optimization with the PM3 method. ................................ ................................ 160 6 2 6 Molecular orbital transitions calculated by ZINDO of 8 molecules based on the unit cell of the monoclinic molecular crystal of TNT and geometry optimized using the PM3 method. ................................ ................................ ................................ .................... 161 6 27 Calcu lated molecular orbital transitions by ZINDO of a TNT unit cell (8 molecules) of the orthorhombic molecular crystal. ................................ ................................ ............ 161 6 28 Molecular orbital transitions of 8 TNT molecules based on the unit cell of the orthorhombic molecular crystal and geometry optimized using the PM3 method calculated by ZINDO. ................................ ................................ ................................ ...... 162 6 29 Summary of the calculated molecular orbital transitions by ZINDO for two T NT molecules in the A and B conformation of the monoclinic molecular crystal at varying intermolecular distances. ................................ ................................ .................... 163 6 30 Calculated molecular orbital transitions by ZINDO for pairs of TNT mo lecules in the A and B conformations of the orthorhombic crystals as they would appear in the crystal at varying intermolecular distances. ................................ ................................ ..... 164

PAGE 15

15 6 31 The longest wavelength (reddest) calculated molecular orbital transition by ZINDO versus the intermolecular distance of a pair of TNT molecules in the A conformation (A) and B conformation (B) of the monoclinic crystal. ................................ ................... 164 6 32 Summary of the calculated molecular orbital transitions by ZINDO for pairs of TNT molecules in the A conformation (A) and B conformation (B) of the orthorhombic molecular crystal at varying intermolecular distances. ................................ .................... 165 6 33 Molecular orbital transitions for a pair of orthorhombic conformation A TNT molecules stacked like pancakes at an intermolecular distance of 1.5 calculated by ZINDO. ................................ ................................ ................................ ............................ 165 6 34 Calculated molecular orbital transitions by ZINDO of the monoclinic unit cell of TNT with one TNT molecule replaced with a DNT molecule. ................................ ....... 166 6 35 Molecular orbital transitions calcu lated by ZINDO of the monoclinic unit cell of TNT with a MNT molecule in the place of one TNT molecule. ................................ ..... 166 6 36 The calculated molecular orbital transitions by ZINDO of a modified monoclinic molecular crystal unit cell of TNT where one TNT molecule is replaced with a toluene molecule ................................ ................................ ................................ .............. 167 6 37 Molecular orbital transitions calculated by ZINDO of a unit cell of the monoclinic molecu lar crystal of TNT with a benzene molecule replacing one TNT molecule. ........ 167 6 38 Calculated molecular orbital transitions by ZINDO of a monoclinic molecular crystal unit cell of TNT where one T NT molecule has been removed leaving a vacancy .......... 168 6 39 The molecular orbital transitions for a unit cell of the monoclinic molecular crystal of TNT where two TNT molecules have been removed l eaving a plane vacancy calculated by ZINDO ................................ ................................ ................................ ....... 168 6 40 Calculated molecular orbital transitions by ZINDO of the monoclinic molecular crystal unit cell of TNT where four TNT molecules in the stack ing sequence have been removed and four TNT molecules with orthorhombic molecular crystal conformations have been inserted creating a stacking fault ................................ ............ 169 6 41 Molecular orbital transition calcu lations by ZINDO of one RDX molecule from the molecular crystal stable at room temperature ................................ ................................ .. 170 6 42 The calculated molecular orbital transitions by ZINDO for two RDX molecules as they would ap pear in the molecular crystal stable at room temperature .......................... 171 6 43 Molecular orbital transitions of one unit cell (8 molecules) of the RDX crystal stable at room temperature calculated by ZINDO ................................ ................................ ..... 171 6 44 The molecular orbital transitions of a HMX molecule from the molecular crystal that is stable at room temperature calculated by ZINDO ................................ ....................... 172

PAGE 16

16 6 45 Calculated molecular orbital transitions by ZINDO for two HMX molecules from the molecular crystal ................................ ................................ ................................ .............. 173 6 46 Molecular orbital transitions for four HMX molecules from t he molecul ar crystal calculated by ZINDO ................................ ................................ ................................ ....... 173 6 47 The calculated molecular orbital transitions by ZINDO of one Tetryl molecule from the molecular crystal stable at room temperature. ................................ ........................... 174 6 48 Molecular orbital transitions calculated by ZINDO for two Tetryl molecules as they would appear in the molecular crystal. ................................ ................................ ............ 175 6 4 9 Calculated molecular orbitals transitions by ZINDO of an entire unit cell (eight molecules) of Tetryl ................................ ................................ ................................ ......... 175 6 50 The molecular orbital transitions of one PETN molecule as calculated by ZINDO. ...... 176 6 51 The results of calculating the molecular orbital transitions by ZINDO for a pair of PETN molecules as they appear in the molecular crystal ................................ ................ 177 6 52 Molecular orbital transitions of four PETN molecules as they would appear in the molecular crystal calculated by ZINDO. ................................ ................................ ......... 177 6 53 The calculated MO transitions by Z INDO of one molecule of 2, 4 DNT (A), 2, 4 DNT with geometry optimization (B), 2, 6 DNT (C), and 2, 6 DNT with geometry optimization (D). ................................ ................................ ................................ .............. 178 6 54 Molecular orbital transitions calculated by Z INDO of one molecule of several low power explosive materials: 2 MNT (A), 2 MNT with geometry optimization (B), 3 MNT (C), 3 MNT with geometry optimization (D), 4 MNT (E), and 4 MNT with geometry optimization (F). ................................ ................................ .............................. 179 6 55 The results of the calculated molecular orbital transition by ZINDO of one molecule of toluene (A), toluene with geometry optimization (B), benzene (C), and benzene with geometry optimization (D). ................................ ................................ ...................... 180 6 56 Calculated molecular orbital transitions by ZINDO of derivatives of the TNT molecules as they would appear in the molecule: the nitro group (A) and the methylene group (B) and after geometry optimization of the nitro group (C) and the methylene group (D). ................................ ................................ ................................ ....... 181 A 1 Simplified molecular orbital energy diagram for the combination of two 2p atomic orbitals 93 ................................ ................................ ................................ .......................... 200

PAGE 17

17 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy INVESTIGATION OF SELECT ENERGETIC MATERIALS BY DIFFERENTIAL REFLECTION SPECTROMETRY By Anna Marie Fuller August 2007 Chair: Rolf E. Hummel Major: Materials Science and Engineering continue to target buildings and mass transit systems with explosive devices. The detection of these energetic materials is necessary to insure national security and welfare. D etection techniques such as X ray scanners, Raman spectroscopy, Terahertz spectroscopy and ion mobility spectrometry are in current use or development; h owever, none of these are appropriate for all ne cessary applications. These techniques include. The present document provides an overview of the current detection techniques and describes a new technique for detecting energetic mate rials called differential reflection spectrometry (DRS). DRS essentially measures the optical absorption of energetic materials. The use of DRS has led to the discovery of previously unreported optical characteristics for some energetic compounds that a re unique to the individual material. These optical characteristics consist of absorption shoulders between 270 and 420 nm, e.g. near 420 nm for 2, 4, 6 trinitrotoluene (TNT). In the presented research, the origin of the differential reflection spectra o btained was investigated using several techniques including UV Visible spectrophotometry (transmission and reflection) and computer molecular modeling. Experimental DRS spectra of TNT, hexahydro 1,3,5 trinitro 1,3,5 triazine (RDX), octahydro 1,3,5,7 tetra nitro 1,3,5,6 tetrazocine (HMX),

PAGE 18

18 pentaerythritol tetranitrate (PETN), and 2, 4, 6, n tetranitro n methylaniline (Tetryl) were taken and analyzed. From the experimental results and verification by molecular modeling, it was found that the absorption featu res observed in the redder region of the UV range (270 420 nm) are due to molecular orbital transitions in the nitro (NO 2 ) groups of the measured explosives. These transitions only occur in specific conditions, such as high concentration solutions and sol ids, where the normally forbidden transitions are allowed. The unique optical characteristics of the energetic materials presented in this dissertation are observed only in the solid or relatively high concentrated states suggesting the interaction of se veral molecules. Therefore these absorption features are proposed to be due to a charge transfer self complex. T his phenomenon can be interpreted in the same manner as the accumulation of atoms and be modeled using quantum mechanics

PAGE 19

19 CHAPTER 1 INTRODUCTION 1.1 Introduction to the Presented Research impact on our daily lives. Terrorists continue to target buildings and mass transit systems with explosive devices. In war torn areas, live landmines injure more than 2000 civilians a month 1 Former wea pon manufacturing sites typically contain toxic levels of explosive residue within the soil and surrounding watershed. Detection of these energetic materials is necessary to insure the security and health of our society. 1.1.1 Current Detection Techniqu es There are several detection systems currently employed at airports, sea ports, and for forensics. Ion mobility spectrometry is currently in use for passenger and carry on luggage screening by sampling with air puffers and swabbing. In addition, X ray scanners check for unusual (i.e. denser) items in carry on luggage. A new passenger screening system based on millimeter wave technology is being tested at a few airports 2 4 The technology detects the temperature changes from a person (millimeter wave radiation) and boasts improved resolution compared to current X ray technology. Color test kits are often used as a field test or in forensic investigations where certain reagents cause a color change in particular energetic materials allowing them to be identified 5 The sensitivity of these tests can be improved using the photoluminescence of the colored product 6,7 Also in use are handheld Raman spectrometers that can chemically identify several different types of chemicals 8 Electronic noses have been developed to imitate the response of a bomb sniffing dog and are being considere d for deployment in Iraq 9

PAGE 20

20 All the previously mentioned techniques have advantages and disadvantages for application to explosive materia ls detection. Moreover several current detectors require a sampling technique such that the material of interest must be collected and often processed before a measurement can occur. This sequence of steps increases measurement time and reduces the appli cability to standoff or contact less sensing. For example, in order for ion mobility spectrometry to be effective the energetic material must be removed from a surface and placed inside the instrument. False positives are also quite common with the ion m obility scanners especially for nitrate containing fertilizers often found on golf balls and shoes. X ray and millimeter wave systems do not detect the energetic material itself but rather an anomaly (an item with a higher density or lower thermal respons e) on a person or in a package that also contributes to false positives. These particular systems also tend to be quite large in size as well as expensive. The Raman scanner, while compact and portable, needs to be within a few centimeters of the questio nable item in order to detect an energetic material. Other common techniques are also time consuming and only a small sample of the passengers and luggage entering an airport terminal is actually surveyed. Many other proposed instruments are not portable and therefore cannot be readily adapted to several different applications including standoff detection All of the previous examples demonstrate the need for a simple, fast, accurate, portable, and contact less detection system. 1.1.2 Novel Detection Met hod: Differential Reflection Spectrometry The differential reflection spectrometer (DRS), also called a differential reflectometer, was developed at the University of Florida in 1970 and has the ability to fulfill the above stated criteria 10 The DRS essentially measures the imaginary part of the dielectric constant of a material known as the absorption. Characteristic DRS spectra for several common explosives including 2, 4, 6 trinitrotoluene (TNT), hexahydro 1,3,5 trinitro 1,3,5 triazine (RDX), octahydro

PAGE 21

21 1,3,5,7 tetranitro 1,3,5,6 tetrazocine (HMX), pentaerythritol (PETN), and 2,4,6, n tetranitro n methylaniline (Tetryl) have been found in the present research 11 15 The instrumentation is relatively simple in construction utilizing a white ligh t source, a modulation of the sample area, a spectrometer, and a detector. Two individual reflectivities are collected from two spots in close proximity (~2 mm) creating a differential reflection spectrum. Collecting or processing a sample material is no t required and the instrument does not have to be in contact with the suspicious item; a beam of light can simply be directed at the target area. The differential nature of the technique yields a high sensitivity which can detect and characterize a trace amount of explosive 2 ) in a short period of time, on the order of 100 ms. 1.2 Proposed Detection Techniques for Energetic Materials As science advances there are several new techniques that have been proposed for detection of energetic materials. Many of these techniques may not be in wide use but they are worth mentioning. The methods are divided as general chemistry based techniques and optical techniques. 1.2.1 General Techniques The development of explosive detection devices has stretched to nearly every part of science. Many of the techniques that have been proposed are for application in specific situations. These can be generally divided into several different analytical techniques: vapor and trace detectors (chemistry derived techniques ), bulk explosive detectors, and atmospheric sensors. Pure analytical chemistry techniques are used for trace detection in forensics research and in chemical analysis in research. Liquid chromatography (LC) can be used to identify energetic materials. Th inlayer, paper, and high performance (also called high pressure) represent different types of LC; all techniques are based on the separation of the chemical compounds into organic

PAGE 22

22 molecules and compounds. In LC, the chemical compounds are separated by a m obile solvent phase that utilizes the mass of the individual molecules and compounds to differentially separate them. The flow rate of each compound can be calculated using the distance a given spot traveled from the original spot on the chromatography pl ate and is used for identification 5 Gas chromatography (GC) is based on the same principle but instead a sample in a vapor state is injected into a chamber where a mobile carrier gas separates the compounds present in the sample 5 Analytical testing can be performed on the separated compounds by inserting them into a photoluminescence chamber or mass spectrometry system for LC and GC. Vapor and trace energetic materials detectors chemically identify the presence of energetic materials by either coupling a gas chromatographer with an electron capture detector, an ion mo bility spectrometer, or a mass spectrometer. Ion mobility and mass spectrometry are based on the unique mass of individual ions 16 In ion mobility spectroscopy, an ion is created in a reactor and then injected by an electric field gate into a drift region. This drift region is composed of a m agnetic field that separates the ions by their mobility, a function of ion mass, that are then detected by an ion sensitive plate. The time the ion takes to get from the gate to the detector yields the mobility of the ion; so it can be identified. Nitrat e containing energetic materials are chiefly detected with this technique by the presence of NO 2 and NO 3 ions. Mass spectrometry is based on the mass to charge ratio of a charged particle. The charged particles are accelerated through a curved magnetic field mass analyzer, where the radius of the charged particle path is characteristic for the mass to charge ratio. An ion collector or photographic plate measures the position of a particle after the mass analyzer, yielding the mass to charge ratio and i dentifying the ion where in general to heavier ions have a larger radius 17

PAGE 23

23 The detection of bulk energetic materials, for example in suitcases and cars, has been proposed to be performed by neutron analysis, nuclear quadrupole resonance, and X ray detection. Neutron analysis inc ludes thermal neutron (TNA), fast neutron (FNA), and pulsed fast/thermal neutron (PFTNA) and is based on the neutron excitation of atoms in energetic materials. TNA measures the characteristic gamma rays produced by the low energy neutrons interacting wit h the nitrogen in the explosive. In contrast, FNA uses high energy neutrons in order to create the gamma rays from the nitrogen. PFTNA uses the same technique but the neutron generator is pulsed thereby increasing the number of nuclear reactions. Gamma rays from these reactions are detected, as well as those created by fast and thermal neutrons, and measured between the neutron generator pulses giving PFTNA a higher selectivity 17,18 Nuclear quadrupole resonance (NQR) is based on the interaction of radio frequency energy with quadrupole energy levels of an atomic nucleus. In the case of energetic materials, the detector would concentrate on the presence of nitrogen isotopes in the sample, e.g. 14 N for TNT 19 X ray detection systems employ the low angle x ray scattering from a sample in order to se e density profiles inside an object such as a suitcase 17 This is the same concept that is applied to medical X ray machines. New X ray technologies can detect the higher resolution backscattered X rays 4 This allows lower doses of radiation for detection; though any amount of ionizing radiation is harmful. However, one determent to this technique is the anatomical detail produced during passe nger screening. These techniques can only detect the presence of an unexpected object, such as a bomb or firearm, but not the explosive chemical explicitly. Detection of energetic materials in the atmosphere, as vapors or gases, can be conducted by immun ochemical tests, amperometric gas sensors, and voltammetry sensors. The immunochemical test involves the reaction of a target compound (the explosive compound) with

PAGE 24

24 a specific antibody 9,20 The amperometric gas sensor measures the electrical current in an electrochemical sample cell positioned between two electrodes at a constant potenti al. The presence of explosive decomposition products is indicated by an increase in the current at a specified potential. The closely related voltammetry sensors measure the potentials (voltage) corresponding to the peaks in the current that occur in an electrochemical cell when an explosive is present. The potential intensity at a particular current peak yields the concentration of the explosive 17 1.2.2 Optical Techniques Due to simplicity and potential portability, several optical techniques have also been proposed for detect ion of energetic materials. Ultraviolet(UV) visible range absorption spectra are widely used for chemical analysis of organic materials in a laboratory 21 23 Techniques that exploit the absorption of energetic materials in the UV and visible range have been proposed. Infrared spectroscopy has also been used extensively in characterization of energetic materials by measuring the resonance of vibrational modes related to individual molecules and specific functional groups 17,24,25 3mm or 0.1 THz 10THz) for spectroscopy and imaging of energetic materials are also being aggressively researched. Techniques not directly related to the optical properties of energetic materials use the re action of energetic materials with specific assays and subsequent fluorescence or the Many energetic materials have absorption in the UV. The optical signals collected in this range commonly are low in intensity and have high amounts of noise. Several instruments have been developed incorporating methods to increase the signal strength and accuracy 22,26,27 Todd e t al. demonstrated the use of cavity ringdown spectroscopy (CRDS) to detect trace explosive vapors 1,20 CRDS employs a pulsed laser injected into an optical cavity containing the sample

PAGE 25

25 gas. After a laser pulse, the radiation in the cavity decays and is measured. The direct absorption spectrum of the sampl e gas can then be determined from the decay measurement and then related to the absorption of known explosives. The molecular structure of energetic materials often leads to optical quenching properties and can be used indirectly for their detection 7,28 32 The energetic materials are placed on, or streamed by, a material that is photoluminescen t in the UV or visible range. The intensity of the luminesc ence decreases due to quenching indicating the presence of an explosive compound. This is an indirect optical technique where neither the absorption nor the emission of the explosive compound is measured. The photoluminescent materials used vary from porous silicon to selected polymers. One such study utilizes porous Silicon, which has a photoluminescence with a maximum at approximately 640nm 28 When a piece of porou s Si is exposed to TNT vapor for a period of time, the initial photoluminescence is que nched. Cleaning of the porous s ilicon does not restore the original intensity of the photoluminescence therefore the quenching is attributed to bonding of the TNT to th e porous s ilicon surface. The TNT molecule donates electrons to the porous silicon which in return reduces the intensity of the luminescence. The same quenching phenomenon can be observed when TNT is added to pyrene in solution or a metalloe containing p olymer 31,32 Luminescence quenc hing in these circumstances was so large that the prior luminescent features were completely eliminated after TNT was added. The use of classical infrared range instruments for detection yields a relatively low signal for solid samples (mentioned above). A method proposed by Stahl and Tilotta uses Infrared spectroscopy with solid phase microextraction (SPME) 25 The explosive compounds

PAGE 26

26 are extracted from water onto a silicone polycarbonate copolymer film. The resulting sample is then examined by infrared spectroscop y to yield quantitative data with increased signal strength. Raman spectroscopy has also been explored for energetic material detection, as mentioned in Section 1.1 8,33 38 Unfortunately, the Raman Stokes shift has a low intensity and due to light scattering and can be difficult to resolve in ambient light. The sensitivity of a Raman instrument can be improved by absorbing the target molecules onto a specially designed surface. Surface enhanced Raman spectroscopy or (SERRS) is a relatively new technique, adapted for several classes of energetic materials at varyious wavelengths in the UV, visible, and infrared ranges 39 43 While SERRS increases the signal strength it also lim its the capability for standoff detection. The terahertz range of the electromagnetic spectrum may not be strictly considered an optical range. However, the instrumentation is quite similar to those used in the previously mentioned techniques so it is des cribed here. The use of terahertz frequency technology can yield spectral information and imaging for energetic material detection 44,45 Using an electro optic or similar terahertz emitter and an appropriate antenna the characteristic s pectrum of an energetic material is obtained 46 For imaging, the THz waves are directed at the object or person and an image similar to an x ray image is received 47 The terahertz spectra have a higher resolution than traditional x rays spectra especially for plastics, because materials have more var ied responses (absorption, transmission, or reflection) to terahertz waves. Terahertz radiation is non ionizing, so it is safe for biological systems, and can penetrate through several layers of clothing, although doing so reveals the anatomical detail of the imaged person. Fluorescence spectroscopy coupled with immunoassay reactions creates an optical sniffer for the detection of energetic materials 9 Energetic materials themselves have not been reported

PAGE 27

27 to demonstrate strong fluorescence in the UV to infrared range of the spectrum. Therefore the fluorescence of other materials must be employed to detect explosive materials. Bakken et al. employ an array of immunoassay sensors to the distal ends of an optical fiber bundle 20 The individual fibers in the optical bundle transmit both an incident lig ht to the sensors and the fluorescence response of the sensors, due to the optical excitation, to the detector. When the array is in contact with an explosive compound the respective sensors fluoresce creating a pattern unique to the individual explosive compound. As demonstrated above, very few explosives have luminescence in the UV and visible regions. Photofragmentation and laser induced fluorescence (PF LIF), a new technique, has been developed that utilizes the luminescence of the smaller molecular components of the explosive molecules 48 55 Basically high energy photons hit the target molecule and cause bond disas sociation which results in fragmentation of the molecule. A second laser of relatively high energy excites the electrons of one of the fragmented species to a higher energy level and subsequent fluorescence occurs at a higher energy or smaller wavelength. For common explosives like TNT, RDX, and PETN a 227 to 248 nm wavelength can be used for the first photofragmentation 54 The fragments of interest are the NO 2 groups that are in turn fragmented to vibrationally excited NO fragments and O. The second fragmentation is caused by a photon with a wavelength of 300 nm or less. A fluorescence wavelength of 193 nm or close to 227 nm is indicative of the NO fragment. The concentration of NO and therefore of the sampled explosive can be calculated from the intensity of the fluorescence 54 Due to the high power laser, this technique damages the surface that is being sampled. Laser induced breakdown spectro metry is very similar in nature to the PF LIF technique described ab ove in that a high energy light source is used to break apart the target

PAGE 28

28 molecule 56,57 However, in the case of LIBS, individual atoms are fragme nted from the molecule using a high power pulsed light source such as a frequency doubled Nd:YAG laser. The fragmented atoms are in a gas state and the unique light emission from them is detected using the same wavelength of light as the fragmenting light These elements have optical emission, in the form of sharp peaks, at specific wavelengths that are collected with a spectrometer and sent to a detector. The wavelength of light emission is unique for each element: C at 247 nm, H at 656 nm, N at 746, 82 1, and 869 nm, and O at 777 and 844 nm 57 The relative intensity of each element emission peak to the others identifies the original target molecule. 1.3 Concentration of the Presented Research The work contained in this research began as an investigation of a method to detect energetic materials. In the process, previously unreported optical properties of some explosive compounds were measured. Therefore the focus of the research was redirected to determining the physical origin of the optical properties and why th ey are seemingly limited to the energetic materials studied. The presented optical properties were observed mainly in the solid or relatively high concentrated solution states. This phenomenon is not uncommon in materials systems. For example, in order for ferromagnetism to be observed in an alloy containing iron, there is a defined concentration of iron atoms needed to obtain a positive exchange energy integral 58 A laser requires the presence of several atoms, in solid state or gas, to cause an electron population density in the conduction band large enough to stimulate additional electron excitation that in turn causes lasing. As demonstrated with these examples, the transfiguration of phenomena from a single atom to several atoms is frequent. In the reported research the optical characteristics of interest were seen when there was an accumulation of several molecules. The question that

PAGE 29

29 remains to be answered is can this phenomenon (the reported optical absorption) be interpreted in the same manner as the accumulation of a toms and modeled using quantum mechanics? 1.4 Summary The number of current and proposed energetic material detection techniques is great, but today there is no one instrument that can detect every explosive material in all applications. DRS has the pot ential for rapid standoff detection of small amounts of energetic materials using their distinct and unique optical absorption. The diverse amount of detection methods is due to the unique chemistry of this class of materials. Chapter 2 will explore the chemistry and optical properties of energetic materials.

PAGE 30

30 Table 1 1. Summary of current and proposed energetic material detection techniques and their capabilities. Proposed Techniques Directly Measure Energetic Material? Sample Processing? Portable? Stan doff? Bulk X Ray No No No Yes Terahertz Image: No Spectral: Yes No No Yes Millimeter Wave No No No Yes TNA/FNA/PFTNA Yes Yes No No Analy tical LC Yes Yes Yes No HPLC Yes Yes No No GC Yes Yes No No Vapor and Trace GC ECD Yes Yes No No IMS Yes Yes Yes No MS Yes Yes No No Atmo spheric Amperometric Yes Yes Yes No Immunochemical Yes Yes Yes No Voltammetry Yes Yes Yes No Optical CRDS Yes Yes No No Quenching No Yes Yes No IR SPME Yes Yes No No SERRS Yes Yes Yes No Sniffing No Yes Yes Yes PF LIF Yes No No Yes LIBS Yes No Yes Yes DRS Yes No Yes Yes Notes: TNA/FNA/PFTNA= Thermal/Fast/ Pulse Fast Thermal Neutron Analysis, LC/HPLC/GC= Liquid/High Pressure Liquid/Gas Chromatography, CRDS= Cathode Ring Down Spectrometry, IR SPME= I nfrared spectrometry with Solid Phase MicroExtraction, SERRS= Surface Enhance Raman spectrometry, PF LIF= Photofragmentation and Laser Induced Fluorescence, LIBS= Laser Induced Breakdown Spectrometry, DRS= Differential Reflection Spectrometry.

PAGE 31

31 CHAPTER 2 BACKGROUND INFORMATION FOR ENERGETIC MATERIALS The study of energetic materials has fascinated people for many centuries. The first recorded discovery of explosive material was in 220 BC by Chinese alchemists 59 They accidentally made blackpowder (sulfur, charcoal, and a nitrate such as KNO 3 saltpetre) that exploded while they were separating gold from silver in a low temperature reaction. Development of blackpowder continued f or several centuries and was the military standard explosive by the 13 th century, used in civil engineering in the 16 th century, and for mining beginning in the 17 th The majority of the common modern explosives such as TNT, RDX, HMX, picric acid, nitrogl ycerin, and PETN were not developed until the 19 th century, a very active period for explosives research. Several explosive materials were refined in process and stabilization and eventually brought into commercial manufacturing. One of the most famous w as the invention of nitroglycerine. It was first discovered by Italian professor Asciano Sobrero, but upon finding its explosive properties he ceased his studies. Swedish inventor Immanuel Nobel took this knowledge, developed a manufacturing process, and opened a small plant in 1863 with his son Alfred. Manufacturing nitroglycerin proved to be a difficult task due to periodic accidental initiations of the explosive. This resulted in the loss of two manufacturing plants over three years including one tha explosion, Alfred Nobel ceased manufacturing the nitroglycerine and instead concentrated on stabilizing the explosive. This resulted in the invention of dynamite in 1875. Dynamite is not made of TNT, a co mmon misconception, rather its chemical composition is actually nitroglycerine and kieselguhr (an absorber).

PAGE 32

32 2.1 Characteristics of Energetic Materials Explosives are part of a group of materials termed energetic materials 16 This class of materials receives its name from the large amount of e nergy that is stored within the molecules defined amount of energy in order to form, termed the heat of formation. For energetic materials, the molecular formation rea ction is endothermic, meaning it gains heat or energy from the exterior environment. The larger the heat of formation, the greater amount of energy is stored within the molecule. Explosion of a material is caused by the spontaneous release of energy upon decomposition, called heat of decomposition 16 The heat of decomposition is simply the energy from the exothermic reaction where the molecule decomposes to smaller molecules. It is equal in absolute value, but opposite in sign, to the heat of formation if the molecule decomposes to the same m olecules from which it was originally formed. If the molecule decomposes to different molecules, then the heat of decomposition may be larger or smaller than the heat of formation. The heat of reaction is also a description of the chemical reaction that occurs upon decomposition, which is calculated by subtracting the heat of formation of the reactant from the heat of formation of the products 16 In order for decomposition to occur, an energetic material needs to absorb a relatively small amount of energy (in comparison to the heat of decompo sition) from its surroundings, called the activation energy. This is shown by a potential energy curve in Figure 2 1. The decomposition can be rapid, depending on the material, once the activation energy is absorbed. Energy is released and heats the pro ducts to high temperatures that, in the case of gas products, causes a very large pressure increase. For example 2, 4, 6, trinitrotoluene (TNT) has a chemical formula C 7 H 5 N 3 O 6 The heat of reaction for an unconfined decomposition is 144.32 kcal/m ol. It produces 10.0 moles of H 2 N 2 and CO gas from one mole of solid TNT.

PAGE 33

33 Pressure increases from 1 atmosphere to 67.95 atmospheres during the decomposition, which causes an expansion from 137 mL to 224,000 mL of gas or an increase of 16,351%. The total rea ction expends 633 cal/g of heat into the surrounding environment. Using this thermodynamic information one can see how dramatic an explosion of TNT can be. Figure 2 1. Potential energy diagram demonstrating the activation energy (E a ) and reaction ener gy (E r ) of a molecule. 2.2 Chemistry of Energetic Materials Energy produced in the formation of explosive compounds is stored in the bonds of the molecule 16 By comparing the structures of several common explosives, one can see in Figure 2 2 that they all have similar compositions. Of particu lar importance are multiple bonds (double or triple bonds) and nitro (NO 2 ) groups which have double bonds between the nitrogen and one of the oxygen atoms. Due to these bonding characteristics, explosives usually have high densities as shown in Table 2 1. These characteristics are not just a coincidence, but are due to the largely endothermic reactions of the individual components (atoms or small molecules) upon formation of a larger molecule. The orientations of dipoles related to functional groups and resulting bond angles are also partial contributors to the high energy 60

PAGE 34

34 Figure 2 2. Molecular compositional structures of several common energetic materials. Table 2 1. The density and nitrogen percentage of several common energetic materials including those studied 60 Compound Density (g/cm 3 ) %N TNT 1.654 18.5 RDX 1.82 37.84 HMX (alpha) 1.87 37.83 Tetryl 1.73 24.39 Nitroglycerine 1.591 18.5 PETN 1.76 17.72 Energetic materials can be easily separated into different groups by their physical properties. In this study they will be simply classified as pure compounds and mixtures. Pure compounds are further subdivided into primary explosives, high explosives, and non explosive activation energy. A compound with a nearly spontaneous reaction or a highly unstable compound will be called a primary explosive such as lead azid e, Pb(N 3 ) 2 These compounds usually initiate the decomposition of a high explosive. They are highly unstable, due to a small

PAGE 35

35 activation energy, and do not release a large amount of energy compared to high explosives, but the energy released is in the ord An explosive with a larger activation energy is a high explosive like TNT. These compounds react rapidly, but slower than primary explosives, and release a great amount of energy forming a large volume of hot gasses. High explosive materials are then divided according to their molecular composition into nitro compounds (molecules containing nitro groups), nitrate esters (nitrogen containing derivatives of acids), nitramines (containing two to three alkyl (R) groups bonded to nitrogen), and salts of inorganic acids as seen in Figure 2 3 61 Non explosive compounds are those added to an explosive to stabilize or plasticize them making them easier to form. Common non explosive additives are diphenylamine (stabilizer) and polyisobutylene (plasticizer) 62 Figure 2 3. Characterization of the explosives investigated by chemical and explosive properties 5 Ammonium Nitrate

PAGE 36

36 Other explosive materials exist that are not easily categorized as the common explosives that are studied here. A fairly comm on homemade explosive is ammonium nitrate with fuel oil (ANFO) 63 The ammonium nitrate (fertilizer) is a nitrogen containing compound that when combined with fuel oil (diesel fuel for example), which donates oxygen to the compound, can be rather explosive. Relatively new energetic materials are liquid explosives lik e those identified in the attempt on the Heathrow Airport plot 64 These include triacetone triperoxide (TATP), hexamethyle ne triperoxide diamine (HMTD), and triam m oniotrinitrobenzene (TATB) and are quite unstable, see Table 2 2. Table 2 2. The explosive velocities and powers of several common energetic materials 63 Explosive Detonation velocity (m/s) Specific energy (kJ/kg) Explosive power (%) TNT 6900 870 116 RDX 8750 1394 169 HMX 9100 1387 169 Tetryl 7570 1200 132 PETN 8400 1220 167 TATB 7760 101 Nitroglycerine 7750 1139 170 HMTD 7350 897 1 00 TATP 4900 104 Picric Acid 7350 987 ANFO 3100/4000 982 2.2.1 TNT As shown in Figure 2 4 TNT has an aromatic or cyclic molecular structure with one methyl and three nitro groups spaced around the ring. The aromatic structure is based on a benzene ring that contains electrons that are delocalized, that is, the molecular orbitals are all interconnected and the electrons travel freely through all of the orbitals and are not attached to a particular carbon atom. Methyl and nitro groups are considered substituents because they replaced a hydrogen atom on the benzene ring and contribute characteristics to the molecule consistent with their own molecular structure 61 The methyl group is an electron

PAGE 37

37 donating group in that it supplies a partial electron to the benzene structure while the nitro groups are electron withdrawing. The positions of these substituents can vary so that TNT occurs in six different isomers such as 2, 3, 5 3, 4, 5 and 2, 4, 5 but 2, 4, 6 is the symmetric isomer, the most commonly used for explosives, and will be referred to as simply TNT 62 Figure 2 4. The molecular structure and molecular crystals of TNT present at room temperature. A) The TNT molecule conformation B as it appears in the molecular crystal. B) Th e ac or the (101) plane view of the monoclinic crystal unit cell. C) The bc or (011) plane view of the monoclinic molecular crystal. D) The ac plane view of the orthorhombic crystal. E) The bc plane view of the orthorhombic molecular crystal. The atoms in this and the next molecules are colored according to the atom type: gray is carbon, white is hydrogen, blue is nitrogen, and red is oxygen. A B C D E c a c

PAGE 38

38 Table 2 3 Unit cell parameters of crystalline TNT at room temperature 65 Monoclinic Orthorhombic Space Group Pc2 1 /c Pb2 1 a No. of Molecules 8 8 a 14.91 15.075 b 6.034 20.024 c 20.881 6.107 90 90 110.365 90 90 90 TNT is crystalline at room temperature in the form of pale yellow need l es 59 The molecular crystal structure is either monoclinic or orthorhombic where orthorhombic is a meta stable crystal phase 19,66 68 The cell parameters and crystal structures are shown in Table 2 3 and Figure 2 4 respectively. Both molecular crystal structures contain two conformatio ns of the TNT molecule, A and B. These molecules differ by the degree of torsion of the ortho NO 2 groups outside the plane of the benzene ring, as shown in Figure 2 5 and described in Table 2 4. The molecules are found in the crystals in a stacking seque nce of ABAB in the monoclinic structure and ABBA in the orthorhombic. A molecular crystal differs from the more common metal structures in that the molecules occupy lattice positions instead of individual atoms. Similar to metal crystals, molecular cryst als can contain defects such as vacancies, inclusions, and stacking faults, for example twinning. TNT is usually found as a monoclinic crystal with twinning defects where an orthorhombic stacking sequence is found within the monoclinic. The intermolecula r bonds in the crystal are non covalent in nature and are commonly reported as hydrogen bonding but are not traditional hydrogen bonds 68 70 TNT is most commonly used as a general explosive. It can be mixed with RDX or alumina creating different explosi ve properties. In addition, TNT also increases the explosive power in improvised or homemade explosives made from ammonium nitrate fuel oil (ANFO) and an ignition source 63

PAGE 39

39 Table 2 4. Torsion angles of the NO 2 groups in several different conformations of the TNT molecule. Torsion angle of NO 2 group () Molecule Ty pe 2 4 6 Optimized 29.00 0.74 23.57 Monoclinic A 23.00 10.96 20.30 Monoclinic B 17.80 15.90 25.96 Orthorhombic A 23.60 10.92 20.29 Orthorhombic B 25.69 15.56 18.29 Figure 2 5. Trinitrotoluene molecules in conformation A and B from the monoclinic molecular crystal superimposed on one another. Note the difference in torsion angles of the NO 2 groups for each conformation so that four oxygen atoms appear instead of two. 2.2.2 RDX RDX (1, 3, 5 Trinitro 1, 3, 5 triazacyclohexane) also known as hexo gen, cyclonite, or the primary ingredient in C 4 is a heterocyclic nitroamine. As shown in Figure 2 6, the molecule is composed of a ring of alternating carbon and nitrogen atoms with three substituent nitro groups. Unlike the aromatic molecular structur e of TNT, the carbons in the ring are saturated so that the electrons are localized to each bonded atom pair 61 RDX is considered a tertiary derivative of the

PAGE 40

40 ammonia molecule. This suggests that the chemistry of the molecule is dominated by the nitrogen atom creating the amine and the electron pair located on that nitrogen atom. Figure 2 7. The RDX molecular structure and crystal structure stable at room temperature. A) RDX molecule as it appears in the molecular structure. B) The ab or (101) plane view of the molecular crystal unit cell. C) The bc or (011) plane view of the unit cell. D) An off axis view of the unit cell to demonstrate the eight molecules present in the unit cell. Table 2 5. Unit cell parameters for the RDX molecular crystal at room temperature 71 Space Group Pbca No. of molecules 8 a 13.182 b 11.574 c 10.709 90 90 90 A B C D

PAGE 41

41 RDX is a white crystalline solid at room temperature and has an orthorhombic crystal structure 71 There are eight molecules in the unit cell of the Pbca space group with cell parameters shown in Table 2 5. The RDX molecule has a chair formation (shaped with the center of the molecular ring flat and the top and bottom of the ring extended in the upward and downward direction so it appears like a chair) with the nitro groups extending out of the plane of the chair as seen in Figure 2 6. RDX has been shown to have significant amounts of molecu lar defects in the form of vacancies which cause dislocations and dislocation pile ups in the crystal 72,73 RDX is most commonly known as a component in C 4, a plastic explosive. C 4 is moldable and easily formed into improvised explosive devices. RDX is also often mixed with TNT and desensitizers to make Composition B or without the desensitizers forming cyclotol 5 2.2.3 HMX HMX (1, 3, 5, 7 tetranitro 1, 3, 5, 7 tetraazacyclooctane) was originally discovered as a by product of the manufacturing of RDX 74 As seen in Figure 2 7, the molecule is composed of an eight sided ring of alternating nitrogen and carbon atoms. Each carbon h as two hydrogen molecules bonded outside the ring and each nitrogen atom is bonded to a nitro group. Similar to RDX the cyclic structure is saturated so all electrons in the ring are localized 61 HMX is also considered to be a t ertiary derivative of ammonia. There are four polymorphs or related crystal structures known for HMX 74 The beta form is stable at room temperature with a monoclinic crystal structure having 2 molecules in the unit cell. The space group is P2 1 /c with the cell parameters given in Table 2 6. In the molecular crystal, the HMX molecule is a puckered ring (a lternating atoms either above or below the ring plane) with the nitro groups placed in varying directions off of the ring. Research has shown that

PAGE 42

42 relatively small intermolecular distances exist between molecules which make intermolecular hydrogen bonds o f the form C HO 74 Figure 2 HMX that is the stable HMX molecule in the conformation seen in the crystal. B) The bc or (011) plane view of the molecular crystal unit cell. C) The ab or (110) plane view of the molecular crystal unit cell. HMX is a high molecular weight, very powerful, and relatively insensitive high explosive. It is used almost exclusively in military applications including as a detonator in nuclear weapons b c a A B C

PAGE 43

43 and in the moldable plastic form as solid rocket prop ellant. It is combined with TNT to make a mixture called octol 5 Table 2 6. Unit cell parameters for the beta HMX molecular crystal at room temperature 74 Space Group P2 1 /c No. of molecules 4 a 6.54 b 11.05 c 8.70 90 124.3 90 2.2.4 Tetryl Tetryl (2, 4, 6, N tetranitro N methylaniline) is the most closely related explosive to TNT of the studied explosives. It is composed of an aromatic carbon ring so that all the ring molecular orbitals are delocalized. There are three nitro groups substituted on the benzene ring and nitrogen atom bonded to another nitro group and a methyl group as seen in Figure 2 8. Even though Tetryl has an aromatic structure it is still considered a nitramine due to the three bonds o n the first nitrogen. Tetryl is seen as light yellow crystals at room temperature often appearing as needles 63 It has a monoclinic crystal structure of space group P2 1 /c like beta HMX 75 The other cell parameters are outlined in Table 2 7. The Tetryl molecule, in the molecular crystal, is stacked offset to the surroundin g molecules as seen in Figure 2 8. The nitro groups are non planar with the benzene ring. This is suspected to be due to the formation of an intermolecular hydrogen bond between the hydrogen bonded to a ring carbon and the oxygen atom of the nitroamine 75

PAGE 44

44 Figu re 2 9. The molecular structure and crystal structure of Tetryl in the stable phase at room temperature. A) The Tetryl molecule in the conformation seen in the crystal phase. B) The ac or (101) plane view or the molecular crystal unit cell. C) The bc o r (011) plane view of the unit cell. Table 2 7. Unit cell parameters for the Tetryl molecular crystal at room temperature 75 Space Group P21/c No. of Molecules 4 a 14.129 b 7.374 c 10.614 90 95.07 90 A B C

PAGE 45

45 Tetryl is a fairly sensitive medium explosive velocity material as demonstrated in Table 2 2. It is used primarily alone in detonators and explosive booster charges. At one time it was combined with TNT to create tetrytol that was us ed in landmines 5 2.2.5 PETN PET N (1, 3 dinitrato 2, 2 bis(nitr o methyl)propane) also known as pentaerythritol tetranitrate is the most unstable of the studied explosives. As shown in Figure 2 9, PETN has a unique molecular structure. There is one central carbon atom that forms four bo nds to separate methylene groups. Each methylene group is also bonded to an oxygen atom that is then bonded to a nitro group. PETN is considered an ester of a tetrahydroxylic alcohol where the OH group of tetrahydroxylic alcohol was replaced by the nitr o group. They are characterized by participating in hydrogen bonding as acceptors but cannot hydrogen bond to each other because they have little ability to act as hydrogen bond donors 5,61 Table 2 8. Unit cell parameters of the PETN molecular crystal at room temperature 76 Space Group P 42 1 c No. of Molecules 4 a 9.380 b 9.380 c 6.710 90 90 90 A white or colorless crystal at room temperature, PETN is commonly used in detonation cords 63 It has a tetragonal crystal structure with a P 42 1 c space group including two molecules per unit cell; see Table 2 8 for other cell parameters 76,77 The PETN molecule in the crystal appears so that each nitro group makes the corner of a square. As shown in Fi gure 2 9 the carbon atoms from the methyl groups are not in the same plane to each other. Research has shown that PETN is prone to edge dislocations and general dislocation planes due to the inherent slip plane in the crystal structure 78

PAGE 46

46 Figure 2 10. The molecular structure and molecular crystal structure of PETN. A) The P ETN molecule conformation as it appears in the crystal. B) The ab or (110) plane view of the molecular crystal. C) The bc or (011) plane view of the molecular crystal. 2.2.6 Nitroglycerine Nitroglycerine, also called glycerol trinitrate, is a nitrate es ter similar to RDX and HMX 62 It is made from the nitration of pure glycerine resulting in the molecule shown in Figure 2 2, which is a branched linear chain of three carbons with H 2 ONO 2 attached to the end carbons and HONO 2 to the center carbon. This is unlike the other nitrate esters s tudied which are cyclic. Nitroglycerine is a yellow or cream colored oil at room temperature with a fairly high vapor A B C

PAGE 47

47 pressure 63 Therefore there is little information on the crystal structure or intermolecular bonding in the solid state, although a great amount of research has been performed concerning the medicina l uses of nitroglycerine. Nitroglycerin is the primary component of dynamite and the major component of gelatinous industrial explosives. It is relatively stable but incredibly sensitive to shock. There aight, and gelatin 62 The compone dynamite have been described at the beginning of Chapter 2 while the straight and gelatin both contain nitroglycerin, sodium nitrate, a carbon based combustible material such as wood, and antacid (CaCO 3 ). Gelatin dynamite contains nitrocell ulose that is bonded to the nitroglycerine making a gelatin. 2.3 Optical Properties of Energetic Materials It is shown in Section 2.2 that energetic materials have interesting chemical compositions and properties. This also yields intriguing optical prope rties for the materials studied. The optical properties of energetic materials have been studied in nearly all areas of the electromagnetic spectrum. These studies have been mostly used for the characterization and research of explosive compounds and not necessarily for detection. A thorough investigation of these prior studies contributes a large knowledge base for the research. 2.3.1 Ultraviolet and Visible Range (180 750nm) The ultraviolet (UV) region of the electromagnetic spectrum is commonly defin ed from 180 nm to 400 nm and the visible region from 400 nm to 750 nm 62 These two regions are often measured together due to the availability of detectors and light sources that can accommodate both regions. The energy of electromagnetic radiation in this range is in the order of the band gaps or difference in energy levels in materials. Therefore, several spectroscopy techniques in this region are based on the absorption of the electromagnetic radiation by a material 79 This

PAGE 48

48 absorption is caused by an energy transfer from the incident photon (electromagnetic radiation) to an electron in the sample causing the electron to move to an unoccupied, higher energy state. This phenomenon will be discussed in greater detail in C hapter 3. A spectrometer of this type uses an UV Visible light source such as a Xe, Hg, or a Deuterium lamp 2 The light source may also be coupled with a stronger visible spectrum emitter for example a tungsten halogen lamp that emits predominantly the r ed part of the spectrum. In a commonly used spectrometer, the UV Visible light beam is split into two equal parts prior to the sample chamber. One light beam is directed through the sample in while the other light beam does not and is used as a reference beam. A detector on the other side of the spectrometer from the light source records the transmittance of the light through the sample and compares that to the reference light beam. The data is usually analyzed using the Beer Lambert law (Equation 2 1) which states that the absorbance, A, of a material is dependant on the concentration of the absorbing species in the sample, c, and the path length through the sample, (2 1) 62 Most spectra are displayed as the molar extinction coefficient or transmittance versus the wavelength. The samples used are generally solutions of the desired material in a solvent such as 95% ethano l (5% water), acetonitrile, or acetone. When a solution is used the solvent is placed in front of the reference beam to subtract the absorption of the solvent from that of the sample solution. This being said the solvent may have an effect on the absorpt ion of the sample due to interactions between the sample and solvent or its absorption characteristics. An educated choice of solvents is necessary to reduce the interactions as much as possible 79

PAGE 49

49 TNT, RDX, HMX, Tetryl, PETN and Nitroglycerine are defined as organic molecules and have certain characteristics in the ultraviolet (UV) and possibly the visible region. These characteristics are due to the bonding in the organic molecules 79,80 The absorption spectra of materials usually contain broad peaks due to the vibrational and rotational states in the molecules that superimpose on the electron transitions and broaden them. Du e the great amount of information available, UV visible spectrophotometry has been used for the chemical analysis of explosive materials. A study by Schroeder and colleagues investigated the UV absorption of several different explosives 21 The samples were in solutions of ethyl alcohol of various concentrations. Cyclic nitro based explosives namely TNT, RDX, HMX, and Tetryl demonstrated absorption maxima from 213 nm to 229 nm as shown in Table 2 8. PETN and NG were also measured but th ey did not demonstrate significant absorption in the UV region. The UV absorption of solutions with several solvents, performed by a separate group on TNT had an absorption maximum at 225 nm or 233 nm (depending on the solvent) 81 These results were comparable to the findings of Schroeder 21 Similar meas urements performed by Kamlet et al. and Conduit et al. confirmed absorption maxima for TNT at 227 nm and 232 nm in solution with methanol, and ethanol and water respectively 82,83 TNT was measured by Usachev and colleagues for UV absorption in the gaseous or vapor state 22 They used a thermal ta pe to heat the sample solution creating a vapor and measured this in a UV Vis absorption spectrometer and a Cathode Ring Down Spectrometer. Both techniques measure a broad peak with a maximum at 216 nm for the vapor TNT sample which is shifted to shorter wavelengths (blue shift) from the absorption maximum observed in solution. UV and visible absorption of solid state explosives is a difficult task and reported sparingly in the literature. The UV absorption of HMX, RDX, and PETN has been measured in KCL

PAGE 50

5 0 p ressed pellets 26 Absorption maxima were found at 227 nm for HMX, 243 nm for RDX, and 192 nm for PETN. Crystals of HMX and RDX have been measured in reflectance yielding the previously mentioned high energy peaks but also a feature at 35 5 nm 84 In scanning the literature it is noted that essentially all spectrometric measurements of high energetic materials seem to yield absorption peaks only in the 190 230 nm (UV) range and not in the visible spectrum even though UV Visible instruments were used. This is an important observation considering the results reported in Chapter 5. Specifically the present research rev ealed optical structure in the blue end of the visible spectrum in addition to the UV peaks. Table 2 9. Ultraviolet absorption maxima for TNT, RDX, HMX, and PETN as reported in the literature. Solvents are given for those measured in solution. The dat a of the present research are not included. TNT (nm) RDX (nm) HMX (nm) PETN (nm) Acetonitrile 193, 234 26 85 195,236 86 190, 227 85 192, 260, 290 26 Ethanol (95%) 225 81 227 21 202, 236 85 213 21 201 85 228.5 21 Methanol 227 82 204, 234 85 Water 233 81 KCl pellets 190, 243 26 194, 233 26 193 26 Thin Films 209, 248 87 224, 247 87 Single Crystals 220, 255 84 202, 245 84 Gas/Vapor 216, 244 22 2.3.2 Infrared Properties (0.750 The infrared region is generally defined as the waveleng th range from 750 nm to 1 mm 58 Frequencies in the infrared region are of the same order of magnitude as molecular vibrational frequencies 62 The molecular vibrational frequencies or modes are based on the atom and spring molecular model 58 Atoms in a molecule are bonded, demonstrated by springs; these springs have a defined vibratio nal frequency. When the infrared radiation frequency is equal to the vibration frequency of the spring the frequency is absorbed whereas all other frequencies are transmitted 62

PAGE 51

51 These vibrations can be associated with the entire molecule or specific functional groups within the molecu le 62 For example, TNT has a vi brational mode related to the entire molecule and vibrational modes associated with the NO 2 and CH 3 functional groups. Most infrared spectra are displayed using wave numbers. The wave number is in the units of inverse s the wavelength in microns (see Equation 2 2). The (2 2) vibrations associated with the entire molecule usually yield absorption bands at or below 1300 cm 1 f the molecule and provide a fingerprint. Vibrations related to the functional groups have absorption bands in the region greater than 1300 cm 1 specific functional groups within the molecule. The presence of these bands aid in the identification of unknown molecules and are recorded in several tables. 2.3.2.1 Fourier Transform Infrared Spectroscopy As mentioned in Section 2.2, the energetic materials are composed of nitro compoun ds, nitrate esters, and nitramines. There are IR bands assigned to the particular explosive groups as follows: symmetric trinitro compounds (such as TNT) at 1081 cm 1 nitramines (RDX and HMX) at 1282 cm 1 and nitrates in general (such as esters, PETN) a t 1666, 1282, and 833 cm 1 These bands are assigned due to characteristic vibrations (symmetric and asymmetric stretching etc.) for the entire molecule 88 The wavenumber of the band may shift slightly due to substitutions on the molecule. Bands that are related to the functional groups, specifically the NO 2 groups, in the studied explosives have scissor frequencies at 734, 825, and 907 cm 1 and stretch frequencies at 1533 and 1545 cm 1 Molecules with C N bonds have stretching bands at 938, and 907 cm 1 N

PAGE 52

52 N stretching at 1080 and 1045 cm 1 and O N O bending at 930 cm 1 These band frequencies create fingerprints for the individual explosive molecules 24,25,77,83,89 2.3.2.2 Raman Spectroscopy Raman spectroscopy measures the Stokes shift in wavelength of a monochromatic incident light by the inelastic collisions with electrons in a m aterial. A Raman spectrum is characteristic of the material and is fairly well defined for the solid phase for many explosive materials such as TNT, HMX, and RDX 35,90,91 Researchers have used near infrared and red photon sources to obtain Raman spectra. Fourier transform can be used to enhance the sensitivity and resolution of Raman shifts 92 An FTIR, modified for Raman measurements is commonly used. TNT, RDX, and HMX have strong bands in the same regions, 1360 1309 cm 1 884 824 cm 1 and 157 148 cm 1 The smallest wavenumber bands are resolved due to the Fourier transform and specialized filters. Measurements of TNT vapors and solutions were measured using surface enhanced resonance Raman scattering (SERRS) 39 SERRS involves using an analyte, that contains a component that fluoresces, absorbed onto a roughened metal surface. Since TNT does not fluoresce in the specified region and does not readily absorb onto a metal surface the SERRS process needed to be modified. An azo dye was used to modify the TNT creating a highly colored TNT derivative with the ability for strong interactions with the metal surface. Raman spectra were recorded f rom TNT in solution and in vapor form that had fingerprints nearly identical to those taken of solid TNT. Raman was also used to characterize explosives in the shocked form 91 This yielded spectra of detonated explosives and detonation products which allows further understanding of the character istics of the explosive products.

PAGE 53

53 2.4 Summary The chemistry and optical properties of energetic materials are unique. Most importantly, it has been shown that explosive materials have significant absorption in the UV range of the spectrum. The different ial reflectometer has the ability to probe this property and provides the absorption of the solid phase of energetic materials. Chapter 3 will explain the electronic structure of the organic molecules studied and how this affects the presented research.

PAGE 54

54 CHAPTER 3 ELECTRONIC THEORY OF ENERGETIC MATERIALS The present research is concerned with the optical properties of energetic materials which are generally organic materials. The electronic structure of these materials is related to inorganic materials s uch as silicon, but unique due to the presence of molecules with covalently bonded atoms of several different elements. This unique structure has many implications on the optical properties of organic materials. Therefore the electronic structure that gi ves rise to these optical properties should be addressed. 3.1 Electronic Structure of Explosives A depiction of the molecular orbital energy diagram of mononitrobenzene (MNB) based on the molecular bonding and created by the author, is seen in Figure 3 1. The molecular orbital energy diagram for TNT is more complicated and diffucult to represent simply. It is clearly quite complicated with interacting orbitals and many molecular orbitals that do not interact. The molecular orbital transitions in this sys tem can be between both the interacting and non interacting molecular orbitals and are potentially rather complicated. This energy diagram can be compared to that of materials with only one atomic species, such as silicon, where the highest energy state o ccupied is the upper part of the valence band and the lowest unoccupied energy state is the lower end of the conduction band 58 These energy states are separated by an energy gap like that 1. In organic molecules, the bonding and non bonding orbitals are considered similar to the valence band but instead are termed the highest occupied molecular orbital (HOMO) 79 The lowest unoccupied molecular orbital (LUMO) is analogous to the conduction band and is composed of the anti bonding orbitals. Electrons within the molecule can be localized on a particular atom if they are non bonding, shar ed between several atoms in bonding orbitals, or shared between a small number of atoms 93

PAGE 55

55 This is in contrast to an inorganic material where the electrons are either shared between the nearest neighbor atoms, as in covalently bonded materials, or freely distributed thr oughout the material, as in metals, and therefore provide more complicated electronic transitions. Figure 3 1. Molecular orbital energy diagram of mononitrobenzene (MNB) demonstrating the complexity of energy diagrams of large molecules (created by auth or) The molecular orbital energy diagram shown in Figure 3 1 is based on the combination of the atomic orbitals of several different atoms. This is performed through the linear combination of atomic orbitals which is simply the addition and subtraction of the wave functions that describe the two atoms. Quantum mechanics requires that the combination of two atomic orbitals yields two molecular orbitals. This is the origin of the anti bonding molecular orbitals that act as the conduction band. It is impo rtant to note that molecular orbitals are characterized by their symmetry. Similar to crystalline materials, all molecules demonstrate elements of symmetry. The symmetry operations included in the molecule are classified into point groups like the space groups for

PAGE 56

56 crystalline materials. The influence of the symmetry operators on the molecule determines the symmetry of its molecular orbitals. This symmetry is reflected in the wave equations that describe the molecular orbitals. As will be shown in Secti on 3.2 this has a large effect on the allowed molecular orbital transitions. 3.2 Photon Interactions The complex molecular orbital energy diagram in Figure 3 1 demonstrates that the possible electronic interactions are not straightforward for organic molec ules. A discussion of this begins with the interaction of photons with the material. If photons have a large enough energy to promote an electron to a higher energy orbital, the photons are absorbed. This absorption is not only determined by the energy of the incident photon but on several criteria including molecular orbital symmetry among others 79 The theoretical treatment of the absorption of photons begins with calculating the probabili ty of the absorption between two orbitals 93 This is given in Equation 3 1 where B if (3 1) denotes the absorption probability from an initial (i) to a final (f) state, G f is the statistical weight of the final absorption state, and D if th e dipole strength. The statistical weight of the absorption state is defined by the number of degenerate wave functions to which absorption can occur. An absorption probability can be transformed from Equation 3 1 according to Mulliken to give a measure of the intensity given in Equation 3 2. The oscillator strength (f) (3 2) is an indication of The dipole strength, D if can be further obtained by Equation 3 3 (3 3)

PAGE 57

57 showing that it is proportional to the intensity ( I ). The total wave functions of the initial and i f e volume element product. The rotational and vibrational wave functions are assumed to not be a factor due to their long time scale compared to electron transitions, so that the total wave functions are reduced as shown in Equation 3 4. (3 4) This equation provides the basis of quantum mechanical approximations of molecular orbital reduced into a series of one electron orbitals (functions) ; se cond, only one electron is excited in the transition ; j are the same in the ground and excited states. The first and third assumptions are significant but they greatly reduce the problem to a manageable size Equation 3 5 shows Equation 3 4 reduced according to the stated assumptions. When Equation 3 5 (3 5) is zero, then the transition is forbidden and according to the theory should not occur. The symmetry properties of the molecular orbital expressed in the wave function can determine whether Equation 3 5 goes to zero. These are summarized in selection rules. The first selection rule only concerns molecules with a center of symmetry. This rule gives that a transition can only occur between an antisymmetric orbital and a symmetrical orbital or vice versa and not between two antisymmetric or symmetric orbitals. The multiplicity of states gives that transitions from singlet to triplet states are forbidden while singlet to singlet transitions are allowed. This is due to the fact that transitions which require a change in spin are not allowed. The last selection rule is the most intriguing as it involves the symmetry of the molecular states. A transition is forbidden if i f does not convert, as the irreducible

PAGE 58

58 representation, in the same way as at least one of the components of the dipole strength vector (M x M y or M z i f is an even function and M x is also even than the transition is allowed howeve r, if M x is odd then it is not allowed. Or, the transition is allowed if a direct i f contains the totally symmetric representation. This illustrates the importance of the molecular orbital shape and symmetry involved in a particular transition. In addition, if the symmetry of an orbital is changed due to bonding with another molecule or electrostatic interactions a forbidden transition can be allowed, although usually at a low intensity. It is important to note that many electron transitions are considered forbidden if they are expressed as one electron wave function products, however, the transition can be allowed if a linear combination of these orbitals are made. This is due to the non vanishing probability of a many electron excitation. Vibrational motion in a molecule is also a consideration in the electron transitions. These motions can be described as symmetrical or a nti symmetrical and are at a time scale significantly longer than electronic transitions. Reg ardless, there are occurrences where vibrational interaction causes a forbidden electronic transition to become allowed. The weakly allowed nature of this transition is present when the electronic excitation is accompanied with a vibrational motion in the particular direction of the orbital. This is analogous to silicon where absorption of a phonon (lattice vibration) is required for an electron to be promoted to the conduction band across the indirect band gap. 3.3 Summary The molecular orbitals of a lar ge molecule, such as the energetic materials investigated, are complex. This leads to intricate molecular orbital transitions that may be difficult to interpret. The research performed investigates these MO transitions and uses the presented theory to ex plain the origin of the reported absorption.

PAGE 59

59 CHAPTER 4 EXPERIMENTAL PROCEDURE This chapter describes the methods and procedures used in the procurement of data. 4.1 Ultraviolet and Visible Techniques The UV Visible properties are measured using two different instruments. A Differential Reflectometer was used to find the characteristic absorption of the energetic materials that had not been previously reported. The DRS measurements were then complimented using a UV Visible spectrophotometer in transmission and reflection modes. 4.1.1 Differential R eflection Spectroscopy (DRS) The unique optical properties of explosive materials give several options for research. Due to previous research in the field of UV and visible optical properties of semiconductors and metals, the researchers decided to utiliz e this region. Differential Reflection Spectrometry is the main technique used for the presented research. The DRS was originally developed to 10,94 96 Two metal samples containing different amounts of alloying are placed side by side with virtually no space between them. A beam of light is scanned between the two samples as a monochromator scans through the appropriate wavelengths. The differential reflectogram (DR) produced displays the normalized difference in reflectivity versus the wavelength. Features in this DR represent energies where photons are absorbed by electrons within the material and can show how the alloying effects these transitions. The DRS probes the first few atomic layers of a material and reveals the electronic structure near the Fermi surface. 4.1.1.1 Instrument The concept of the DRS is straightforward and in turn the basic instrument is relatively simple in construction as seen in Figure 4 1. The experiments performed are in the UV and

PAGE 60

60 visible range (200 800 nm) but the instrument can also be used through the infrared range with proper adjustments of the light source, grating, and detector. A high pressur e xenon lamp is the unpolarized UV to visible light source. This is placed in front of a scanning monochromator that individualizes the wavelengths of the light and, while continually varying the wavelength, directs the light beam onto a vibrating mirror. The vibrating mirror oscillates the beam to two spots (close in distance to each other) on the sample, at near normal incidence. This creates an area of between 2 x 2 mm 2 and 2 x 4 mm 2 determined by the degree of vibration of the mirror. A lock in ampl ifier is tuned to the frequency of the vibrating mirror. The reflected light from the two sample spots are collected by a mirror and directed onto the face of a photomultiplier tube (PMT) with a light diffuser on the front to prevent any effects due to th e varying sensitivity across the PMT face 10,94 97 Figure 4 1. D iff erential reflection spectrometer used in the research. The different colo rs and shapes in the samples within the inset indicate the inhomogeneous surf aces of the sample allowing differential measurements. The output signal with a direct current component which can be observed by an oscilloscope is split into two channels. One channel contains a low pass filter that removes a

PAGE 61

61 frequency specific square wave from the signal leavi ng the average in reflectivity (Equation 4 1), (4 1) which is held constant by a servo ( at 10 volts) that determines the high voltage applied to the PMT. The frequency of t his square wave is equal to the vibrational frequency of the mirror, in this case 60 Hz. The second channel is connected to the lock in amplifier that gives an output 2. (4 2) A divider circuit forms the desired ratio of the two signals yielding The resulting spectrum is termed a differential reflectogram (DR) and displays the normalized difference in reflectivity versus t he wavelength. The normalized difference in reflectivity, (4 3) is calculated electronically from the output of the PMT The normalization of the signal has several benefits most importantly elim inating possible fluctuations of the line voltage and spectral effects of the optics in the instrument including the intensity changes of the light source. This signal is then sent to the computer by way of a data acquisition card where it is recorded alo ng with the corresponding wavelength 94 96 4.1.1.2 Theory and Interpretation Differential Reflectometry is based on the absorption of photons of a particular energy by electrons in a material similar to UV Visible spectrophotometry. As explained in Sect ion 2.5.1, the DRS measures the normalized difference in reflectivity. This difference in reflectivity yields, 2 that is, the imaginary part of the complex dielectric constant,

PAGE 62

62 (4 4) where 1 (4 5) of the complex dielectric constant (polarization) and the imaginary part, 2 (4 6) are given where n is the (real) index of refraction and k is the damping constant. The above parameters are material properties and dependent on the wavelength of light. The difference in reflectivity could or iginate from several sources. For the reported work, the sample (e.g. TNT) was usually in the form of crystals that were placed on an optically absorbent carbon pad. The crystals are not evenly distributed on the carbon pad that is, the amount or concent ration of the TNT on two spots may not be the same resulting in a difference in the reflection of photons. Alternately, the sample crystals are small and not aligned so that an area may contain several different crystal faces in varying directions. Incid ent light will be reflected off of these crystal faces where the index of refraction is different for each crystal face 98 4.1.1.3 Line Shape analysis The DR measured by the DRS, as mentioned in Section 4.1.1, gives the normalized difference in reflectivity. The absorption of the sample can be found by means of a line shape analysis of the DR. A technique of a line shape analysis has been well documented and finds the exact transition wavelengths for photon absorption 94,97,99 This analysis requires the optical constants, n and k, for the sample materials over the wavelength range of the measured spectra. For energetic materials, the optical constants are not well reported and when they are known it is only for one wavelength, usually the sodium D line. Therefore a line shape analysis of the me asured data based on the shape of the transition and not the calculated values was used in the research.

PAGE 63

63 Figure 4 2. Curves representing the functio n of the photon energy (s) and an o ptical parameter 97 Previous investigations involving a line shape analysis revealed that all differential 2 1 2 type shape displays a maximum (or minimum) at the transition energy for electrons, see Figure 4 1 t the transition energy. In general, all measured curves are combinations of these two extremes for example, those shown in Figure 4 2. An exact line shape analysis has not been attempted in the presen t work since certain spectral parameters (Seraphin coefficients) are not yet known for energetic materials 100 This is of no detriment because the measurements of the Seraphin

PAGE 64

64 coefficients of energetic materials are not the goal o f this research. Instead, the general line shape analysis of energetic materials is used for their identification. 4.1.1.4 Sample Preparation and Operating Parameters Solid energetic materials were placed on a black carbon pad that is adhesive on both sid es. One side was attached to a glass slide for mechanical support and the solid sample placed on the opposite exposed side. The carbon pad, manufactured for scanning electron microscopy, was chosen due to its featureless optical absorptive properties in the UV and visible range, after a survey of several different substrates. Samples of energetic materials were received from several sources in different forms. The TNT received from Chem Service Inc. was in solid powder form with 30% included water for s hipping. A solution of TNT in acetonitrile was made and the solution is deposited on a thoroughly cleaned glass slide or the entire bottle of solution is placed in a low temperature vacuum oven. For the glass slide preparation, the solution was continual ly deposited onto the slide at approximately 10 minute time intervals until a critical concentration was reached and crystals of desired size were formed. While the samples for vacuum oven preparation were placed in an oven set to a temperature lower than the boiling point of the solvent and a vacuum of 25 mmHg was constantly pulled using a small mechanical vacuum pump. In both cases, the solution was allowed to evaporate and dehydrate the TNT resulting in small pale yellow dendritic crystals. The proces s can be repeated to insure better purification as indicated in the literature, but a chemical analysis on the above prepared sample yielded a purity of 97% making additional rounds essentially unnecessary 101 Other energetic materials such as HMX, R DX, and PETN are received from Accustandard in a solution with methanol, ethanol, acetonitrile, or a combination of these solvents. For experiments requiring solid samples the deposition dehydration procedures outlined above are used.

PAGE 65

65 Additional samples w Federal Bureau of Investigation (FBI). These samples include TNT, C 4 (RDX and plasticizer), PETN, Detasheet, Atlas, and other common commercial energetic materials. These were general ly used as received in the DRS where small amounts of the samples were placed on black carbon pads. The prepared sample was placed at the sample holder of the differential reflectometer. The monochromator was then set between 450 nm to 500 nm so the light beam is visible to the human eye and the sample position was adjusted so that the beam is on the energetic material sample crystals. An oscilloscope that was connected to the output of the photomultiplier tube can be used to find an optimum spot on the s ample. The optimum spot was defined as one that produces a near square waveform on the oscilloscope showing that there is a large enough difference in signal between the two oscillating light beams to yield a large response without going into saturation. After the instrument was aligned, the monochromator was scrolled back to 200 nm. The measurement was taken by scanning the monochromator in the decreasing energy direction (increasing wavelength) and the resulting data was displayed as the wavelength ver sus the normalized difference in reflectivity. This sequence was repeated for each measurement whether it is a different sample or simply a different spot on the same sample. Generally, three sequential measurements were taken for each sample (same spot on the sample) to ensure repeatability of the results. One experiment performed on the DRS was to determine the effect of UV light exposure on the DR of a TNT sample. For this experiment the instrument was used as one of the UV light source to limit sampl e handling. The monochromator was set to a specified wavelength in the UV range (250 nm 350 nm), a protective shutter was removed, and the sample was exposed to

PAGE 66

66 the light for a specified amount of time (from 30 seconds to 15 minutes). After each UV light exposure the above experimental procedure was followed. The monochromator has different scanning speeds that can be used to increase the resolution of the measurements. For normal measurements, 200 nm per minute is used due to the relatively high resol ution without significantly increasing the measurement time; 100 nm per minute is a higher resolution while 500 nm per minute provides a low resolution speed used primarily for scrolling back the monochromator. The sensitivity of the DRS can also be adjus ted by changing the voltage of the signal channel for the lock in amplifier between 0.1 and 10 mV. Most measurements are made using 5 or 10 mV which is the lowest sensitivity settings. 4.1.2 U ltraviolet Visible Spectrophotometery The UV Visible spectropho tometer is most often used in the transmission mode where the incident light passes through a sample in solution. Four different instruments were used throughout the research: Perkin Elmer, Hitachi, and two separate Varian Cary spectrometers. A He lamp i s used in conjunction with a scanning monochromator to individualize the wavelengths of light. The light then passes through a beam splitter which directs both beams through the sample chamber. A sample is placed in one beam and the reference (the pure s olvent of the sample solution) in the other beam within the sample chamber. The transmitted light is collected, digitalized, and calculated as the absorbance. Generally, the path length for the measurement is 0.1 cm. 4.1.2.1 Transmission UV Visible spec trophotometry For transmission measurements, the instrument is turned on and properly warmed up (approximately 20 minutes). A zero measurement is taken with no samples in the sample chamber for calibration. The experimental parameters, such as the integr ation time speed and wavelength range, are entered into the operating computer and a background measurement is

PAGE 67

67 made, which will be subtracted from all subsequent measurements in the experiment. Then a reference sample of the solvent is placed in the sampl e chamber, and a baseline measurement is made. To measure an experimental sample, the sample solution is deposited in a quartz cuvette and placed in the sample spot (first beam) of the sample chamber and a spectrum is taken. This is repeated for each sam ple in the experiment with proper cleaning of the sample cuvette in between and a new baseline measurement between chemically different samples (not concentration variations). When the concentration of the solution is continuously varied for an experiment then the lowest concentration solution is measured at first and the highest concentration last. This eliminates the potential for error because the next measurement made is of the same solution at a significantly higher concentration. 4.1.2.2 Reflectanc e UV Visible spectrophotometry A reflectance stage can be installed into the UV Visible spectrophotometer to measure solid samples. The reflectance stage is comprised of a mirror to direct the incident light onto the sample holder at a set angle and a sec ond mirror to direct the reflected light towards the detector. This angle can be varied from 45 to 90 degrees to the detector. The reference sample spot is left empty or a density filter can be used to increase the sensitivity. The reflected light from the sample collected by the detector is calculated in virtually the same manner as the transmission, that is, where I 0 is the incident light intensity is and I R is the intensity of the reflected light. (4 5) 4.1.2. 3 Sample Preparation A solution of the sample material is required for UV visible transmission spectrophotometry. If the hydrated TNT is to be measured, the dehydrating process described in Section 4.1.1 is used. Then the solid sample is carefully massed out and placed into a clean

PAGE 68

68 sample bottle and the specified amount of the appropriate solvent is added by a micropipette. The Accustandard samples are received as a solution therefore they are diluted as needed or used as received. A UV visible reflectan ce spectrophotometry sample is virtually the same as a DRS sample however, the black carbon pad is much larger to account for the lager spot size in the spectrometer. 4.1.3 Infrared Spectroscopy The infrared properties of the energetic materials were als o briefly explored to compliment the DRS results. Many instruments exist to investigate the infrared properties such as Raman spectroscopy and Fourier transform infrared spectroscopy (FTIR). For the reported research an FTIR investigated the fingerprint region of the spectrum and the bonds of the molecules. The FTIR instrument used was a Thremo Magna with microreflectance attachment. The FTIR included a microscope attachment that was used in reflection mode. A solid sample of the energetic material was placed on a stainless steel plate and positioned under the microscope objective lens. Stainless steel was the substrate for the samples because it is optically reflective and inactive (no features) in the infrared region. The microscope was focused so th at a white light spot imitating the position of the infrared beam illuminated the area of interest. Before the spectrum is taken, the visible light is turned off to prevent interference with the response of the sample material. When a measurement is perf ormed the incident beam is directed through the objective lens and the infrared light reflected from the sample travels back through the objective lens of the microscope to the detector and signal processing is performed. The resulting data is presented a s the wavenumber versus the percent of reflection.

PAGE 69

69 4.2 Chemical and Crystalline Structure Analysis To ensure the integrity of the reported results, great lengths were taken to determine the purity and crystal structure of the studied energetic materials. A confirmation of the purity of the energetic materials was performed by High Pressure Liquid Chromatography and Mass Spectroscopy (HPLC/MS). The crystal structure of the samples was determined by differential scanning calorimetry (DSC) where applicable. These crystals were also characterized by optical microscopy and optical profilometry. 4.2.1 High Performance Liquid Chromatography and Mass Spectrometry (HPLC/MS) The samples were analyzed by Dr. Jodie Johnson in the Spectroscopy Services of the chemist ry department. Solutions of TNT and acetonitrile were submitted for analysis where a High Performance Liquid Chromatography (HPLC) coupled with Mass Spectrometry (MS) was performed. HPLC is a technique based on chromatography, where a mobile phase is mov ed through a sample to separate out the components. In HPLC the mobile phase was a solvent or combination of solvents that are mixed and released into a chamber. The sample being analyzed was injected into the system using a small syringe and the solvent was released from the chamber through the sample into a column. The sample was then separated into its components and displayed in a chromatograph where each is separated by the time it took to flow through the chamber. Then the resulting compounds were analyzed using UV Visible absorption by recording the absorption spectra. After this analysis a definite determination of the sample structure was obtained by removing the separated components from the HPLC column and placing them (separately) in a mass spectrometer 102 4.2.2 Differential Scannin g Calorimetry (DSC) The crystal structure of solid TNT samples was investigated using the differential scanning calorimeter (DSC). The DCS is a technique that determines the thermal transitions in a material

PAGE 70

70 such as the melting point or glass transition. A DSC is composed of two plates in a controlled atmosphere where one plate holds the sample and the other is empty. Each plate was set on individual heaters that are controlled by a computer. As the temperature of the heaters was raised (scanned), the a mount of heat required for the sample pan heater to maintain the same temperature as the reference may change at certain temperatures. The resulting difference in sample pan heater heat output was recorded as a function of the temperature. This was displ ayed as a graph of the heat output versus temperature indicating peaks where thermal transitions in the sample occurred. 4.2.3 Optical Microscopy Sample crystals were analyzed, measured, and digitally photographed using an Olympus digital optical microscop e. 4.2.4 Optical Profilometry The shape and morphology of the samples was further characterized by optical profilometry. A Veeco Wyko profiler housed in MAIC was used for the measurements. Optical profilometry produced a measurement of the topography an d morphology of a sample using the optical interference pattern produced from the interaction of the light source and the sample. The instrument was constructed like an optical microscope with an objective lens where the incident light, beam splitter, ref erence mirror, and detector are housed. A light source, usually a red to near infrared laser, was split and one beam was directed at the sample and the second was directed on the reference mirror that is perfectly flat. The difference in the phase and in tensity of the light reflected from the sample and the reference mirror created an interference pattern. A piezoelectric ceramic was used to slowly move the reference mirror (in the z direction) and the interference patterns at several positions were reco rded. When the movement was complete the

PAGE 71

71 computer analyzed the interference patterns and constructed a three dimensional false color image of the sample topography. 4.2.5 Particle Size analysis Particle size analysis was performed on solutions of the ener getic materials in appropriate solvents using a Nanotrac autosampler which implements dynamic light scattering. In a solution of particles and solvent, the particles in the solution experience Brownian motion due to random collisions with thermally excite d solvent molecules. The resulting direction and velocity of the particles are random but the velocity distribution over a period of time approaches a known functional form representative of the particle size distribution. A laser diode with a wavelength of 780 nm was directed through a probe that is inserted directly into the solution. Some of the incident light was reflected off the sapphire lens that separates the probe from the solution, and was used as a reference frequency signal. The incident lig ht of the specified frequency was scattered by the particles in the solution and frequency shifted due to the Doppler effect. The frequency shift was in agreement with the velocity of the particle which was recorded by a photodetector in the probe after b eing recombined with the reference light. A Fast Fourier Transform (FFT) was performed on the resulting interference spectrum that yielded the particle distribution. A thermocouple in the probe records the solution temperature to account for thermal exci tation in the velocity calculations. The Nanotrac was able to measure particles larger than 0.8 nm for fairly high solids loading (concentrations) solutions and 3 nm for low solids loading solutions. 4.3 Computational Chemistry The molecular orbital struct ure of the energetic materials where explored using computation chemistry or molecular modeling using the computer software CAChe (computer

PAGE 72

72 assisted chemistry) by Fujitsu. While these were not experimental results, they are included within this section in order to explain the procedure for the computational methods. The molecule of the energetic material is drawn in the computer interface using a molecule builder tool atom by atom and then connecting the atoms with the proper bonds. Valence configuration of the atoms and hybridization of the bonds were then defined by the computer. Other parameters such as, inserting a double bond on one oxygen in the nitro group, were corrected by the user. Then the geometry of the molecule was optimized using the semi empirical PM3 (parameterized method 3) method. In addition, the optimized geometry was obtained from literature and inserted into the program as fractional coordinates. The molecular orbitals and the resulting transitions are then calculated using Zerne overlap (ZINDO) method more specifically intermediate neglect of differential overlap for spectroscopy (INDO/S) model 103 The resulting ultraviolet and visible transitions are displayed in a graph of wavelength versus oscillator strength. These transitions are also displayed graphically as the molecular orbital surfaces involved in specific transitions. In investigation and support of the computed MO transitions, the molecular orbital energies were calculated with two separate methods, ZINDO and DFT 103 The first method used was the semiempirical parameterized method iteration 3 (PM3) for the optimization of the molecular structure. Then INDO/1 (intermediate neglect of differential overlap one) was applied for the calculation of the molecular orbital energies and electron occupations which is a model of the ZINDO method. ZINDO is based on quantum mechanic equations where some of the two electron integrals that involve the overlap of different basis functions, are ignored. Some o f the values assumed in the calculation are based on experimental values, therefore ZINDO is a semi empirical method. The second method employed was a density functional theory (DFT) method

PAGE 73

73 for geometry optimization and molecular orbital calculations. Me anwhile, the DFT is based on the electron density functional rather than the quantum mechanical equation. DFT tends to over delocalize the electrons and therefore over estimate the energies of the molecular orbitals. The ZINDO methods have the trend of o ver localizing the electrons and underestimating the molecular orbital energies. Using the two methods, their results will not be exactly the same but if they are both treating the system consistently, they should be relatively close. Several of the teste d energetic material molecules and related molecules were modeled using the computational chemistry. The computed molecular orbital transitions wavelengths were then compared to the experimental data. The graphical computer interface was also used to dem onstrate the construction of the molecular crystals of the energetic materials.

PAGE 74

74 CHAPTER 5 EXPERIMENTAL RESULTS 5.1 Differential Reflectograms of Energetic Materials The differential reflectogram of a typical solid TNT sample is shown in Figure 5 1 (on page 88 ). Several spectral features are present including a double peak near 250 nm a small peak at 300 nm, and most distinct, an absorption edge or shoulder between 390 and 400 nm. This shoulder will be termed from now on as the 420 nm feature, has an onset at about 440 nm, an inflection point at 420 nm and a maximum near 390 nm. Sin ce the 420 nm feature is situated at the long wavelength end of the UV spectrum, it will be also referred below as the long wavelength shoulder or red transition. It is unique to TNT in comparison to other inert materials, organic and inorganic, (includin g vegetation) that have been measured (Figure 5 2 and Figure 5 3). Several other common explosive materials, RDX, HMX, PETN, Tetryl, and Nitroglycerin have also been investigated. Their differential reflectograms are shown in Figure 5 4. Each explosive has a unique absorption edge at a large wavelength (low energy) that is distinct from the other explosives and inert materials. Specifically, the inflection point of these features are located near 330 nm for RDX, 320 nm for HMX, 275 nm for PETN, 300 nm f or Tetryl, and 310 nm for nitroglycerin. The inflection point is the minimum of the first derivative of the curve in the region of interest and gives a specific wavelength for comparison to the other energetic materials. Due to the wavelength of the shou lders for these materials the features will also be termed the long wavelength shoulder. The energetic materials were obtained from chemical standard companies, see Section 4.1. For confirmation, additional samples were provided on loan by the Alachua Cou nty Sheriff Office Bomb Squad and the Federal Bureau of Investigation (FBI). These samples were commercial or military grade explosive materials that demonstrated the same features as before

PAGE 75

75 and seen in Figures 5 5 and 5 6 for the materials obtained from the chemical standards companies. Nitroglycerin was obtained in the form of a heart medication tablet and was therefore mixed with inert materials. It is concluded that the observed features are independent of the supplier of the particular energetic mat erial. The long wavelength features are independent of the solvent used for recrystallization (if it is needed). This is shown particularly for TNT where the sample was recrystallized separately from solutions of acetonitrile, acetone, and ethanol, see Fi gure 5 7. TNT has different solubilities with each of the solvents that yield different masses of material from the same volume of solution. Therefore the samples shown in Figure 5 7 have similar masses of TNT rather than recrystallization from equivalen t volumes of solution. It may be argued that the 420 nm feature is occurring due to the substrate or to a substrate sample interaction. To investigate this TNT was placed on several different substrates and the differential reflectivity was measured as seen in Figure 5 8. All of the fabrics were black in color except where indicated. It was found that the 420 nm feature is maintained regardless of the substrate. 5.2 Characterization of TNT Purity, Crystal Structure, and Morphology The purity of the tes office bomb squad were evaluated using HPLC/MS. From the mass spectrometry results (Figure 5 9) it was found that the Chem service Inc. TNT was 93.8% TNT and 5.3% 2, 4 dinitrotoluene (DN T). The Bomb squad TNT was found to have a higher purity than the chemical standard (99.2% TNT and 0.2% 2, 4 DNT). Dinitrotoluene can be an impurity in TNT due to incomplete nitration of the toluene molecule during formation.

PAGE 76

76 Differential scanning calori metry (DSC) was used to characterize and confirm the presence of TNT. The DSC graph shown in Figure 5 10, displays a transformation, or change in weight, near 82C that is according to the literature, the melting point of TNT 8 9 X ray diffraction (XRD) spectrometry is routinely utilized to determine the crystal structure of materials 11 displays monoclinic and orthorhombic phases are labeled, of a TNT sample 68,101 The graph shows that the monoclinic and orthorhombic phases are both present while the majority of the sample is monoclinic. The samples of the energetic materials were also characterized using optical micrographs and optical profilometry. T hese two techniques show a closer view of the dendritic and needle like structure of the TNT samples. The optical micrograph for TNT (Figure 5 12) displays sharp thin needles of the TNT crystal. These needles are on average 20 long. The smaller crystal deposits clearly show additional dendritic behavior that cannot be seen by the naked eye. The varying morphologies of the other energetic materials are shown in Figure 5 12. PETN is distinctly different than the other materials in that it recrystallizes in a fairly smooth crystal. This unique morphology is also observed with the naked eye in the fine powder. HMX and RDX display dendritic morphologies similar to TNT but on a smaller scale. Tetryl is characterized by small dist inct crystals within this sample. The optical profilometry of TNT (Figure 5 13) displays how the individual needles grow on top of each other in stacks. 5.3 Behavior of Energetic and Other Materials with a Ring Structure Figure 5 14 shows the spectra of s everal energetic and non energetic materials that have ring or cyclic molecular structures, see Figure 5 15. The spectra infer that the presence of a ring structure is not the only prerequisite to having the absorption features in the long wavelength

PAGE 77

77 rang e. In fact, the only cyclic materials with a feature below 300 nm, in addition to a small wavelength feature, are TNT, RDX, HMX, and Tetryl. Therefore additional molecular compositions must be contributors to the observed features. 5.4 Mixtures of Severa l Energetic Materials Many explosive devices, commercial and military, can contain mixtures of several common energetic materials in order to tailor the explosive power needed in an application 5 Figure 5 16 shows differential reflectograms of dry binary mixtures of energetic materials. It is observed that each energetic material retains their individual characteristic features. Specifically, the shoulder of TNT and that of C 4 are individually identifiable in the same differential reflectogram. The features are even amplified compare d to the individual substances due to the differential character of the DRS. 5.5 Effect of Rotating or Tilting the Sample Due to the novel use of the DRS for powder samples, the question was raised if the rotation or tilt of the sample would affect the me asured spectrum. Thus, a TNT sample placed in the DR was rotated around its center point and measured at each degree of rotation seen in Figure 5 17. The shape of the spectrum varied slightly with the rotation angle but the presence and wavelength of the characteristic 420 nm feature was not affected. Variation of the spectrum shape is due to sampling several spots on the material where different crystal faces and concentrations of TNT are present at each spot. In addition, a sample was tilted with resp ect to the incident angle. This was achieved by placing a glass slide underneath one side of the sample glass slide at varying distances from a fixed point causing the glass slide and subsequently the sample to tilt (depicted in Figure 5 18). As seen in Figure 5 19, there is virtually no change in the DR spectrum with tilt angle. The wavelength of the characteristic feature remained virtually the same and only the intensity somewhat decreased. It is suggested that, as the angle was

PAGE 78

78 increased a smaller a mount of photons are collected by the detector causing the decrease in intensity. Since there was virtually no change in the measured differential reflectograms with rotation or tilt angle, then the reflected photons detected and shown in the DR are due to the tested material. This further assures that the characteristic features originate from their optical properties. 5.6 Dependence on the Number of Nitro Groups Present The DR of solid samples of 2, 6 DNT, 2, 4, DNT and 1, 3 DNB were measured and compa red to TNT. Mononitrotolulene was not measured because it is a liquid at room temperature making DRS measurements, utilizing a vertical sample stage, quite difficult. The measured spectra for the materials containing two nitro groups are displayed in Fig ure 5 20. An absorption edge similar to that of TNT is observed for each sample around 420 nm. With subtraction of the number of NO 2 groups, the onset of the absorption is slightly blue shifted (to a smaller wavelength) from that of TNT and subsequently the maximum and point of inflection are also blue shifted. Specifically, the inflection point is at 397 nm for 2, 6 DNT, 402 nm for 2, 4 DNT, and 403 nm for 1, 3 DNB. A comparison of the inflection points for TNT and the dinitro samples is displayed in Figure 5 21. From the measurements, it is concluded that the presence of the 420 nm feature does not appear to have a dependence on the number of nitro groups in the molecule, however there may be a relationship with the wavelength of this feature. 5.7 C omparison of Differential Reflectograms with Results from Other Optical Techniques The reflection mode of a UV visible spectrophotometer was used to compliment the DRS findings. The resulting spectrum, shown in Figure 5 22, displays the characteristic 420 nm feature of the TNT sample. Features in the bluer region (200 300 nm) are not observed with the

PAGE 79

79 UV Visible instrument in reflection mode. This may be due to the fairly low sensitivity of the spectrometer in this range or the absorption of light by the carbon substrate. The effect is eliminated in the DR measurements where many of the absorption features of the substrate are removed from the spectrum due to the differential technique. The transmission mode of the UV visible spectrophotometer was used to measure solutions of energetic materials. The solutions were chosen to be at a relatively high concentration (on the order of g/L) when placed in the spectrometer. Figure 5 23 shows UV visible absorbance spectra of TNT in an acetonitrile solution at a concentration of 3.2 g/L using four different spectrophotometers. The feature near 420 nm known from DRS measurements is slightly blue shifted. Specifically, the minimum of the first derivative is at 395 nm as compared to 420 nm. The Varian Cary 5e wa s used for the remainder of the UV Visible absorbance experiments. Transmission measurements are conventionally performed at smaller concentrations than used above. Therefore the concentration of the TNT solution was diluted in steps. Figure 5 24 displ ays resulting UV visible absorbance spectra of the continual dilution of the solution. It is observed that with decreasing TNT concentration, the 420 nm feature decreases in intensity until it is no longer observed. This phenomenon will be further invest igated in Section 5.8.1. It is noted that particle size analysis of the highest concentration acetonitrile solution did not measure any crystals of a size above the detection limit of the instrument (3 nm). Therefore, the TNT was quite dissolved in the s olvent. The other energetic materials measured previously using DRS were also investigated using the UV Visible transmission spectrophotometer. Solutions of high concentrations (1 g/L) were investigated (see Figure 5 25). They demonstrate long waveleng th absorbance near the features

PAGE 80

80 observed in the respective differential reflectograms. The shoulders or peaks in solution are blue shifted compared to the DRS spectra similar to TNT. In addition, the concentrations of the sample solutions were diluted to see if the long wavelength features (shoulders) were affected. The spectra taken for the continual solution dilution of several energetic materials are shown in Figure 5 26. All of the materials except PETN demonstrate a large dependence on the intensit y of the long wavelength feature with concentration. It is noted that the highest concentration PETN solution is 50 ppm which is an order of magnitude below that of the other energetic materials. This may affect the ability to see a measurable change in the long wavelength feature. Solutions of 2, 4 DNT, 2, 6 DNT, and 1, 3 DNB, in a solvent of equal amounts of methanol and acetonitrile, were also measured with the UV visible transmission spectrophotometer as shown in Figure 5 27. A feature around 420 n m is found blue shifted compared to the TNT solution at the same concentration. The inflection point of these features are blue shifted by an average of 25 nm compared to the DRS spectra. In addition, spectra of solutions of 2, 3, and 4 MNT, and nitroben zene (NB) were taken in a methanol and acetonitrile solvent, see Figure 5 28. A feature near 400 nm was again observed in these solutions where the 2 MNT shoulder was red shifted, at 380 nm, than the 3 and 4 MNT, observed at 363 and 370 nm, respectively a nd NB, at 367 nm. These compounds are a liquid at room temperature with a very high vapor pressure and due to this are not suitable to be measured with the DRS. A graph comparing the inflection points of the 420 nm feature for TNT, DNT, DNB, MNT, and NB is shown in Figure 5 29. It is noted that the inflection point for TNT is clearly at longer wavelengths compared to the other nitrotoluene or nitrobenzene samples. In addition, the position of the nitro groups on the benzene ring in reference to the met hylene group also has an

PAGE 81

81 effect. For example, 2, 6 DNT and 2 MNT have transitions closest to TNT. These molecules have nitro groups located in the ortho position adjacent to the methylene group. In summary the long wavelength shoulders found in the dif ferential reflectograms of TNT, RDX, HMX, PETN, and Tetryl are replicated in spectra taken with UV Visible transmission spectrophotometry of high concentration solutions. This needs to be emphasized due to possible criticism that the observed features may be artifacts of DRS. In addition, these results shine light on the reason why essentially no one else has found these UV and blue range features. This is because very dilute solutions are routinely used in UV Visible transmission spectrophotometry where the long wavelength shoulders do not exist. As for use as a explosives detection method, the UV Visible transmission spectrophotometry is not applicable to standoff detection and has a low sensitivity; in short DRS is clearly superior to UV Visible spect rophotometry. 5.8 Experiments Leading to Possible Answers for the Observed Transitions The presence of the 420 nm feature in TNT has been established and shown to be from the TNT and not an impurity or instrumental artifact. What remains to be explained i s the physical origin of this feature and the interaction with its environment. 5.8.1 Experimental Modifications of UV Visible Transmission Measurements As discussed above in Section 5.7 the large wavelength features observed in the DR have also been seen in high concentration solutions of the energetic materials but blue shifted from the differential reflectograms. The concentrations of the energetic material solutions were varied to see the effect on the 420 nm transition. UV visible transmission spectr a of TNT in solution at decreasing concentrations are shown in Figure 5 24 and 5 30. Figure 5 24 displays the raw absorbance data of the solutions where the intensity of the 420 nm feature decreases gradually with concentration to where it is no longer de tected at 26 ppm. The molar absorbtivity has been

PAGE 82

82 calculated using the Beer Lambert law, Equation 2 3. The result is graphed in Figure 5 30. This spectrum also demonstrates how the height of the 420 nm feature decreases with concentration and shows clea rly that the transition is no longer seen at smaller concentrations. Similar behavior is observed in the other energetic materials and DNT, DNB, MNT, MNB, but not PETN, see Figures 5 26. Fewer concentration variations were performed with these materials but the trend is still observed. Data analysis of the TNT solution concentration variation was performed by graphing the absorption intensity at a specific wavelength defined by the inflection point for the highest concentration solution, versus the solu tion concentration. Wavelengths near other spectral features, namely 210, 300, and 385 nm were also graphed as intensity versus solution concentration for comparison in Figure 5 31. The 385 nm graph has two regions of linearity, one for the high concentr ations to 850 ppm and one for the lower concentrations. A similar dual linearity is seen for the 300 nm graph but the slope change is at a much lower concentration (213 ppm) in the 234 nm graph. It is important to emphasize that the absorbance for higher concentrations in the 210 nm graph is above the sensitivity limit of the spectrometer. Therefore no conclusions on the linear behavior can be made. According to the Beer Lambert law the intensity of an absorbance feature must vary linearly with the conc entration of the measured solution. A deviation from the linearity indicates that the absorbance feature is not due to a single molecule but rather to a molecular complex with the solvent or other molecules. In order to determine if a complex was formed w ith the solvent, TNT was put in solution separately with several different solvents. The UV visible transmission spectra for these solutions with solvents of acetone, methanol, ethanol, isopropanol, water, and toluene are seen in Figure 5 32. Figure 5 33 is a graph of the solvent polarity, given in Table 5 1, versus the

PAGE 83

83 wavelength of the inflection point. There is no significant shift of the wavelength with the polarity of the solvent and all points are within 5 nm of each other. The exceptions to this are water (not shown in Figure 5 33) and ethanol. TNT has a low solubility in ethanol compared to the other solvents and is sparingly soluble in water. In fact TNT crystals are observed floating on the surface of the water indicating very little is disso lved in the water. It is noted that any variation in the wavelength is not consistent with a particular change in polarity. Therefore it has been shown that the 420 nm feature in TNT is not due to a molecular complex of the molecule and the solvent. It is possible that the transition can be due to a molecular complex between two of the same molecules. 5.8.2 Exposure of TNT to UV Light Visual observation of aged TNT shows a brown color compared to the pale yellow tint of new prepared TNT. This will be i nvestigated in more detail in the following experiments. It has been shown in the literature that TNT may degrade with exposure to UV light 104 106 A TNT sample was exposed to UV light in the DR instrument with the monochromator set at 200 and also to 250 nm wavelength As demonstrated in Figure 5 34, there was virtually no change in the observed spectrum of the TNT sample with increasing exposure time to the UV light. The power of the light source at these wavelengths was however, too small to be measured by the powe r meter. In another experiment, TNT was exposed to sunlight behind a window for several hours and up to 60 days. There was a decrease in the intensity of the 420 nm feature after several hours as shown in Figure 5 35. A spectrum of the sunlight with and without involvement of window glass, displayed in Figure 5 36, indicates that radiation by a wavelength longer than 250 nm may cause the decrease in the 420 nm feature. In regards to this, a TNT sample was exposed to a defocused HeCd laser beam, which em its UV light at 325 nm, for varying amounts of time, see Figure 5 37. The feature intensity at 420 nm decreases in size (height) with

PAGE 84

84 increasing exposure time until at 4 minutes there is only a slight inflection and at 8 minutes there is no distinguishabl e shoulder. Figure 5 38 compares the ratio of the normalized differential reflectivity at 380 nm with that at 450 nm versus the HeCd laser exposure time. When the ratio reaches a value of one, 380 nm and 450 nm are equal so there is no longer a feature i n this range. A significant drop in the ratio is seen between 15 and 30 seconds displaying an exponential decrease with increasing time. The total decrease is about 21 % in 1000 seconds. There are several reactions that may be occurring in the sample. On e of the first reactions to consider, due to the crystal morphology of TNT, is optical traps. Optical traps occur when photons are absorbed by the electrons in materials causing them to be raised to higher energy levels and subsequently trapped in energy levels caused by defects in the crystal. These electrons can not absorb any additional photons causing an absorption transition to no longer be observed. These traps, like many defects present in crystals, can be annealed out and the electrons can return to their ground state. To test this hypothesis, the irradiated TNT samples were annealed at 40 to 75C (the melting point of TNT is 82C) in a low temperature oven for varying amounts of time. The differential reflectograms of the annealed TNT samples a re shown in Figure 5 39. There is no observable increase of the 420 nm feature regardless of the time or temperature of the annealing treatment. Figure 5 40 further demonstrates this by displaying the 380 to 450 nm normalized differential reflectivity ra tio versus time for one temperature. A control TNT sample that had not been irradiated was also annealed, Figure 5 41, with no observable change in the 420 nm feature in the differential reflectogram. It is concluded that the loss of the 420 nm feature d ue to exposure to UV light is essentially not due to optical traps. Another reaction to consider is photolysis of the TNT molecule. It has been reported that the photodissociation energy for the ring C N bond is 3.2 eV or 488 nm 104 106 This results in the

PAGE 85

85 severing of nitro groups from the ring structure. In order to examine this phenomenon, a TNT sample was irradiated with the 457 and the 488 nm lines of an Ar ion laser. The 488 nm line is the most powerful of the Ar laser with a laser power near that of the HeCd las er (21 mW) whereas the 457 nm line is a secondary emission wavelength with a much lower power (0.57 of TNT irradiated with either the 457 or 488 nm laser line in Figure 5 42 does not change with increasing r adiation time for either laser line. This is more recognizable in Figure 5 43 where the ratio of 380 and 450 nm normalized differential reflectivity is plotted versus the exposure time for the 488 nm laser line The total decrease is 6.95% from the first to second measurement which could be well within the error of the measurement. From this data, the lower energy laser lines should be sufficient to cause photodissociation of the C N bond, but there is no significant change in the observed DRS spectrum. Therefore it is proposed that the power of the 457 nm laser line is too low and the 488 nm line is on the borderline of the required energy therefore no photodissociation is occurring. The reduction of TNT to DNT, MNT, or toluene should be observable in F ourier transform infrared spectrometry of the irradiated sample. Changing of the functional group bonds and the breaking of others should shift or decrease the intensity of the respective absorption bands. Figure 5 44 displays the FTIR results of a contr ol TNT sample and one that has been irradiated by the 325 nm HeCd line for 16 minutes so that the 420 nm feature is no longer observed. The FTIR bands of importance are the NO 2 symmetric and asymmetric stretches 88 at 1530 cm 1 and 1350 cm 1 There is virtually no difference in the FTIR spectra of the two samples which suggests that there is no appreciable change in bonding in the samples and therefore no dissociation of the ring C N bond. Due to the wavelength of the irradiation source (325 nm), only the surface of the sample may be effected. The change in the surface layer is detected by

PAGE 86

86 DR because it i s a surface analytical technique that samples the first few atomic layers of a material. Meanwhile in FTIR, the wavelengths are much longer so that the collected signal is characteristic of the bulk material and dominates over the signal of the surface la yer. In order to confirm the products produced during the photolysis of the TNT samples by the 325 nm line of a HeCd laser, chemical analysis by HPLC/MS was performed. Samples of control TNT (not irradiated), irradiated TNT with no 420 nm feature, and TNT irradiated in the acetonitrile solution were characterized. The HPLC/MS results shown in Figure 5 45 demonstrate that for all samples no DNT or MNT was detected. The only molecules that were present were TNT. This confirms that the process causing the disappearance of the 420 nm feature is not due to photodissociation of the TNT molecule. However, the possible photodissociation would occur predominately at the surface of the sample whereas HPLC/MS analyzes the entire (bulk) of the substance. This is a lso visually observed. Only the outer layer of exposed or aged TNT has a brown tint whereas the interior remains pale yellow. It is important to mention that it was shown in Section 5.6 that solid 2, 4 and 2, 6 DNT have features in the 420 nm region altho ugh they are somewhat blue shifted from TNT. In other words, creating a DNT molecule out of TNT may only slightly decrease the intensity of the observed TNT feature. In summary the cause for the brownish tine of aged and/or UV irradiated TNT has not bee n completely clarified in this research. Other models and experiments need to be designed for this. 5.9 Line Shape Analysis of Differential Reflectograms A differential reflectogram does not immediately yield the absorption of a material. As discussed in Section 4.1.2, only a line shape analysis of the DR reveals the wavelengths of the electron transitions. Since spectral optical constants of TNT and other explosives are not readily available, a simple graphical line shape analysis was instead performed which is based on Figure

PAGE 87

87 4 2. This line shape analysis may not reveal the exact wavelength of a transition; instead, it may vary only up to 10 nm. Figures 5 46 through 5 53 display the result of such a line shape analysis 2 (absorption) of TNT, RDX, HMX, Tetryl, PETN, 2,4 DNT, 2,6 DNT, and DNB. These graphs can be used to make a more direct comparison with the model calculations displayed in the next chapter. The wavelengths of the electron transitions, determined by this line shape analys is, are shown for all of the investigated energetic materials in Table 5 2.

PAGE 88

88 Figure 5 1. Differential reflectogram of a typical solid sample of dehydrated TNT on carbon tape. The spectral features of interest are seen as a double peak near 250 nm, a small peak at 300 nm, and a shoulder near 420 nm.

PAGE 89

89 Figure 5 2. Differential reflectograms of common organic and inorganic materials on carbon tape compared to TNT (A and B). There are no materials demonstrating the TNT spectral feature near 400 nm. Many do display a feature near 250 nm. The individual curves have been staggered for clarity.

PAGE 90

90 Figure 5 3. Differential reflectograms of vegetation from several plants compared to that of TNT. None of the vegetation cur ves display all of the fe atures of TNT. Individual curves have been staggered for clarity.

PAGE 91

91 Figure 5 4. Differential reflectograms of several common energetic materials on carbon tape. Each material has a unique shoulder similar to TNT but at a characteristically diff erent wavelength. The individual curves have been staggered for clarity.

PAGE 92

92 Figure 5 5. Differential reflectograms of TNT received from several sources including pure standards and military compositions on carbon tape. All spectra display essentially the same features. Figure 5 6. Differential reflectograms of several commercial explosive materials on carbon tape. All materials display their characteristic absorption features. The individual curves have been staggered for clarity.

PAGE 93

93 Figure 5 7. Differential reflectograms of TNT on carbon tape, recrystallized independently from solution with different solvents. The 420 nm feature is displayed in all the recrystallized materials.

PAGE 94

94 Figure 5 8. Differential reflectograms of TNT on sever al different substrates. All are black in color except where indicated. Individual spectra have been staggered for clarity. Note that the 420 nm feature appears on all surfaces.

PAGE 95

95 Figure 5 9. Chemical analysis results of high pressure liquid chromat ography mass spectra for TNT from a standards company (A) and a military explosive (B). The samples are mostly TNT with small 2, 4 DNT impurities.

PAGE 96

96 Figure 5 10. Differential scanning calorimetry spectrum of TNT, recrystallized from an acetonitrile soluti on. Figure 5 11. Spectrum from x ray spectrometry of solid TNT recrystallized from an acetonitrile solution. Significant lattice planes have been labeled indicating that both the monoclinic and orthorhombic phases are present.

PAGE 97

97 Figure 5 12. Optical mi crographs of the measured energetic materials at 20X magnification. A. TNT. B. RDX. C. HMX. D. PETN. E. Tetryl.

PAGE 98

98 Figure 5 13. Optical profilometry using false color to indicate height from the base plane (pale yellow in color) of a solid TNT sam ple recrystallized from an acetonitrile solution. Figure 5 14. Differential reflectograms of several materials with cyclic or ring molecular structures on carbon tape. Only the energetic materials display long wavelength shoulders. The individual curve s have been staggered for clarity.

PAGE 99

99 Figure 5 15. Molecular structures of the materials with cyclic molecular structures shown in Figure 5 12. Figure 5 16. Differential reflectograms of PETN mixed with TNT (A) and C 4 mixed with TNT (B) on carbon tape. The materials retain their individual absorption features in the mixtures.

PAGE 100

100 Figure 5 17. Differential reflectograms of a solid TNT sample, on carbon tape, at different rotation angles with respect to the incident light beam. There is no change in the sig nificant features of the TNT spectrum. Figure 5 18. Schematic representation of the tilting of a sample resulting reflection in the direction of the dotted arrows.

PAGE 101

101 Figure 5 19. Differential reflectograms of TNT, on carbon tape, at different tilt angles with respect to the incident light. The spectra intensity is steadily decreased but the features do not shift in wavelength. Figure 5 20. Differential reflectograms of solid samples of 2, 4 DNT, 2, 6 DNT, and 1, 3 DNB on carbon tape. A feature near 42 0 nm is seen similar to TNT. The individual curves have been staggered for clarity.

PAGE 102

102 Figure 5 21. The wavelength of the inflection point of the 400 nm feature for TNT, 2, 4 DNT, 2, 6 DNT, and 1, 3 DNB. There is some difference in the wavelength of these samples. Figure 5 22. Reflection spectrum of solid TNT on the reflection stage of a Perkin Elmer UV Visible spectrophotometer. The 420 nm shoulder is displayed in the spectrum. A control measurement (no sample on stage) is added for comparison.

PAGE 103

103 Figure 5 23. Molar absorbtivity of TNT measured in transmission utilizing an acetonitrile solution with at a concentration of 3.1 g/L taken with four different spectrophotometers. Essentially the same spectrum is found for each spectrometer, especially the 420 nm feature. The diamonds indicate the sensitivity limits of the instruments.

PAGE 104

104 Figure 5 24. Absorbance spectra of varying concentrations of TNT in acetonitrile solution measured by a Varian UV visible spectrophotometer. As the concentration is decreased the 420 nm shoulder appears to blue shift and decrease in intensity until it is no longer observed. The shaded region indicates a spectral artifact caused by competing absorbancies of the solvent and sample. This shaded area in this and the following figures of UV Visible transmission spectrometry should therefore be ignored or at least treated with some caution.

PAGE 105

105 Figure 5 25. Molar absorbtivity spectra of various energetic materials in solution of equal parts of ethanol and acet onitrile at concentrations of 1 g/L. Spectral features are observed that are similar to those found in the differential reflectograms.

PAGE 106

106 Figure 5 26. Molar absorbtivity spectra with varying concentration of the sample in the solution for (A) RD X, (B) HMX, (C) PETN, and (D) Tetryl. All materials except PETN display a decrease in wavelength of the long wavelength feature with decreasing concentration.

PAGE 107

107 Figure 5 27. Absorbance spectra of separate solutions of 2, 4 DNT, 2, 6 DNT, and 1, 3 DNB in high 1 g/L (A) and low 0.1 g/L (B) concentrations with equal parts ethanol and acetonitrile. Absorbance features similar to those in differential reflectograms are seen in the high concentration solutions. Figure 5 28. Absorbance spectra of separate s olutions of 2 MNT, 3 MNT, 4 MNT, and nitrobenzene with equal parts of ethanol and acetonitrile at a concentration of 1 g/L. A feature is demonstrated near 420 nm for each sample although the wavelength is shifted for each sample. Note the shaded part of the spectrum is an artifact due to the competing absorbencies of the solvent and sample.

PAGE 108

108 Figure 5 29. Wavelength of the inflection point for the 400 nm feature taken with an UV Visible transmission spectophotometer of solutions of TNT, 2, 4 DNT, 2, 6 DNT, 1, 3 DNB, 2 MNT, 3 MNT, 4 MNT, and nitrobenzene with their respective solvents at a concentration of 1 g/L. There is an observed dependence of the wavelength on the sample. The different color regions indicated the number of nitro groups present in the molecule.(compared to Figures 5 25, 5 27, and 5 28)

PAGE 109

109 Figure 5 30. Molar absorbtivity for varying concentrations of TNT in acetonitrile. The previously assumed blue shift of the 420 nm feature with decrease in concentration is now observ ed as an intensity decrease.

PAGE 110

110 Figure 5 31. Absorbance intensity versus the concentration of TNT in solution with acetonitrile at several wavelengths namely 384 nm (A), 300 nm (B), 210 nm (C), and 234nm (D). Parts E through H display the same re lationship for concentrations between 3.4 g/L and 1.7 g/L. The dashed line in (C) and (G) indicated the sensitivity limit of the instrument.

PAGE 111

111 Figure 5 31 continued.

PAGE 112

112 Figure 5 32. Molar absorbtivity of TNT in solution with several different solven ts at the same concentration. There is little influence of the solvent on the 420 nm feature except for water. Figure 5 33. Wavelength of the inflection point of the 420 nm feature for TNT in solution with several solvents versus the solvent polarity. There is a noticeable blue shift for the lowest polarity solvent compared to the other solvents.

PAGE 113

113 Figure 5 34. Differential reflectograms of solid TNT with increasing time of exposure to the monochromated 200 nm (A) and 250 nm (B) light of a high pressur e Xe lamp. There was virtually no change in the observed spectra with increased exposure. Figure 5 35. Differential reflectometry of a solid TNT sample exposed to the sun at increasing time intervals (A) short term and (B) long term. There is a noticea ble decrease in intensity of the 420 nm feature after a very short time.

PAGE 114

114 Figure 5 36. Spectrum of sunlight behind and in front of a lab window. There is a significant amount of light measured at wavelengths longer than 300 nm.

PAGE 115

115 Figure 5 37. Differential reflectograms of solid TNT with increasing time of exposure to the 325 nm line of a defocused HeCd laser. After several minutes the 420 nm feature is significantly decreased in intensity.

PAGE 116

116 Figure 5 38. Ratio of the intensity of the normalized differential reflectometry at 380 nm to 450 nm of TNT versus the time of exposure to the 325 nm HeCd laser line. The ratio indicates the size of the 420 nm feature.

PAGE 117

117 Figure 5 39. Differential reflectograms of solid TNT that h as been irradiated for 16 minutes with the 325 nm HeCd laser line followed by annealing for various times at 45C. There is no restoration of the 420 nm feature due to increased annealing times. Figure 5 40. Ratio of the normalized differential reflecti on intensity at 380 nm and 450 nm of TNT that has been irradiated with the 325 nm line of the HeCd laser for 16 minutes The samples have been subsequently annealed for various times at 45C. Asterisks indicate samples annealed at 75C.

PAGE 118

118 Figure 5 41. Diff erential reflectograms of solid TNT that has not been exposed to a UV laser versus annealing time at 45C. There is virtually no change in the observed spectra with increased annealing time. Figure 5 42. Differential reflectograms of solid TNT with incr easing time of exposure to the 457 nm (A) and 488 nm (B) lines of an Ar ion laser. There is no distinguishable decrease in the 420 nm feature with increasing exposure time.

PAGE 119

119 Figure 5 43. Ratio of the normalized differential reflection at 380 and 450 nm of solid TNT versus increasing time of exposure to the 488 nm line of an Ar ion laser. Compare to Figure 5 38 where exposure to a HeCd laser (325 nm ) is shown.

PAGE 120

120 Figure 5-44. Fourier transform infrared spectrometry of solid TNT on stainless steel with and without 16 minutes of exposure to th e 325 nm line of a HeCd laser. There is not a noticeable shift in the absorption bands or a large decrease in intensity that would indicate a bonding change. Figure 5-45. Chemical analys is results by high pressure liquid chromatography and mass spectrometry. The expected peaks (fro m constants) for DNT (A), MNT (B), and the measured spectrum fo r UV irradiated TNT (C). 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 Time (min) 0 50 100 Relative Intensity 0 50 100 0 50 100 0 Expected m/z 181 [M-H]of DNT NOT DETECTED 36.86 181.2 Expected m/z 136 [M-H]of NT NOT DETECTED 36.45 135.8 MW 227 TNT RT: 36.62 BP: 0.0 TNT Time (min) A B C Relative Intensity

PAGE 121

121 Figure 5-46. Absorption transiti ons of TNT based on a graphical line shape analysis of the respective differential reflectogra m, that is, of Figure 5-1. Figure 5-47. Graphical line shap e analysis of the respective differential reflectogram of RDX (Figure 5-4) revealing th e absorption transitions.

PAGE 122

122 Figure 5-48. Absorption transitions of PETN based on graphical line shape analysis of the respective differential reflectogra m, that is, of Figure 5-4. Figure 5-49. Graphical line shap e analysis of the respective differential reflectogram of HMX (Figure 5-4) giving the absorption transitions.

PAGE 123

123 Figure 5-50. Absorption transiti ons of Tetryl based on graphical line shape analysis of the respective differential reflectogra m, that is, of Figure 5-4. Figure 5-51. The absorption transitions of 2, 6 DNT found using graphical line shape analysis of the respective differential reflec togram that is Figure 5-4.

PAGE 124

124 Figure 5 52. Graphical line shape analysis of the respective differential reflectogram of 2, 4 DNT (Figure 5 4) revealing the absorption transitions. Figure 5 53. Absorption transitions of 1, 3 DNB based on graphical line shape analysis of the respective differential reflectogram, that is, of Figure 5 4.

PAGE 125

125 Table 5 1. Polarities of the solvent used in UV Visible transmission spectrophotometry. Solvent Polarity Acetone 21 Ethanol 24 Acetonitrile 37 Isoproponal 18 Methanol 33 Toluene 2.4 Table 5 2. Electron transitions of various energetic materials determined by the graphical line shape analy sis of the differential reflectograms. Energetic Material First transition (nm) Second transition (nm) Third transition (nm) Fourth transition (nm) TNT 417 300 263 243 RDX 314 256 244 PETN 308 285 236 HMX 276 241 Tetryl 276 241 2, 6 DNT 397 286 270 254 2, 4 DNT 402 293 275 258 1, 3 DNB 403 296 259

PAGE 126

126 CHAPTER 6 COMPUTATIONAL CHEMISTRY RESULTS The experimental results discussed in Chapter 5 give a reasonable idea of where the characteristic differential reflectogram traits may originate. Moreover, computational chemistry is a tool that can assist in di scovering almost exactly where the features are originating and under what circumstances they can occur 103 As outlined in Section 4.3 the ZINDO method was utilized for this task except where noted. 6.1 Single Isolated Molecule of TNT The computational analysis began w ith modeling a simple system of a single isolated TNT molecule drawn into the user interface and optimized using PM3 geometry. The calculated molecular orbital transitions are shown in Figure 6 1 as the wavelength versus the oscillator strength. Oscillat or strength is an indication of the probability of the transition and subsequently the intensity of the absorption as explained in Section 3.2. Transitions that are forbidden in that the oscillation strength is zero, are indicated in the graphs with an as terisk at the wavelength of the transition. As can be seen in Figure 6 1, most of the molecular orbital (MO) transitions are at wavelengths smaller than 310 nm. Further, a forbidden transition is located at 460 nm. For comparison, one monoclinic molecul e A was drawn using the fractional coordinates. The geometry was optimized using the PM3 method and the resulting molecular orbital transitions were calculated. The pertinent MO transitions versus the oscillator strength are shown in Figure 6 2 and are i dentical to those in Figure 6 1. A modeled absorbance spectrum based on this calculation is shown in Figure 6 3 with six features highlighted. The molecular orbitals that correspond to these transitions are displayed in Figure 6 4. The forbidden transit ion does not appear in Figure 6 3 because such transitions are not included in modeled spectra. A small intensity transition is observed near 300 nm where Figure 6 4 displays the molecular orbitals

PAGE 127

127 mostly involved are the nitro groups. One orthorhombic m olecule A is also drawn and optimized, with the same results as the other two molecules, see Figure 6 5. In order to compliment and confirm the calculated results by ZINDO, a time dependent density functional theory (TDDFT) calculation of a single TNT mole cule was also performed. The molecule was geometry optimized and the MO transitions were calculated (Figure 6 6). Figure 6 7 displays the MO energies versus occupancy that agree quite well with the ZINDO calculations to further prove this point. 6.2 One Isolated TNT Molecule from Monoclinic and Orthorhombic Crystals The molecular crystal structure of TNT was discussed in Section 2.2.1 as being monoclinic or orthorhombic at room temperature. A single TNT molecule from each of the molecular crystal structu res were entered into the program using the calculated fraction coordinates. The molecular geometry from the fractional coordinates is termed the current geometry. Molecules of both conformations A and B for each crystal structure (see Table 2.4) were mo deled separately. The calculated MO transitions for conformation A are shown in Figure 6 8 for monoclinic and Figure 6 9 for the orthorhombic structure. There are transitions near 420 nm in both conformation A molecules with small oscillator strengths. A modeled absorbance spectrum for the MO transitions for the monoclinic conformation A molecule is given in Figure 6 10 with five transitions highlighted; corresponding molecular orbitals participating in these transitions are depicted in Figure 6 11. It can be seen that all the long wavelength features originate from transitions within the NO 2 groups. The calculated MO transitions for the B conformation of the molecules from both crystals are displayed in Figures 6 12 and 6 13. The longest wavelength tr ansition is at 360 nm for the monoclinic crystal and 370 nm for the orthorhombic. This result demonstrates the effect of the difference in torsion angles of the NO 2 groups in the A and B conformations.

PAGE 128

128 6.3 Two Isolated TNT Molecules from the Monoclinic an d Orthorhombic Crystals A pair of A and B molecules were put into the user interface using fractional coordinates so that the pair of molecules appears just as they would in an ideal crystal. Figure 6 14 shows the molecular orbital transitions in the mono clinic crystal without geometry optimization where the longest wavelength transition is found to be at 320 nm. The majority of the MO transitions occur between 275 nm and 300 nm and near 200nm which is similar to a combination of the individual molecules A and B MO transitions. The exception is that no transitions appear above 320 nm. When the geometry of the molecular pair is optimized, the A molecule becomes nearly perpendicular to the benzene plane. The resulting MO transitions are similar to the non optimized ones where many transitions occur near 200 nm and 275 nm, see Figure 6 15. However, there is a weak oscillator strength transition near 350 nm. Molecular pairs of A and B molecules for the orthorhombic structure were also modeled. The non opti mized molecular pair has MO transitions shown in Figure 6 16 that are quite similar to the non optimized monoclinic pair. Several MO transitions appear in the 200 nm region and between 275 nm and 300 nm with the longest wavelength transition at 300 nm. O ptimization of the molecular pair causes a weak MO transition at 360 nm and a reduction in the MO transitions near 275 nm, see Figure 6 17. This behavior is comparable to the monoclinic molecular pair. The treatment of the molecular pair by ZINDO raises q uestions to how it considers the pair of molecules. This is observed where the 450 nm MO transition of the A molecule does not appear in the pair but several of the other transitions are present. Therefore time dependant density functional theory (TDDFT) and density functional theory (DFT) calculations were performed to determine the energy and occupation of the molecular orbitals. The calculation using TDDFT was performed on an in plane pair of monoclinic molecules similar to Figure 6 15. Results of thi s calculation (Figure 6 18) are very similar to ZINDO where the longest

PAGE 129

129 wavelength transition is near 320 nm, although the transition is forbidden in contrast to the ZINDO results. In addition the energy of the molecular orbitals of two molecules was calc ulated using DFT and were compared in Figure 6 19 to the same calculations performed by ZINDO method INDO/1. There is virtually no difference between the occupied molecular orbital energies of the single molecule and pair in both the ZINDO and DFT calcula tions. The unoccupied molecular orbitals of the ZINDO and DFT calculations have higher energies for the single molecule than the pair, but only for the highest energy molecular orbitals which are most likely not involved in the lower energy or long wavele ngth transition of interest. 6.4 Several (more than two) Molecules Involving Monoclinic and Orthorhombic Crystals In consideration of the above calculations, increasing numbers of TNT molecules were added into the MO transition calculations in ZINDO. Th e number of monoclinic A and B molecules were increased and calculated separately and are summarized in Figures 6 20 and 6 21 as the longest wavelength MO transition versus the number of molecules. It is observed that as the number of molecules is increas ed the wavelength of the longest transition decreased (blue shift). The calculated results for the orthorhombic molecules are summarized in Figures 6 22 and 6 23. A similar trend as the monoclinic molecules is observed where the wavelength of the longest wavelength MO transition decreases with increasing number of molecules. For both the monoclinic and orthorhombic crystals, the longest wavelength MO transition reaches a point after which it remains constant. Therefore it is assumed that as more molecul es are added the wavelength of the transition will no longer decrease significantly nor will it increase to 420 nm. 6.5 Entire Molecular Crystals of TNT: Monoclinic and Orthorhombic The entire unit cell of both the monoclinic and orthorhombic molecular cry stals contains eight molecules. Section 6.4 discussed calculated MO transitions for several molecules up to

PAGE 130

130 seven molecules. The entire unit cell then involves a continuation of these calculations and the results are in turn very close to the seven molec ules results. The monoclinic crystal is considered first and the unit cell from several directions is shown in Figure 2 4. In the non optimized unit cell, the longest wavelength transition is at 270 nm, as depicted in Figure 6 24. The majority of the M O transitions occur between 250 and 275 nm with several weak oscillator strength transitions near 200 nm. When the geometry of the unit cell is optimized several molecules turn so that there are nearly perpendicular to the neighboring molecules and the (1 00) plane, as can be seen in Figure 6 25. Figure 6 26 shows the MO transitions of this optimized geometry unit cell. The longest wavelength transition is slightly longer than the non optimized at 280 nm and the remaining transitions are mostly between 25 0 and 275 nm. The orthorhombic unit cell is seen in Figure 2 4 from several different directions. It is closely related to the monoclinic unit cell and in turn there are several similarities. From this, one might assume that the calculated MO transitio ns would be analogous to the monoclinic unit cell. The non optimized unit cell has several large oscillator strength transitions. They are longer then 275 nm which is slightly different than the monoclinic case seen in Figure 6 27. In the optimized geom etry unit cell, the MO transitions, depicted in Figure 6 28, are essentially the same as the monoclinic unit cell. Several of the molecules in the cell rotate in the same manner as those in the monoclinic unit cell. This result is not unexpected. 6.6 C hanging the Intermolecular Distance between Pairs of TNT Molecules As described in Sections 6.4 and 6.5, increasing the number of molecules does not cause a transition near 420 nm. In fact, more than one molecule causes the MO transition at 450 nm to be n o longer present. Due to the large amounts of defects found in a typical TNT crystal, the transition of interest may be due to intermolecular distances different from the equilibrium

PAGE 131

131 position used in the previous calculations. Pairs of molecules in the s ame plane and stacked (like pancakes) as they appear in the molecular crystal were modified and the MO transitions were calculated. The in plane molecular pair of A and B conformations in the monoclinic molecular crystal were calculated as the intermolec ular distance was increased and decreased from the equilibrium. These calculations were performed with the molecules with and without geometry optimization and summarized in Figure 6 29. For both geometries, as the intermolecular distance is increased th e wavelength of the longest wavelength transition decreases at first substantially and then remains about the same towards larger distances. In contrast, decreasing the intermolecular distance causes an increase in the wavelength of the longest transition at a fairly steep rate. Similarly orthorhombic molecules results involving in plane A and B pairs are summarized in Figure 6 30. As the intermolecular distance increases, the trend of the wavelength is equivalent to the monoclinic result, although the l ongest wavelength transition increases less dramatically than the monoclinic pairs. This is especially noted in that the longest wavelength MO transition is near 500 nm for the optimized geometry and below 400 nm for the non optimized. In the monoclinic pairs, the longest wavelength MO transition extends to nearly 2500 nm in the non optimized pair and 1250 nm for the optimized geometry. The molecules in the monoclinic crystal are stacked in the direction of the z axis in the A then B molecular sequence. Along the b axis direction the (010) plane is repeated so that an A molecule lies directly below, at an equilibrium spacing, another A molecule and the same for the B molecules. Thus, pairs of molecules that appear to be stacked on top of each other lik e a pancake stack are pairs of A molecules and pairs of B molecules. This is also true for the orthorhombic unit cell even though the stacking sequence in the z axis direction is ABBA.

PAGE 132

132 Calculated results for pairs of monoclinic A molecules are presented in Figure 6 31. For the intermolecular distance increasing from the equilibrium, the same trend of a stable longest wavelength transition as in the in plane pairs is observed. An interesting feature is that the wavelength of the optimized geometry pair MO transition is about 100 nm longer than the non optimized pairs down to a intermolecular distance of 1.29 The wavelength of the low energy MO transitions goes to 1500 nm as the distance decreases for the optimized geometry and to near 900 nm for the non optimized. The calculated results of monoclinic pairs of B molecules are illustrated in Figure 6 31. A similar trend to the A molecules is seen where the optimized geometry transitions lay about 75 nm longer than the non optimized geometry pairs. A t about 1.7 the non optimized geometry pair surpasses the longest wavelength transition at 350 nm and quickly reaches extremely long wavelengths. These wavelengths are much longer than those reached by the pair of A molecules. Calculated results from the orthorhombic stacked A molecule pairs are summarized in Figure 6 32. It is observed that the results are slightly different than the equivalent monoclinic pair. This is best demonstrated by the smaller wavelength difference (about 60 nm) of the optim ized and non optimized molecular pairs. As the intermolecular distance decreases, the maximum wavelength achieved is distinctly lower than those found for the monoclinic pairs. For the B molecular stacks of the orthorhombic crystal, the separation of the current and optimized geometry longest wavelengths is the same as the A molecular pairs, 60 nm (see Figure 6 32). The longest wavelength transition for the optimized B molecular pairs at the smallest intermolecular distance is at 1827 nm which is larger than the same for the A molecular pair at 620 nm.

PAGE 133

133 In summary, a few distinct intermolecular distances display MO transitions at wavelengths near 420 nm especially for the stacks of A molecules in both crystal structures. These are all seen at intermolec ular spacings approximately one smaller than the equilibrium spacing. This result is depicted in Figure 6 33 for a pair of orthorhombic molecules in the A conformation at an intermolecular distance of 1.5 which is 1.2 smaller than the equilibrium d istance. A MO transition is found near 420 nm, specifically at 390 nm. From these observations, it is proposed that the 420 nm transition is more likely to occur when the molecules are in compressive stress than when they are at or further apart than the equilibrium distance. 6.7 Modifying the TNT Unit Cell to Yield Defects in the Crystal It has been reported that the TNT crystal readily exists as a monoclinic crystal with large amounts of stacking faults, in particular twinning 65 This is where an orthorhombic crystal is essentially formed inside of the monoclinic crystal causing stacking defaults and other defects. The modeling results of unit cells, as discussed in Section 6.5 were for perf ect crystals. Using the information from Section 6.6 it is reasonable that these calculations of perfect crystals did not yield MO transitions near 420 nm, due to their equilibrium intermolecular distances. It is possible that the 420 nm transitions only occur in non ideal situations. Consequently each unit cell was modified in order to simulate non ideal conditions. The first modification considered molecular defects, that is, one molecule within the unit cell was not an ideal TNT molecule. A TNT mol ecule was deconstructed within the unit cell by systematically removing substituents of the molecule and finally creating a vacancy. Figure 6 34 depicts the MO transitions of a monoclinic unit cell where one TNT has had a nitro group removed so that it is now a dinitrotoluene molecule. The system with optimized geometry is shown in Figure 6 34. Another nitro group was removed creating a mononitrotoluene molecule whose MO transitions of the current and optimized geometry are displayed in Figure 6 35. A

PAGE 134

134 t oluene molecule is created when the last nitro group is removed. The corresponding MO transitions are found in Figure 6 36 for the current and optimized geometry. Then the methylene (CH 3 ) molecule was removed which left only a benzene ring in the crystal MO transitions for the current and optimized geometry are given in Figure 6 37. Finally the entire molecule was removed leaving a vacancy in the TNT unit cell. These MO transitions are shown in Figure 6 38 for the current and optimized geometry. The l argest wavelength MO transitions are generally observed when the modified unit cell geometry is not optimized. For example, the largest wavelength MO transition occurs when the mononitrotoluene impurity is in the unit cell, at 333 nm. Several longer wave length forbidden transitions are noticed when toluene and benzene molecules are present. The toluene impurity has the longest forbidden transition at 364 nm. When the vacancy is present there are no MO transitions forbidden or allowed at wavelengths long er than 300 nm. The second modification of the TNT unit cell was to insert a stacking fault and calculate the MO transitions. This was performed in two instances, the first being an entire plane vacancy where two planar molecules were removed from the u nit cell. The resulting MO transitions are depicted in Figure 6 39 for the current and optimized geometry. Second, four orthorhombic molecules from the unit cell were inserted into the stacking sequence of the monoclinic unit cell. These molecules essen tially took the place of four monoclinic molecules but with the properties of the orthorhombic unit cell. Figure 6 40 shows the calculated MO transitions for the current and optimized geometry. The plane vacancy configuration resulted in MO transitions at and below 288 nm for the non optimized geometry. Apparently, a stacking fault yielded the most dramatic change in the MO transitions with a fairly strong transition at 409 nm (and several forbidden transitions at 460,

PAGE 135

135 480, and 482 nm). In addition, MO transitions are found near 250 nm and 300nm which are in good agreement with the experimental results for TNT in DRS (see inset of Figure 6 1). When the geometry of the system was optimized several of the TNT molecules rotated out of the plane moving the longest wavelength MO transition to 270 nm. This rotation is similar to what was observed in the geometry optimization of the TNT monoclinic and orthorhombic unit cells. It is therefore concluded that stacking faults inserted into the monoclinic molecula r crystal, are a distinct possibility for explaining the observed DR features of TNT. 6.8 Other Energetic Materials The experimental results presented in Chapter 5 also included data taken on other energetic materials related to TNT. Several of these expl osives were also briefly investigated with molecular modeling, namely RDX, HMX, Tetryl, and PETN. Additional modeling calculations for all the energetic materials investigated, similar to what was performed for TNT, is proposed as future work. 6.8.1 RDX T he RDX molecule, pairs, and unit cell was entered into the user interface using fractional coordinates and cell parameters from the literature 71 The MO transitions for the single molecule for the current and optimized geometry are shown in Figure 6 41. Calculations for the current geometry molecule show several long wavelength transiti ons at 398 and 412 nm and the next shortest transition is at 255 nm. The optimized geometry molecule has a long wavelength transition at 384 nm and a very weak or forbidden transition at 369 nm. These MO transitions are found at a much longer wavelength than the experimental results by nearly 60 nm which is unlike the TNT molecule. Next, two RDX molecules are placed in the user interface using the fractional coordinates so that they are as they would appear in the molecular crystal. The calculated resu lts for the

PAGE 136

136 current and optimized geometry are depicted in Figure 6 42. Similar to two TNT molecules, the longer wavelength transitions disappear when the two molecules are considered. In the current geometry pair a forbidden or very low oscillation stre ngth transition is seen at 381 nm and an allowed transition at 309 nm. When the geometry is optimized the longest wavelength transition is now seen at 282 nm. The current geometry RDX pair has MO transitions fairly close to the experimental results. Thi s behavior again is dissimilar to the TNT molecule pair. The RDX unit cell contains eight molecules similar to the TNT crystals unit cells. Figure 6 43 displays the calculated MO transitions for the current and optimized geometry. Analagous to the TNT u nit cells the longest wavelength MO transitions are far into the blue region. For the current geometry, the longest wavelength transition is at 245 nm and for the optimized geometry at 254 nm. There are many forbidden MO transitions near 200 nm. 6.8.2 HM X A single HMX molecule and subsequent combinations of molecules were entered into the modeling program using fraction coordinates obtained from the literature 107 The calculat ed MO transitions for current and optimized geometry are shown in Figure 6 44. For the current geometry the longest wavelength MO transition at 385 nm has a fairly small oscillator strength. When the geometry is optimized this transition becomes stronger and red shifts slightly. In addition more transitions appear in the 200 nm region. This behavior is unusual compared to the closely related RDX molecule. The modeling of two HMX in plane molecules is shown in Figure 6 45. There are noticeably no MO tra nsitions found for either geometry at wavelengths larger than 300 nm. In fact, the longest wavelength transition for both geometries is at 283 nm with nearly the same oscillator strength.

PAGE 137

137 The beta HMX molecular crystal stable at room temperature has onl y two molecules in the unit cell so modeling a pair of molecules is equivalent to modeling a unit cell. Therefore four molecules have been modeled to represent the effect of several molecules in a space. The calculated MO transitions are shown in Figure 6 46 for the current and optimized geometries where there are no MO transitions above 250 nm for either geometry. 6.8.3 Tetryl As discussed in Chapter 2 the Tetryl is the closest energetic material to TNT in molecular structure. However, its crystal struc ture is very similar to HMX. A single Tetryl molecule is modeled and the calculated MO transitions are shown in Figure 6 47 for the current and optimized geometry. The most notable characteristic of these molecules is the large number of MO transitions e specially for the current geometry. In the current geometry graph several transitions occur beyond 300 nm with the longest wavelength transition at 377 nm. The transition is red shifted after geometry optimization and now at 391 nm with the next transiti on at 292 nm. The calculated MO transitions for an in plane pair of Tetryl molecules are displayed in Figure 6 48. The large number of MO transitions is retained and the longest wavelength transition is blue shifted and now at 315 nm for the current geo metry. In the optimized geometry the longest wavelength transition is blue shifted past 300 nm to 290 nm. One unit cell of the Tetryl crystal contains eight molecules and was modeled. The calculated MO transitions are shown in Figure 6 49 for the curren t and optimized geometry. Several MO transitions occur in the current geometry with the longest wavelength transition at 315 nm. When the geometry is optimized this is blue shifted to 275 nm.

PAGE 138

138 6.8.4 PETN The PETN molecule is quite distinctive from the oth er tested energetic materials and is only similar in that it has several nitro groups. It is not cyclic and has almost a square like structure. This uniqueness is further affirmed by the calculated MO transitions depicted in Figure 6 50. The current geo metry calculations show many MO transitions above 200 nm with several closely grouped transitions near 235, 260, and 340 nm with the latter being the longest wavelength transition. When the geometry is optimized many of the transitions below 200 nm reduce greatly in oscillator strength or become forbidden. The longest wavelength transition is now at 215 nm. Next, two PETN molecules are placed in the modeling program and the MO transitions are calculated, as seen in Figure 6 51. The MO transitions reduc e to four closely packed groups, where the longest wavelength transition is within a group at 285 nm. While the geometry optimization reduces the MO transitions even more to two groups both below 200 nm. The PETN molecular crystal has two molecules in it s unit cell. Therefore, like HMX, four molecules have been modeled in order to observe the influence of several molecules in a space (Figure 6 52). The number of transitions increases greatly from the molecular pair, however, the oscillator strength decr eases significantly as well. The MO transitions may have been present in the pair but too small in strength to be seen in the graph. In contrast the geometry optimized calculation displays almost the exact same behavior as the molecular pair. 6.8.5 Trini trotoluene Derivatives: DNT and MNT The dinitrotoluenes and mononitrotoluenes have a considerably smaller explosive power than all the other energetic materials described above, but they are still considered energetic molecules. They are the most common i mpurity in TNT due to incomplete nitration, which also makes them of great interest.

PAGE 139

139 Dinitrotoluene is observed in the 2, 4 or 2, 6 confirmations. The first molecule investigated was the 2, 4 conformation of DNT. Figure 6 53 shows the calculated MO tra nsitions of the current and optimized geometry of the molecules. The current geometry 2, 4 DNT molecule was formed by removing the 6 nitro group of the TNT molecule A from the monoclinic structure. From Figure 6 53 it is noted that the longest wavelength transition occurs at 447 nm with two forbidden transitions at 419 and 410 nm. When the geometry is optimized the longest wavelength MO transition red shifts to 454 nm and a forbidden transition occurs at 460 nm. The calculated MO transitions for the cu rrent geometry 2, 6 DNT model are presented in Figure 6 53. Several of the MO transitions are very similar if not exactly like the 2, 4 DNT molecules with similar geometry. The longest wavelength MO transition is at 449 nm with another transition close b y at 445 nm. After geometry optimization the longest wavelength transition is now at 456 nm and the next transition has a very low oscillator strength at 455 nm. The mononitrotoluenes are observed in the 2, 3, and 4 conformations otherwise known as the ortho, meta, and para. In a similar fashion as the DNT molecules, the MNT molecules were formed by removing all but the desired nitro group from the TNT molecule A of the monoclinic crystal. The 2 MNT was modeled and the MO transitions that were calculat ed are seen in Figure 6 54. It is notable that there is roughly no difference between the MO transitions of the current and optimized geometry molecules. The only deviation from the current geometry is a red shift in the longest wavelength transition fro m 446 nm to 460 nm. There is also a blue shift of the very low oscillator strength transition at 417nm to 391 nm in the optimized geometry molecule. Calculations of the MO transitions for the 4 MNT are depicted in Figure 6 54. The MO transitions of the current geometry molecule are closely related to those of the same geometry 2

PAGE 140

140 MNT molecule. The longest wavelength transition occurs at 443 nm and the next transition is at 420 nm. After geometry optimization, the longest wavelength transition has a ver y low oscillator strength and is red shift to 461 nm. These results confirm what was presented in Sections 5.6 and 5.7 where the dinitrotoluenes and mononitrotoluenes have features near 400 nm. These molecules do display computational results unlike the single TNT molecules. Several long wavelength transitions exist for the molecules even after geometry optimization. 6.9 Non High Explosive Materials (Components of the TNT Molecule) Several parts of the TNT molecule, which are also in several other energ etic materials, were individually modeled. The resulting MO transitions for methylene (CH 3 ), toluene, benzene, the nitro group (NO 2 ), and methylene (CH 3 ) with current and optimized geometry are shown in Figures 6 55 and 6 56. It is clearly demonstrated t hat any transitions longer than 275 nm are due to the nitro groups. This is seen in the current geometry calculation, where the longest wavelength transition of interest is at 495 nm in the nitro group (Figure 6 56). Other MO transitions were found at wa velengths up to 3170 nm for this molecule. When the geometry is optimized the longest wavelength transition occurs at 670 nm and the longest wavelength transition in the range of interest is forbidden and at 389 nm.

PAGE 141

141 Figure 6 1. Calculated mole cular orbital transitions by ZINDO of one isolated TNT molecule. The TNT molecule was free drawn into the interface and the geometry was optimized using the PM3 method. The asterisk indicates a forbidden transition. Inset: Absorption for TNT determined from the pertinent differential reflectogram using line shape analysis, see Figure 5 46. The shaded region indicates the UV range that is not measured with DRS and therefore is not considered. This procedure is used in the following figures.

PAGE 142

142 F igure 6 2. Calculated molecular orbital transitions by ZINDO of one TNT molecule based on monoclinic molecular crystal geometry and optimized using the PM3 method.

PAGE 143

143 Figure 6 3. A modeled UV Visible spectrum based on the calculated molecular orbit al transitions by ZINDO for one TNT molecule based on the monoclinic molecular crystal geometry optimized with the PM3 method. Absorbance transitions of interest are indicated numerically and described by the participating molecular orbitals shown in Figu re 6 4.

PAGE 144

144 Figure 6 4. One isolated TNT molecule based on the monoclinic molecular crystal geometry and geometry optimized using the PM3 method with molecular orbitals that participate in absorption transitions highlighted. The solid orbitals are the highest occupied molecular orbitals (HOMO) and the shaded orbitals are the lowest unoccupied molecular orbitals (LUMO). The numerical labels indicate the corresponding transition in Figure 6 3. In this and other diagrams of the molecules the atoms a re color coded where gray is carbon, white hydrogen, blue nitrogen, and red oxygen. The color and shape of the orbitals indicate the type of molecular orbital. The molecular orbitals are colored to better differentiate them from each other and are not ch aracteristic of a particular type. The orbital shape is based on classical orbital shapes e.g. spheres for s orbitals.

PAGE 145

145 Figure 6 5. Molecular orbital transitions calculated by ZINDO for one isolated TNT molecule based on the orthorhombic molecula r crystal geometry and optimized using the PM3 method. Asterisks indicated forbidden transitions.

PAGE 146

146 Figure 6 6. Calculated molecular orbital transitions by time dependent density functional theory (TDDFT) for one isolated TNT molecule with th e geometry optimized.

PAGE 147

147 Figure 6 7. The molecular orbital energies versus orbital occupancy for one isolated TNT molecule calculated by ZINDO and density functional theory (DFT). There is an appreciable difference between the MO energies calculated by the two models. Figure 6 8. Molecular orbital transitions calculated by ZINDO for one isolated TNT molecule having the A conformation in the monoclinic molecular crystal.

PAGE 148

148 Figure 6 9. The molecular orbital transitions for one isolated TNT molecule in the A conformation of the orthorhombic molecular crystal.

PAGE 149

149 Figure 6 10. Modeled spectrum of one isolated TNT molecule with the A conformation of the monoclinic molecular crystal based on the molecular orbital transitions calculated by ZINDO. Arabic nume rals label the transitions of interest further described by Figure 6 11.

PAGE 150

150 Figure 6 11. Molecular orbitals that participate in the MO transitions calculated for one isolated TNT molecule with the A conformation of the monoclinic molecular cryst al. The highlighted molecular orbitals are those involved in the highlighted transitions in Figure 6 10. Solid color molecular orbitals are the HOMO and the shaded orbitals are the LUMO.

PAGE 151

151 Figure 6 12. Calculated molecular orbital transitions by ZINDO of one isolated TNT molecule having the B conformation of the monoclinic molecular crystal. Figure 6 13. Molecular orbital transitions of one isolated TNT molecule in the B conformation of the orthorhombic molecular crystal calculated by ZINDO.

PAGE 152

1 52 F igure 6 14. Calculated molecular orbital transitions of two TNT molecules in both A and B conformations of the monoclinic molecular crystal as they would appear in the crystal as a pair in the same plane.

PAGE 153

153 Figure 6 15. Molecular orbital trans itions of two TNT molecules based on the A and B conformations of the monoclinic molecular crystal and geometry optimized using the PM3 method calculated by ZINDO.

PAGE 154

154 Figure 6 16. Calculated molecular orbital transitions by ZINDO of two TNT molecules in t he same plan having the A and B conformations of the orthorhombic molecular crystal as they would appear in the crystal structure. Figure 6 17. Molecular orbital transitions calculated by ZINDO of two TNT molecules based on the A and B conformations of t he orthorhombic molecular crystal and geometry optimized using the PM3 method.

PAGE 155

155 Figure 6 18. The molecular orbital transitions of two TNT molecules based on the A and B conformations of the monoclinic crystal calculated by TDDFT. Figure 6 19. The energy of the molecular orbitals versus occupancy of one and two TNT molecule systems calculated separately by ZINDO and DFT. The difference in the calculated MO energies is seen only in the LUMO orbitals.

PAGE 156

156 Figure 6 20. Summary of the calculated molecula r orbital transitions by ZINDO for increasing numbers of monoclinic conformation A TNT molecules in a system. The longest (reddest) wavelength transition for each calculation is shown versus the number of molecules present. Geometry optimization using th e PM3 method was also performed and the resulting longest wavelength MO transitions are indicated in red.

PAGE 157

157 Figure 6 21. Wavelength of the longest wavelength (reddest) calculated molecular orbital transition by ZINDO for increasing numbers of TN T molecules in the B conformation of the monoclinic molecular crystal. Geometry optimization using the PM3 method was also performed for each calculation and the resulting longest wavelength MO transitions shown in red.

PAGE 158

158 Figure 6 22. Summary o f the molecular orbital calculations by ZINDO of increasing numbers of TNT molecules in the A conformation of the orthorhombic crystal. The longest wavelength (reddest) MO transition is displayed versus the increasing number of molecules. The calculation s were repeated after geometry optimization using the PM3 method and shown in red.

PAGE 159

159 Figure 6 23. The longest wavelength (reddest) MO transition calculated by ZINDO with increasing number of TNT molecules in conformation B of the orthorhombic cr ystal. Geometry optimization using the PM3 method was also performed and the calculations repeated and displayed in red.

PAGE 160

160 Figure 6 24. Calculated molecular orbital transitions of a unit cell (8 molecules) of the monoclinic molecular crystal of TNT by ZIN DO. The asterisk at 260 nm indicates the forbidden transition. Figure 6 25. Representations of the unit cell of the monoclinic molecular crystal of TNT after geometry optimization with the PM3 method.

PAGE 161

161 Figure 6 26. Molecular orbital transitions calcula ted by ZINDO of 8 molecules based on the unit cell of the monoclinic molecular crystal of TNT and geometry optimized using the PM3 method. Figure 6 27. Calculated molecular orbital transitions by ZINDO of a TNT unit cell (8 molecules) of the orthorhombic molecular crystal.

PAGE 162

162 Figure 6 28. Molecular orbital transitions of 8 TNT molecules based on the unit cell of the orthorhombic molecular crystal and geometry optimized using the PM3 method calculated by ZINDO.

PAGE 163

163 Figure 6 29. Summary of the calcu lated molecular orbital transitions by ZINDO for two TNT molecules in the A and B conformation of the monoclinic molecular crystal at varying intermolecular distances. The longest wavelength (reddest) MO transition for each pa ir is displayed for each inte rmolecular distance. The calculations performed on molecular pairs that have optimized geometry using the PM3 method are shown in red. The inset depicts an enlargement of the wavelength scale to make the transition wavelength around 400 nm more visible. The smallest inset is a representation of the molecular pair used in the calculation at the equilibrium intermolecular distance which is indicated by the arrow in the graphs.

PAGE 164

164 Figure 6 30. Calculated molecular orbital transitions by ZINDO for pairs of TN T molecules in the A and B conformations of the orthorhombic crystals as they would appear in the crystal at varying intermolecular distances. The longest wavelength (reddest) transition for each intermolecular distance is plotted. The calculation was re peated for the pairs after geometry optimization using the PM3 method and are shown in red. Figure 6 31. The longest wavelength (reddest) calculated molecular orbital transition by ZINDO versus the intermolecular distance of a pair of TNT molecules in th e A conformation (A) and B conformation (B) of the monoclinic crystal. The molecules are stacked as they would appear in the molecular crystal. The calculation was repeated after geometry optimization using the PM3 method, of the pairs at each intermolec ular distance and is shown in red.

PAGE 165

165 Figure 6 32. Summary of the calculated molecular orbital transitions by ZINDO for pairs of TNT molecules in the A conformation (A) and B conformation (B) of the orthorhombic molecular crystal at varying intermolecular d istances. The molecular pairs of two A molecules or two B molecules are stacked as they would appear in the crystal (see insets). The calculations repeated after geometry optimization using the PM3 method, are displayed in red. Figure 6 33. Molecular o rbital transitions for a pair of orthorhombic conformation A TNT molecules stacked like pancakes at an intermolecular distance of 1.5 calculated by ZINDO.

PAGE 166

166 Figure 6 34. Calculated molecular orbital transitions by ZINDO of the monoclinic unit cell of T NT with one TNT molecule replaced with a DNT molecule (A). The calculation was repeated for the modified unit cell after geometry optimization using the PM3 method (B). Figure 6 35. Molecular orbital transitions calculated by ZINDO of the monoclinic unit cell of TNT with a MNT molecule in the place of one TNT molecule (A). After geometry optimization using the PM3 method of the modified unit cell, the calculation was repeated (B).

PAGE 167

167 Figure 6 36. The calculated molecular orbital transitions by ZIN DO of a modified monoclinic molecular crystal unit cell of TNT where one TNT molecule is replaced with a toluene molecule (A). The PM3 method was used for geometry optimization of the modified unit cell and the calculation repeated (B). Forbidden transit ions are indicated by an asterisk. Figure 6 37. Molecular orbital transitions calculated by ZINDO of a unit cell of the monoclinic molecular crystal of TNT with a benzene molecule replacing one TNT molecule (A). The calculation was repeated after the modified unit cell geometry was optimized using the PM3 method (B).

PAGE 168

168 Figure 6 38. Calculated molecular orbital transitions by ZINDO of a monoclinic molecular crystal unit cell of TNT where one TNT molecule has been removed leaving a vacancy (A). After geometry optimization of the modified unit cell using the PM3 method, the calculation was repeated (B). Figure 6 39. The molecular orbital transitions for a unit cell of the monoclinic molecular crystal of TNT where two TNT molecules have been removed leaving a plane vacancy calculated by ZINDO (A). The calculation was repeated after geometry optimization of the modified unit cell using the PM3 method (B).

PAGE 169

169 Figure 6 40. Calculated molecular orbital transitions by ZINDO of the monoclinic mol ecular crystal unit cell of TNT where four TNT molecules in the stacking sequence have been removed and four TNT molecules with orthorhombic molecular crystal conformations have been inserted creating a stacking fault (A). The geometry of the modified uni t cell was optimized using the PM3 method and the molecular orbital transition calculations repeated (B). Asterisks designate the wavelength of forbidden transitions.

PAGE 170

170 Figure 6 41. Molecular orbital transition calculations by ZINDO of one RDX mol ecule from the molecular crystal stable at room temperature (A). Molecular orbital transition calculated by ZINDO after optimizing the geometry of the RDX molecule (B). Inset of (A) is the absorption transitions determined by line shape analysis of the r espective differential reflectogram (see Figure 5 47).

PAGE 171

171 Figure 6 42. The calculated molecular orbital transitions by ZINDO for two RDX molecules as they would appear in the molecular crystal stable at room temperature (A). After geometry optimization u sing the PM3 method, the molecular orbital transitions of the pair of molecules was repeated (B). Figure 6 43. Molecular orbital transitions of one unit cell (8 molecules) of the RDX crystal stable at room temperature calculated by ZINDO (A). The calc ulation was repeated after geometry optimization of the unit cell using the PM3 method (B). The forbidden transitions are denoted by an asterisk.

PAGE 172

172 Figure 6 44. The molecular orbital transitions of a HMX molecule from the molecular crystal that is stab le at room temperature calculated by ZINDO (A). The same calculation repeated after the geometry of the HMX molecule was optimized using the PM3 method (B). Inset of (A) is the absorption transitions by line shape analysis of the pertinent DR(see Figure 5 49).

PAGE 173

173 Figure 6 45. Calculated molecular orbital transitions by ZINDO for two HMX molecules from the molecular crystal (A). After the geometry of the pair of molecules was optimized using the PM3 method, the calculation was repeated (B). Figure 6 46. Molecular orbital transitions for four HMX molecules from the molecular crystal calculated by ZINDO (A). The same calculations were repeated after the geometry of the molecules was optimized.

PAGE 174

174 Figure 6 47. The calculated molecular orbital trans itions by ZINDO of one Tetryl molecule from the molecular crystal stable at room temperature (A). After geometry optimization of the molecule using the PM3 method, the calculation was repeated (B). Inset of (A) is the absorption transitions found by line shape analysis of the respective differential reflectogram(see Figure 5 50).

PAGE 175

175 Figure 6 48. Molecular orbital transitions calculated by ZINDO for two Tetryl molecules as they would appear in the molecular crystal (A). The calculation was repeated after the geometry of the pair was optimized (B). Figure 6 49. Calculated molecular orbitals transitions by ZINDO of an entire unit cell (eight molecules) of Tetryl (A). The MO transitions were recalculated after the geometry of the unit cell was optimized (B). Forbidden transitions are indicated by an asterisk.

PAGE 176

176 Figure 6 50. The molecular orbital transitions of one PETN molecule as calculated by ZINDO (A). After geometry optimization of the PETN molecule (B). Inset is the line shape analysis of the pertinent differential reflectogram (see Figure 5 48).

PAGE 177

177 Figure 6 51. The results of calculating the molecular orbital transitions by ZINDO for a pair of PETN molecules as they appear in the molecular crystal (A). The MO transitions calculated after ge ometry optimization of the pair of molecules (B). Figure 6 52. Molecular orbital transitions of four PETN molecules as they would appear in the molecular crystal calculated by ZINDO (A). Calculated MO transitions after the geometry of the PETN molecul es was optimized (B).

PAGE 178

178 Figure 6 53. The calculated MO transitions by ZINDO of one molecule of 2, 4 DNT (A), 2, 4 DNT with geometry optimization (B), 2, 6 DNT (C), and 2, 6 DNT with geometry optimization (D).

PAGE 179

179 Figure 6 54. Molecular orbital tr ansitions calculated by ZINDO of one molecule of several low power explosive materials: 2 MNT (A), 2 MNT with geometry optimization (B), 3 MNT (C), 3 MNT with geometry optimization (D), 4 MNT (E), and 4 MNT with geometry optimization (F).

PAGE 180

180 Figure 6 5 5. The results of the calculated molecular orbital transition by ZINDO of one molecule of toluene (A), toluene with geometry optimization (B), benzene (C), and benzene with geometry optimization (D).

PAGE 181

181 Figure 6 56. Calculated molecular orbital t ransitions by ZINDO of derivatives of the TNT molecules as they would appear in the molecule: the nitro group (A) and the methylene group (B) and after geometry optimization of the nitro group (C) and the methylene group (D).

PAGE 182

182 CHAPTER 7 FURTHER DISCUSSION Chapter 5 introduced the low energy or long wavelength differential reflectogram features of several energetic materials. Additional experiments confirm that the features are not only unique to the individual energetic materia ls but they are also robust. The features exist independent of the sample treatment or supplier and retain their characteristics in mixes of energetic materials. Investigation of these results leads to the question of the physical origin of the long wave length feature. 7.1 Review of TNT Absorption The majority of the presented experiments are performed using TNT so the discussion will begin with its 420 nm feature. Few researchers have reported this result directly. In fact the only references to a la rge wavelength feature were by Cantillo and Felt et al 23,108 Cantillo observed an absorption edge near 400 nm for a solid sample of TNT using a plain reflection spectrometer and attributed it to the NO 2 groups of the molecule, due to spectra of NO 2 gas giving an equivalent absorption edge. The sa me research group had also reported near field reflectance microscopy measurements where a feature in the 395 to 405 nm range is seen 109 Felt et al. observed a broad optical absorbance peak near 400 nm for a TNT and hydroxide reaction in a relatively dilute solution. As the time of the hydroxide reaction increased the intensity of the ab sorption increased until it reached a saturation point and then decreased as the reaction continued. Downs and Forsyth reported an absorption onset for single crystals of TNT near 400 nm 98 Downs and Forsyth and Cantillo attributed this absorpt ion to an NO 2 groups based on the logic that it could not be from the benzene ring. Felt only described the absorption feature to be from a product of the TNT hydroxide reaction.

PAGE 183

183 7.2 Origin of the Long Wavelength Feature Several of the exper imental results in Chapter 5 give an indication of the origin of the 420 nm feature. The first addressed are the UV visible transmission spectrophotometry measurements of varying concentrations of solutions. As discussed in Section 5.8.1 the absorbance a t 420 nm decreased in intensity with the decrease in concentration but this trend was not completely linear. The trend had two regions of linearity; the higher concentrations above 1.7 g/L or 1700ppm and the lower concentrations below this value. This de viates from the Beer Lambert law which states that the absorbtivity of a chromophore or light absorbing species, decreases linearly with the concentration. Essentially the smaller number of chromophores that are present, the smaller the amount of absorpti on. A deviation from this behavior indicates that solvent molecules or other molecules. Measurements of TNT in the DRS after recrystallization from several diff erent solvents and UV visible transmission spectra where the solvent for the solution was varied show that neither affects the wavelength of the absorption feature at 420 nm. Therefore, it is inferred that the cause of the observed absorption is not an in teraction with the solvent. It is possible that the absorption at 420 nm occurs due to an interaction or the formation of a complex between molecules of the energetic material. The nature of this effect needs to be addressed. Marinkas et al. have observe d the same phenomenon as above in solid and supersaturated solutions of RDX and HMX 84,110 An absorption edge near 355 nm for RDX and HMX was found in the solid phase and solvated solutions but as the concentration was decreased the features were no longer observed. These results are compared to those reported in Section 5.1 and 5.8.1 where an absorption feature near 335 nm for RDX and 320 nm for HMX were seen using DRS. Similar wavelength transitions were observe d in UV visible absorbance of high

PAGE 184

184 concentration solutions. As the concentration s of the solutions were decreased, the intensity of the absorbance decreases until it is no longer measurable. Marinkas and his colleagues attributed this phenomenon to a cha rge transfer self complex that has a low intensity due to selection rules that render the transition forbidden and is a solid state effect. A charge transfer self complex is where an electron is exchanged between moieties (components of the molecule and c an be either atoms or small molecules) within the molecule or between large molecules that do not ordinarily share electrons. For illustration, consider an electron from the sp 2 molecular orbital of a ring N in an RDX molecule. This electron normally is shared with the nearest neighbors in the ring but not with the N O sp 2 orbital but instead in a charge transfer complex the electron is transferred to the N O orbital. 7.2.1 Solid State Effect? There are several interesting arguments to this proposal. F irst of all, the long wavelength transition was readily observed in the high concentrated solutions of TNT, RDX, HMX, Tetryl, and PETN suggesting that the transition is not a solid state effect. Marinkas and his colleague may not have observed this due to either using a low resolution spectrometer or solutions that were too dilute. The concentration of the solutions was not reported so the latter cannot be confirmed. In comparison, the particle size analysis of the high concentration solution of TNT meas ured in the UV Visible transmission spectrometry reported, was performed using dynamic light scattering. No particles (assuming a sphere) larger than 3 nm in size, the lower limit of the instrument, were observed in the solution after multiple trials, as described in Section 5.7. Using a sphere 3 nm in diameter and each molecule within the sphere is the equilibrium distance apart; this is the particle size equivalent to roughly 13 TNT molecules, or just shy of two unit cells. A group of thirteen TNT mole cules together is the largest a particle could be in the high concentration solution but there could well be far smaller particles present. The definition of a

PAGE 185

185 crystal is a substance having long range order; groups of less than 13 molecules suspended in a solution do not seem to meet this criterion 58 Second, the absorbance features were observed in concentrations much lower than the 3.3 g/L that was used for the particle size analysis. Therefore it is argued that the transition is an effect that is not constrained to the solid state. 7.2.2 Charge Transfer Self Complex It is well reported in the literature that the energetic materials, such as TNT, RDX, HMX, Tetryl, and PETN do not have strong intermolecular bonds in the crystalline state 41,66,69,89,111 In fact TNT, RDX, HM X, and Tetryl have been observed to have weak intra and intermolecular bonds that are described in the literature as hydrogen bonds although they are not the traditional hydrogen bonds These weak bo nds occur mostly between the O o f a nitro group and the H of either a ring carbon or methylene group (if one exists) where the bonded atoms are in the same molecule for intramolecular bonds or in neighboring molecules for intermolecular bonds. These bonds are in the form of a relatively strong electrostatic o r dipole dipole interaction where no electrons are shared rather, they cause the molecules to attract each other and influence the molecular orbitals. Charge transfer complexes can exist between molecules in highly concentrated solutions or solid crystals In the discussed system the complexes would be between molecules of the same type i.e. TNT to TNT. This supports the self complex portion of It is important to note that in the case of TNT the intermolecular interactions mentioned ab ove occur only with the molecules in the same plane 66,69,112 The interactions cause an increased distortion of the NO 2 groups from that of the single molecule to that of the molecule found in the molecular crystal. This has been reported in the literature and confirmed by computational results, se e Sections 6.1 and 6.8, where the molecular geometry for one isolated molecule of TNT, RDX, HMX, or Tetryl in a crystal changes significantly when the geometry is

PAGE 186

186 optimized in free space. PETN does not demonstrate hydrogen bonds in the crystal or the non linear decrease in absorption intensity with decreasing concentration for the UV visible absorbance measurements 76 78 In addition, the molecular geometry within a cr ystal and the single PETN molecule in free space with geometry optimization are not significantly different. This has been confirmed in the literature 78 Therefore it is proposed that the intermolecular interactions between molecules of TNT, RDX, HMX, and Tetryl contribute to the long wavelength absorption features 7.2.3 Cont ribution of the NO 2 Groups So far it has been determined that the low energy MO transition is not strictly confined to the solid state, is an intramolecular transition, and is affected by the rotation of the nitro group out of the equilibrium position. Th erefore it is deduced that the relevant MO transition occurs within the nitro group of the energetic molecule. In the case of TNT the experiments that have been conducted in conjunction with DNT, DNB, MNT, and MNB, can be used to determine in which nitro group the transition is located. The differential reflectograms of TNT, 2, 4 DNT, 2, 6 DNT, and DNB displayed features near 420 nm (Figure 5 19). This was also observed in the UV Visible absorbance spectra of these molecules and 2, MNT, 3, MNT, 4, MNT, a nd MNB that were summarized in Figure 5 28. TNT has the reddest (longest wavelength) transition of all of the molecules, followed by 2, 6 DNT, 2, MNT, and then the remaining materials which are very close together. Clearly, t he larger the number of nitro groups in the molecule the redder the MO transition. Using this logic, it is intriguing that the 2, 6 DNT and 1, 3 DNB transitions are blue shifted compared to 2, MNT. In addition the 4 MNT, 6 MNT and MNB are also blue shifted even though they have the same number of nitro groups as 2 MNT. It is proposed that not only the number of nitro groups affects the wavelength of the transition but also their position on the benzene ring. The ortho (2 and 6) nitro groups are never exactly planar to the

PAGE 187

187 benzene ring due to the steric hindrance caused by the large methylene group. The charge in the molecular orbitals of the methylene group attracts the lone pair of electrons on the single bonded oxygen and repels the double bonded oxygen. Meanwhile the para, or 4 nitro group is almost exactly planar in the single molecule and is slightly rotated due to an intermolecular hydrogen bond in the molecular crystal. Thus, the closer to the methylene group is to the nitro group, the greater the degree of rotation. This rotation increases when the molecule is in the molecular crystal. When the nitro groups are close to the methylene group then the MO transition is redder. This is most apparent in the DNB and MNB molecules where no methylene group is present on the ring and the transition is greatly blue shifted. In addition, the transition is bluer for 2, 4 DNT than for 2, 6 DNT and bluer for 3 and 4 MNT than for 2 MNT. Therefore it is concluded that the energy of the transition is lowered as the rotation of the nitro group increases and as the number of nitro groups increases. This conclusion affirms that the symmetry of the molecular orbitals of the nitro groups is essential to the wavelength and intensity of the long wavelength feature The computational chemistry results for the varying intermolecular distance between pairs of TNT molecules and the modified unit cell also support the stated conclusion. As the distance between the molecules increased, the interaction of the two molecules deceased and in the optimiz ed geometry calculations of the rotation of the nitro groups decreased as well. The resulting longest wavelength MO transition remained at the same wavelength significantly below 420 nm. But as the distance was decreased the interaction between the molec ules increased and the wavelength of the longest wavelength MO transition increased sharply. Similarly when a stacking fault is introduced several long wavelength transitions appeared in the area of 400 nm. In summary, as the molecules were stressed or f orced to interact with each other, the wavelength

PAGE 188

188 of the longest transition increased. These interactions often caused or maintained the rotation of the nitro groups out of the equilibrium position. Since it has been thus determined that the low energy tr ansition is due to the nitro groups and is dependent on the MO symmetry within these molecules, the specific molecular orbitals involved in the transition should be discussed. The orbitals involved can be inferred from a couple of experiments. One of the se is observing the effect of the solvent on the transition wavelength in UV Visible transmission spectrophotometry. The wavelength of the transition can be shifted if the solvent forms a complex with the molecular orbitals involved in the given transitio n. This occurs more often with polar solvents where the energy of a non bonding orbital is decreased subsequently incr is then observed for the transition with increasingly polar solvents. Other molecular transitions ot shift with increasingly polar solvents. One exception of this is when the molecule is in the excited state, a complex can form with the LUMO orbital causing the MO energy to decrease and the wavelength of the transition to red shift. Figure 5 32 shows the UV Visible transmission spectra for TNT in varying solvents. The 420 nm absorption feature did not significantly change with the polarity of the solvent although there was a slight red shift of about 2 nm. Therefore the The computation modeling allowed the user to visualize what molecular orbitals were contributing to a particular transition. The type of orbital was indicated by its shading and shape. In the case of the isolated TNT molecules where the 420 nm transition was observed, the transition involves several orbitals that are seen in Figure 6 11. The unoccupied (indicated as shaded) orbital for the 420 nm transition, is the sp 2 orbital of the nitrogen to the single bonded

PAGE 189

189 oxygen. The ground state of the electron is the non orbital) that is bonded to the nitro group. Therefore the MO transition is caused by a charge transfer from the ring carbon to the LUMO or the 7.3 Summary From the preceding discussion, the proposed origin of the long wavelength transition for TNT has been defined; a charge transfer self complex causes the transition which is only allowed when the mole cular orbitals have a particular symmetry. This theory essentially agrees with the explanation by Marinkas and his colleagues for the feature seen in solid RDX and HMX. A complete set of experiments was performed only for TNT; however, from the experimen ts performed for the energetic materials, the same theory can be applied. For instance, it has been discussed that RDX, HMX, and Tetryl displayed both long wavelength DR and UV Visible transmission spectral features that decreased non linearly with concen tration similar to TNT. These energetic materials have also been shown to have hydrogen bonding within the molecular crystal similar to TNT. Thus, the long wavelength transition in the energetic materials tested is a charge transfer self complex from a r ing atom to the unoccupied orbital of the nitro group.

PAGE 190

190 CHAPTER 8 SUMMARY AND CONCLUSIONS Energetic materials have unique optical absorption in the UV and blue visible range that is best measured by differential reflectometry. These characteristic features can be exploited for development of a commercial detec tion system for several reasons. Each individual energetic material has its own feature from 270 to 420 nm that is distinct from other organic and inorganic materials tested. Several items of vegetation did have some weak absorption near 400 nm. Howeve r, this was accompanied by additional features in the UV, distinguishing them from the energetic materials. This insures a low number of false positives and proper identification of the individual explosives. The absorption feature is a property of energ etic materials in that it is independent of the source, the solvent (if used), substrate, and degree and kind of impurities. In addition, the absorption can be confirmed by other UV reflection techniques as well as in transmission using sample solutions o f relatively high concentrations. It is important to note that non energetic cyclic materials do not have the described absorption features. The energetic materials retain their individual optical absorption features in mixtures of several different exp losives. It has also been shown that the feature in TNT is independent of crystal face orientation and angle of reflection of the incident light from the sample surface. The wavelength of the TNT feature is dependent on the number and the position of the nitro groups present in the molecule as determined by measuring several DNTs and DNB in the DRS and in UV Visible transmission spectrophotometry. When TNT, HMX, RDX, and Tetryl are in solution the presence and wavelength of the UV feature depends on the d egree of concentration. The concentration dependence deviates from the Beer Lambert law leading to the conclusion that the transition is dependent on its environment. This is further confirmed for TNT by the slight shift in wavelength of the solution

PAGE 191

191 due to the polarity of the solvent. From this shift, it is interpreted that the MO transition is not an the transition is due to a complex between sample molecules The absorption feature of interest in TNT is sensitive to high energy UV light, in that it disappears with prolonged exposure at least at the surface. A change in the measured TNT spectra is seen where only the near 420 nm feature diminished while the r emainder of the spectrum remains unchanged. From FTIR and chemical analysis, it is shown that this photo effect is not due to the photodissociation of the NO 2 groups creating DNT or MNT. Molecular modeling results for TNT from ZINDO and DFT calculations a re essentially similar showing confidence in the reported results. The closest duplication of experimental results and theory for TNT is achieved for three cases namely (1) involving one isolated conformation A molecule, (2) molecular pairs compressed fro m equilibrium spacing, and (3) a unit cell containing a twin n ing stacking fault. In the computational results, the MO transitions that most closely represent the long wavelength transition in TNT is from the HOMO of a ring carbon and the nitrogen oxygen m olecular orbital to the LUMO of the same nitrogen and oxygen molecular orbital. A number of other modeling attempts were conducted to possibly interpret the experimental results including, derivatives of TNT, components of the TNT molecule, varying numbers of molecules in a space, and pairs and unit cells of energetic materials. However, they did not lead to the UV or blue absorption features. Computational chemistry results for other energetic materials show interesting results as well. All the long wave length MO transitions are located within the molecular orbitals of the

PAGE 192

192 nitro groups in all of the materials similar to TNT. Unlike TNT, MO transitions at longer wavelengths than the experimental results w ere found for RDX, HMX, Tetryl and PETN. The experi mental and computational results lead to the conclusion that the observed transitions at the long wavelength end of the UV range are charge transfer self complexes. These complexes occur not only in the solid state but also in highly concentrated solution s in that the particular molecule must interact with a second identical molecule in order for the observed transition to be allowed. This conclusion can be extended to all of the tested energetic materials that display similar experimental and computation al results to TNT namely, RDX, HMX, and Tetryl.

PAGE 193

193 CHAPTER 9 FUTURE WORK The physical origin of the observed 420 nm absorption for TNT has been presented. While this explanation is sound, further work would solidify the proposed theory and applicability to other energetic materials. 9.1 Optical Characte rization The differential reflection work was performed using polycrystalline samples of the energetic materials and unpolarized light. As described in Section 3.2, the components of the dipole strength vector can determine the probability of a particular electron transition. This information can be found for a sample molecule by measuring the absorption using polarized light and a single crystal or thin film sample 93 The molecular orbitals in a molecule can be probed by light that is polarized in the same direction a s their dipole moment components. If the polarity is in a different direction than the dipole moment component of molecular transition, it will not be allowed and no interaction with the molecular orbital will occur. From this, the specific molecular orb itals involved in the transition of interest are determined. In regards to the samples, several processes for growing single crystals of TNT has been describe in the literature. However, the applicability of the crystals to DRS measurements is questiona ble 78,84,98 DRS depends on the presence of an inhomogeneous sample in order for a difference in reflectivity to be detected; a single crystal may be too homogeneous to yield detectable absorption features. Therefore a thin film process is recommended. The DRS instru ment would also need to be fitted with a polarizer for the incident and the reflected light from the sample. The vast majority of the experimental work was performed using TNT due to its availability and common use in explosive devices. Optical characte rization, particularly the

PAGE 194

194 experiments using the UV Visible spectrophotometer, should be repeated with other energetic materials, particularly the recently used liquid explosives based on peroxide and acetone. It is proposed to perform further experimenta tion using the energetic materials in different solvents and investigating more closely the behavior of the absorption peak of interest with decreasing concentration are proposed. The results can then be compared to those of TNT and the applicability of t he proposed theory investigated. As mentioned in Section 4.1.1 the optical constants of energetic materials are not widely known inhibiting the accuracy of line shape analysis. This problem can possibly be addressed by using a spectroscopic ellipsometer t o measure the optical constants for a single crystal sample over a range of wavelengths. 9.2 Photolysis of Energetic Materials The photolysis of TNT was briefly described and investigated in the reported research. There are clearly several processes in th e samples occurring concurrently during high energy UV radiation. A limitation of light sources and their power output allowed only a few experiments to be performed. Using stronger light sources of different wavelengths would make additional experiments more successful and meaningful. In addition, several of the other energetic materials measured by the DRS have a C N bond that could potentially be photodissociated with high energy light. The effect of UV radiation on these materials should also be inv estigated. Additional chemical analysis on the irradiated samples along with the relationship to the molecular crystal structure may provide more information on the observed phenomena. 9.3 Computational Chemistry Chapter 6 demonstrated the great amount of information that can be discovered using computational chemistry methods. It is admitted that a simplified method was used for the calculations. A more complicated method such as time dependant density functional theory

PAGE 195

195 (TDDFT) would probably be able to show more molecular orbital transitions. In addition, calculations involving far more than eight molecules could be accomplished with a more powerful computer, which may better represent a real system

PAGE 196

196 APPENDIX A EXTENDED BACKGROUND OF MOLECULAR ORBITAL THEORY A.1 Electrons An atom is constructed mainly of protons, neutrons, and electrons. The protons and neutrons are primarily located in the mass center of the atom termed the nucleus, while the electro ns occupy the space outside the nucleus as an electron cloud. In essence the electrons orbit the atom nucleus as the planets orbit the sun. This orbit is not a defined path like the planets but instead a position outside the nucleus of the atom where the electron has a high probability of occupying 93 The exact location of an electron cannot be pin pointed. Only the effects of an electron are seen and can be found daily in cathode ray tube televisions and in science by electron microscopy techniques. An electron portr ays both particle and wave characteristics, making it a very unique and fascinating entity. The particle behavior is embodied by the electron having a defined mass and charge. This is seen when the velocity vector of an electron is disturbed by the prese nce of an external electric field, such as in a cathode ray tube. A wave is characterized as having periodicity in space and time and defined by a velocity, wavelength, and a frequency. Electron beams are examples of these wave characteristics. The wave characteristics of an electron are mathematically described by a wave function as contained in the Schrdinger equation. A time dependent Schrdinger equation defines the wave with respect to space and time. The physical behavior of the electron can be c alculated from this equation and is the basis of nearly all the principles in this research. A.2 Atomic orbitals As stated in Section A.1 the electrons of an atom occupy the space surrounding the nucleus in an atom. The manner in which they occupy this space is determined by the number of

PAGE 197

197 electrons present and their energies. Even though the electrons do no have a defined position in the atom, since they exist in an electron cloud, they are expressed to occupy defined energy levels. These levels are of ten termed as orbitals because they are depicted as orbits surrounding the center nucleus much like the solar system. For simplicity, simple integers termed quantum numbers are used to describe the electrons in an atom that are represented in the wave eq uation. The first (principle) quantum number (n) describes the energy as a number that is always positive and whole, the smallest possible being one (ground state). Atomic orbital shape is described by the second (azimuthal) quantum number (l). The value of l is determined by the angular momentum of the electron due to the orbit around the nucleus. By quantum theory the shape of an orbital is restricted by the energy of the orbital. Therefore, the smaller n is the smaller the orbital and consequently th e possible shape of the orbital is more restricted. For example, hydrogen atom n is equal to one and the orbital shape can only be spherical. The second quantum number is most commonly represented by the letters s, p, d, and f that correspond to the diff erent shapes of the atomic orbitals. The last two quantum numbers are the magnetic and the spin quantum numbers m l and m s momentum) is described by the magnetic quant um number. This is not applied to s orbitals because they are spherical and do not have a preferred orientation. These quantum numbers are a positive or negative whole integer between 1 and 1. The spin quantum number has two possible values spin up (+1 /2) and spin down ( 1/2) that represent the spin direction of the electron. The Pauli exclusion principle states that no two electrons in an atom can have the same set of quantum numbers. This simply declares that no two electrons can occupy the same pla ce at the same time, which is intuitive. The exclusion principle also determines how the electrons in

PAGE 198

198 an atom are distributed. For example, an oxygen has eight electrons, so with the exclusion principle in mind, the electrons are distributed as two in 1s two in 2s, and four in 2p. The quantum numbers described determine the symmetry of the atomic orbitals which in turn express the molecular orbitals and have a great impact on the energies of these molecular orbitals and the transitions that are allowed. A.3 Molecular orbitals Atomic orbitals are combined in the bonding of two or more atoms creating molecules. As two atoms approach each other in space and their electrons interact for covalent bonding, their atomic orbitals engage as well. Depending on th e atomic orbitals of the individual atoms, the electrons are redistributed between the two atoms so that the electrons are shared and the total energy of the new molecule is minimized. Consequently the atomic orbitals combine into molecular orbitals where the electrons now shared by more than one atoms in the molecule. The molecular orbitals described above are formed by the linear combination of the contributing atomic orbitals. That is, the wave functions of the molecular orbitals are simply added tog ether to form one new orbital. Quantum mechanics dictates that the linear combination of the two atomic orbitals must yield two molecular orbitals. Hence the atomic orbitals are subtracted, another linear combination, giving the second unoccupied orbital or anti bonding orbital. Non bonding orbitals are not combined because they do not participate in the bonding of the molecule and are usually seen as lone electrons in the valence orbitals of the atom and molecule. The molecular orbitals are named accord ing to their function in the bond 62 Bonding orbitals of the atoms respectively. The non bonding orbitals n contain electrons that are not bonded to another atom, while anti bonding orbita

PAGE 199

199 energy of the molecular orbitals is lower than the energy of the atomic orbitals shown in Figure A 1. This reflects the h igher stability and lower overall energy of a molecule than the individual atoms. Large molecules are composed of several smaller molecules. When these are combine the resulting molecular orbitals can be distributed throughout the entire molecule, among bonded molecules, or remain within the small molecule. This can be observed in the molecular orbital energy diagram of TNT. TNT is synthesized by nitrating toluene where nitro (NO 2 ) groups are substituted on the benzene ring at available sites. The nit ro groups are electron withdrawing in that they take an electron from the benzene ring. This alters the energy of the benzene molecular from the atomic orbi tal of the C of the benzene ring and the N of the nitro group. However, there are no molecular orbitals between the oxygen of the nitro groups and the ring carbons. Similar to space groups for crystal structures molecules have a defined symmetry that be longs to a specific point group 93 The influence of the symmetry operators on the molecule determines the symmetry of its molecular orbitals. This symmetry is reflected in the wave equations that describe the molecular orbitals.

PAGE 200

200 Figure A 1. Simplified molecular orbital energy diagram for the combination of two 2p atomic orbitals 93

PAGE 201

201 LIST OF REFERENCES 1 M. W. Todd, R. A. Provencal, T. G. Owano, B. A. Paldus, A. Kachanov, K. L. Vodopyanov, M. Hunter, S. L. Coy, J. I. Steinfeld, and J. T. Arnold, Applied Physics B Lasers and Optics 75, 367 376 (2002). 2 D. M. Sheen, D. L. McMakin, and T. E. Hall, Abstracts of Papers of the American Chemical Society 230, U328 U329 (2005). 3 G. P. Kulemin and V. B. Razskazovsky, Ieee Transactions on Antennas and Propagation 45, 740 743 (1997). 4 I. Morino, K. M. T. Yamada, and A. G. Maki, Journal of Molecular Spectroscopy 196, 131 138 (1999). 5 J. Yinon and S. Zitrin, The analysis of explosives Vol. 3 (Pergamon, USA, 1981). 6 E. R. Menzel, L. W. Menzel, and J. R. Schwierking, TheScientificWorldJOURNAL 4, 725 735 (2004). 7 E. R. Men zel, K. K. Bouldin, and R. H. Murdock, TheScientificWorldJOURNAL 4, 55 66 (2004). 8 S. K. Sharma, A. K. Misra, and B. Sharma, Spectrochimica Acta Part a Molecular and Biomolecular Spectroscopy 61, 2404 2412 (2005). 9 K. Albert, M. L. Myrick, S. Brown, D. James, F. Milanovich, and D. Walt, Environmental Science Technology 35, 3193 3200 (2001). 10 R. E. Hummel, D. B. Dove, and J. A. Holbrook, Physical Review Letters 25, 290 292 (1970). 11 R. E. Hummel, A. Fuller, C. Schollhorn, and P. Holloway, Applied Ph ysics Letters 88, 1903 1905 (2006). 12 C. Schoellhorn, A. Fuller, J. Gratier, and R. E. Hummel, Applied Optics 46 (2007). 13 C. Schoellhorn, A. Fuller, J. Gratier, and R. E. Hummel, in New developments on standoff detection of explosive materials by diff erential reflectometry Orlando, FL, 2007 (SPIE). 14 R. E. Hummel, A. Fuller, C. Schoellhorn, and P. Holloway, in Trace chemical sensing of explosives edited by R. L. Woodfin (John Wiley, N.Y., 2007), p. 301. 15 A. Fuller, R. E. Hummel, C. Schoellhorn, and P. Holloway, in Stand off Detection of Explosive Materials by Differential Reflectometry Boston, MA, 2006 (SPIE).

PAGE 202

202 16 D. Stull, Fundamentals of Fire and Explosions (American Institute of Chemical Engineers, New York, 1977). 17 J. Yinon, Trends in ana lytical chemistry 21, 292 301 (2002). 18 N. E. Ipe, A. Akery, P. Ryge, D. Brown, F. Liu, J. Thieu, and B. James, Radiation Protection Dosimetry 116, 347 351 (2005). 19 R. M. Deas, M. J. Gaskell, K. Long, N. F. Peirson, M. D. Rowe, and J. A. S. Smith, Pro ceedings of the SPIE 5415, 510 520 (2003). 20 G. Bakken, G. Kauffman, P. Jurs, K. Albert, and S. Stitzel, Sensors and Actuators B 79, 1 10 (2001). 21 W. A. Schroeder, P. E. Wilcox, K. N. Trueblood, and A. O. Dekker, Analytical Chemistry 23, 1740 1747 (19 51). 22 A. Usachev, Applied Spectroscopy 55, 125 136 (2001). 23 D. Felt, S. Larson, and E. Valente, Chemosphere 49, 287 295 (2002). 24 R. L. Prasad and R. Prasad, Spectrochimica acta Part A 58, 3093 3102 (2002). 25 D. Stahl and D. Tilotta, Environmenta l Science Technology 35, 3507 3512 (2001). 26 K. J. Smit, Journal of Energetic Materials 9, 82 103 (1991). 27 K. W. Weissmahr, S. B. Haderlein, R. P. Schwarzenbach, R. Hany, and R. Nuesch, Environmental Science & Technology 31, 240 247 (1997). 28 S. Con tent, W. Trogler, and M. Sailor, Chem Eur J. 6, 2205 2213 (2000). 29 S. Glazier, J. Barron, N. Morales, A. M. Ruschak, P. Houston, and H. Abruna, Macromolecules 36, 1272 1278 (2003). 30 W. J. Salcedo, F. J. R. Fernandez, and J. C. Rubin, Spectrochimica A cta Part A 60, 1065 1070 (2004). 31 H. Sohn, Journal of the American Chemical Society 125, 3821 3830 (2003). 32 J. Goodpaster and V. McGuffin, Analytical Chemistry 73, 2004 2011 (2001). 33 J. C. Carter, S. M. Angel, M. Lawrence Snyder, J. Scaffidi, R. E Whipple, and J. G. Reynolds, Applied Spectroscopy 59, 769 775 (2005).

PAGE 203

203 34 J. Janni, B. D. Gilbert, R. W. Field, and J. I. Steinfeld, Spectrochimica Acta Part a Molecular and Biomolecular Spectroscopy 53, 1375 1381 (1997). 35 I. R. Lewis, N. W. Daniel Jr ., N. C. Chaffin, P. R. Griffiths, and M. W. Tungol, Spectrochimica acta Part A 51, 1985 2000 (1995). 36 M. L. Lewis, I. R. Lewis, and P. R. Griffiths, Vibrational Spectroscopy 38, 17 28 (2005). 37 M. L. Lewis, I. R. Lewis, and P. R. Griffiths, Vibration al Spectroscopy 38, 11 16 (2005). 38 D. S. Moore, Review of Scientific Instruments 75, 2499 2512 (2004). 39 C. McHugh, R. Keir, D. Graham, and W. E. Smith, Chem Communications 580 581 (2002). 40 C. J. McHugh, W. E. Smith, and D. Graham, Abstracts of Pa pers of the American Chemical Society 222, U86 U87 (2001). 41 D. Graham, A. R. Kennedy, C. J. McHugh, W. E. Smith, W. I. F. David, K. Shankland, and N. Shankland, New Journal of Chemistry 28, 161 165 (2004). 42 C. J. McHugh, W. E. Smith, R. Lacey, and D. Graham, Chemical Communications 2514 2515 (2002). 43 C. J. McHugh, R. Keir, D. Graham, and W. E. Smith, Chemical Communications 580 581 (2002). 44 Y. Hu, P. Huang, L. Guo, X. Wang, and C. Zhang, Physics Letters A 359, 728 732 (2006). 45 M. C. Kemp, P F. Taday, B. E. Cole, J. A. Cluff, A. J. Fitzgerald, and W. R. Tribe, in Security applications of terahertz technology Orlando, FL, 2003 (SPIE), p. 44 52. 46 J. Federici, R. Barat, D. Gary, Z. H. Michalopoulou, and D. Zimdars, American Laboratory 38, 2 8 33 (2006). 47 H. Zhong, A. Redo Sanchez, and X. C. Zhang, Optics Express 14, 9130 9141 (2006). 48 T. Arusi Parpar, D. Heflinger, and R. Lavi, Applied Optics 40, 6677 6681 (2001). 49 G. Boudreaux, T. Miller, A. Kunefke, J. Singh, F. Y. Yueh, and D. Mon ts, Applied Optics 38, 1411 1417 (1999). 50 R. L. Pastel and R. C. Sausa, Applied Optics 39, 2487 2495 (2000). 51 M. O. Rodgers, K. Asai, and D. D. Davis, Applied Optics 19, 3597 3605 (1980).

PAGE 204

204 52 J. Shu, I. Bar, and S. Rosenwaks, Applied Optics 38, 4705 4710 (1999). 53 J. B. Simeonsson and R. C. Sausa, Applied Spectroscopy Reviews 31, 1 72 (1996). 54 V. Swayambunathan, G. Singh, and R. Sausa, Applied Optics 38, 6447 6454 (1999). 55 D. Wu, J. Singh, F. Yueh, and D. Monts, Applied Optics 35, 3998 4003 (1 996). 56 C. Bohling, D. Scheel, K. Hohmann, W. Schade, M. Reuter, and G. Holl, Applied Optics 45, 3817 3825 (2006). 57 F. C. De Lucia, R. S. Harmon, K. L. McNesby, R. J. Winkel, and A. W. Miziolek, Applied Optics 42, 6148 6152 (2003). 58 R. E. Hummel, E lectronic Properties of Materials 3 ed. (Springer Verlag New York Inc., New York, 2001). 59 J. Akhavan, The Chemistry of Explosives second ed. (The Royal Society of Chemistry, Cambridge, 2004). 60 R. Gilardi and J. Karle, The structural Investigation o f Energetic Materials: Chemistry of Energetic Materials (Academic Press Inc., San Diego, CA, 1991). 61 H. Hart, L. Craine, and D. Hart, Organic Chemistry: A short course tenth ed. (Houghton Mifflin Company, New York, 1999). 62 J. Yinon and S. Zitrin, Th e analysis of explosives Vol. 3 (Pergamon, USA, 1983). 63 R. Meyer, Explosives Third ed. (VCH, Weinheim, Germany 1987). 64 P. Corfield, Chemistry & Industry 4 4 (2006). 65 H. G. Gallagher, R. M. Vrcelj, and J. N. Sherwood, Journal of Crystal Growth 2 50, 486 498 (2003). 66 W. R. Carper, L. P. Davis, and M. W. Extine, Journal of Physical Chemistry 86, 459 462 (1982). 67 J. Clarkson, W. E. Smith, D. Batchelder, and A. Smith, Journal of Molecular Structure 648, 203 214 (2003). 68 H. G. Gallagher, K. J. Roberts, J. N. Sherwood, and L. A. Smith, Journal of Materials Chemistry 7, 229 235 (1997).

PAGE 205

205 69 W. R. Carper, S. R. Bosco, and J. J. P. Stewart, Spectrochimica Acta Part a Molecular and Biomolecular Spectroscopy 42, 461 466 (1986). 70 P. C. Chen and W. L o, Theochem Journal of Molecular Structure 397, 21 32 (1997). 71 C. S. Choi and E. Prince, Acta Crystallographica Section B Structural Crystallography and Crystal Chemistry B 28, 2857 2862 (1972). 72 M. M. Kuklja and A. B. Kunz, Journal of Physical Chemi stry B 103, 8427 8431 (1999). 73 W. L. Elban, R. W. Armstrong, K. C. Yoo, R. G. Rosemeier, and R. Y. Yee, Journal of Materials Science 24, 1273 1280 (1989). 74 C. S. Choi and H. P. Boutin, Acta Crystallographica Section B Structural Crystallography and C rystal Chemistry B 26, 1235 1240 (1970). 75 H. H. Cady, Acta Crystallographica 23, 601 609 (1967). 76 Y. A. Gruzdkov and Y. M. Gupta, Journal of Physical Chemistry A 105, 6197 6202 (2001). 77 P. S. Makashir and E. M. Kurian, Propellants Explosives Pyrot echnics 24, 260 265 (1999). 78 M. M. Kuklja and A. B. Kunz, Journal of Applied Physics 89, 4962 4970 (2001). 79 D. Pavia, G. Lampman, and G. Kriz, Introduction to Spectroscopy (Harcourt, Inc., Orlando, 2001). 80 P. W. Atkins, Physical Chemistry Third e d. (Oxford University Press, Oxford, 1986). 81 T. Abe, Bulletin of the Chemical Society of Japan 31, 904 907 (1958). 82 M. J. Kamlet, H. G. Adolph, and J. C. Hoffsommer, Journal of the American Chemical Society 84, 3925 3928 (1962). 83 C. P. Conduit, Jou rnal of the Chemical Society 3273 3277 (1959). 84 P. L. Marinkas, J. E. Mapes, D. S. Downs, P. J. Kemmey, and A. C. Forsyth, Molecular Crystals and Liquid Crystals 35, 15 25 (1976). 85 J. Stals, Barraclo.Cg, and A. S. Buchanan, Transactions of the Farad ay Society 65, 904 914 (1969). 86 M. K. Orloff, P. A. Mullen, and F. C. Rauch, Journal of Physical Chemistry 74, 2189 2192 (1970).

PAGE 206

206 87 J. N. Maycock, V. R. Verneker, and W. Lochte, Physica Status Solidi 35, 849 851 (1969). 88 F. Pristera, M. Halik, A. Cas telli, and W. Fredericks, Analytical Chemistry 32, 495 508 (1960). 89 P. S. Makashir and E. M. Kurian, Journal of Thermal Analysis and Calorimetry 55, 173 185 (1999). 90 J. A. Ciezak and S. F. Trevino, Journal of Physical Chemistry A 110, 5149 5155 (2006 ). 91 W. M. Trott and A. M. Renlund, Applied Optics 24, 1520 1525 (1985). 92 N. F. Fell, J. M. Widder, S. V. Medlin, J. B. Morris, R. A. PesceRodriguez, and K. L. McNesby, Journal of Raman Spectroscopy 27, 97 104 (1996). 93 M. Orchin and H. H. Jaffe, Sy mmetry, Orbitals, and Spectra (S.O.S.) (John Wiley & Sons, Inc., New York, 1971). 94 R. E. Hummel, in Encyclopedia of Analytical Chemistry edited by R. A. Meyers (John Wiley and Sons Ltd, Chichester, 2000), p. 9047 9071. 95 R. E. Hummel, Physica Status Solidi a Applied Research 76, 11 44 (1983). 96 J. A. Holbrook and R. E. Hummel, Review of Scientific Instruments 44, 463 466 (1973). 97 R. Enderlein, R. E. Hummel, J. B. Andrews, R. J. Nastasiandrews, and C. W. Shanley, Physica Status Solidi B Basic Rese arch 88, 173 178 (1978). 98 99 R. E. Hummel and R. Enderlein, Physical Review B 29, 1529 1533 (1984). 100 B. O. Seraphin and N. Bottka, Physical Review 145, 628 636 (196 6). 101 R. M. Vrcelj, H. G. Gallagher, and J. N. Sherwood, Journal of the American Chemical Society 123, 2291 2295 (2001). 102 ASM Handbook Volume 10: Materials Characterization ; Vol. edited by R. E. Whan (ASM international, USA, 1986). 103 C. Cramer, Essentials of computational chemistry: theories and models 2nd ed. (Wiley, Hoboken, NJ, 2004). 104 W. R. Mabey, D. Tse, A. Baraze, and T. Mill, Chemosphere 12, 3 16 (1983).

PAGE 207

207 105 D. C. Schmelling, K. A. Gray, and P. V. Kamat, Environmental Science & Techn ology 30, 2547 2555 (1996). 106 M. A. Ronning, R. L. Atkins, and C. A. Heller, Journal of Photochemistry 9, 403 405 (1978). 107 H. H. Cady, D. T. Cromer, and A. C. Larson, Acta Crystallographica 16, 617 623 (1963). 108 C. Cantillo, Thesis, University of Puerto Rico, 2006. 109 L. M. Gomez, C. Osorio, E. Amman, S. P. Hernandez, and M. E. Castro, Chemical Physics Letters 422, 313 316 (2006). 110 P. L. Marinkas, Journal of Luminescence 15, 57 67 (1977). 111 F. D. Lewis and B. A. Yoon, Journal of Organic C hemistry 59, 2537 2545 (1994). 112 C. P. Nash, T. E. Nelson, J. J. P. Stewart, and W. R. Carper, Spectrochimica Acta Part a Molecular and Biomolecular Spectroscopy 45, 585 588 (1989).

PAGE 208

208 BIOGRAPHICAL SKETCH Anna Fuller was born in Albuquerque, New Mexico on July 30, 1980. She graduated from La Cueva High School in 1998 and enrolled in the University of Arizona in Tucson, AZ for the fall semester. Anna graduated from the University of Ari zona in May 2002 with her Bachelor of Science in materials science and engineering specializing in electronic materials. She then began attending the University of Florida in August of 2002 and received her Master of Science in materials science and engin eering in December 2004. Anna greatly anticipates working in research and development to sharpen her skills and looks to eventually becoming a professor.