PHOTOCHEWIICAL RE ACTIONS OF CHLORO AROMATIC COMPOUNDS
A Thesis
Presented to
The Faculty of Graduate Studies
of
The University of Guelph
by
ALEXANDRE KONSTANTINOV
In partial fiilfillment of requirements
for the degree of
Doctor of Philosophy
January, 1999
@ Alexandre Konstantinov, 1999 National Library Bibliothèque nationale du Canada Acquisitions and Acquisitions et Bibliographie Services services bibliographiques 395 Wellington Street 395. nie Wellington Ottawa ON KIA ON4 Ottawa ON K1A ON4 Canada Canada Your Ma Voue derence
Our Ne Nom reterem
The author has granted a non- L'auteur a accordé une licence non exclusive licence ailowing the exclusive permettant a la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sen reproduire, prêter, distribuer ou copies of this thesis in microfom, vendre des copies de cette thèse sous paper or electronic formats. la fome de microfiche/flm, de reproduction sur papier ou sur format électronique.
The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation. PHOTOCHEMICAL REACTIONS OF CHLOROAROMATIC COMPOUNDS
Alexandre Konstantinov Advisor: University of Guelph, 1998 Dr. N.J. Bunce
A novel reaction of successive photocyanation cornmon for various classes of highly chlo~atedaromatic compounds was discovered. The products of photolysis of polychlorinated aromatics in the presence of sodium cyanide were polycyanated hydroxychlorocompounds with various degrees of chlorine replacement. Products from some substrates were isolated, identified and characterized.
The quantum yields of disappearance in the presence of sodium cyanide were determined for a number of structurally diverse chloroaromatic compounds. The quantum efficiency of photocyanations was found to increase with the number of chlorine substituents on a substrate. Sensitization and quenching experiments established the triplet excited state to be reactive for al1 tested compounds, consistent with the suggested
SN2Ar* mechanism for the successive photocyanation. To date, the synthetic potential of the reaction is low because of low selectivity.
In related work, a method for on-site destruction of dioxins in smd quantities of liquid laboratory waste using UV light was developed and applied to both standard congeners and waste sarnples from an analytical laboratory. The novel feaiure was the use of a low power, low pressure mercury arc as the radiation source. The problem of poor matend balance in photolysis of dioxins was resolved by photolyzing tritiated 2,3,7,8-TCDD as a tracer compound and locating the missing material in the polar fractions. These products were identified by ES-MS as hydroxylated aromatic compounds, indicating that as much as 80% of the photolysis proceeds by C-O rather than C-Cl homolysis.
Biological assays demonstrated that the products formed upon photolysis of
2,3,7,8-TCDD lost the toxic effects associated with dioxins' receptor binding ability and did not bind to the estrogen receptor. This information is very important for the practical applications of the UV treatment methods, because it shows that the photolysis products are Likely to have low toxicity, and cm therefore be disposed of by conventional means. I would like to thank a number of people who directly or indirectly contributed to the success of this project. First of dl, Dr. N. J. Bunce, who was an outstanding supervisor during my M.Sc. and Ph.D. studies at the University of Guelph. His thorough knowledge in various fields of chemistry and open-minded style of guidance were a great deal of help in my research. I would like to thank my committee members Dr. 3.
Lipkowski, R. McCrindle and M. Tchir for their assistance, usefûl advise, and encouragement.
Special thanks to Dr. C. Kingsmill and Dr. D. Suh f?om Laboratory Services
Division, University of Guelph, for prompt analysis on a number of samples using the
Electrospray Mass Spectrometer.
Kudos to Dr. G. Ferguson for x-ray crystallography and to Valerie Robinson for running 2-D NMR experiments.
My appreciation to Brian Cox and John Petrulis for successfbl CO-operationin biochemicai assay experiments, and thanks to my other lab-mates for being nice people.
1 would also like to express my gratitude to the outstanding suppon staff of the
Department of Chemistry and Biochemistry, including Teny White - workshop, Yves
Savoret - glassblowing, Karen Shiell -purchasing, Uwe Oehler, Ian Renauld and Steve
Seifried - network support and electronics shop.
My very special thanks to my wife Svetlana for support and patience.
2.3. Conclusions ...... 82
2.4. Expenmental ...... 83 2.4.1. Generai ...... ,...... 83 2.4.2. Irradiations ...... 83 2.4.3. Andysis ...... 83 2.4.4. Mass Spectrometv ...... 84 2.4.5. Photochernical Cyanation of Hexachlorobenzene...... 85 2.4.5.1. Measurement of Quantum Yield of Disappearance of HCB ...... -85 2.4.5.2. Photolysis of Hexachlorobenzene in the Presence of NaCN for Various Penods of Time ...... 87 2.4.5.3. Sensitization of HCB Photocyanation by Acetone and Acetophenone, and Oxygen Removal ...... 88 2.4.5.4. Quenching of HCB Photocyanation by Ferrocene ...... 89 2.4.5.5. Photolysis of HCB at Various Concentrations of Sodium Cyanide .. -89 2.4.5.6. Influence of Water on Photocyanation of HCB ...... 90 2.4.5.7. Determination of Isosbestic Point in Cyanation of Hexachlorobenzene ...... 91 2.4.5.8. Dependance of Quantum Yield of Cyanation of HCB on the Initial Concentration of HCB ...... 92 2.4.5.9. Photocyanation of HCB on Preparative Scale ...... 92 2.4.6. Synthesis of 2,3,4,5,6-pentachlorobenzonitrile...... -96 2.4.6.1. Iodination of 4-chloroaniline ...... 96 2.4.6.2. Polychlorination of 1-iodo-4chlorobenzene ...... 96 2.4.6.3 Cyanation of Pentachloroiodobenzene...... 97 2.4.7. Reactions of 2,3,4,5, 6-Pentachlorobenzonitnlewith Sodium Cyanide ...... -97 2.4.8. Calculation of Quantum Yield of Disappearance of Pentachlorobenzonitnle98 2.4.9. Synthesis and Ground-State Reactions of 1,4 -Dichloro-2,3,5, 6- Tetracyanobenzene and a Trichloro-Tncyanobenzene...... 99 2.4.10. Competitive Photocyanation of HCB, Pentachlorobenzene and 1,2,3, 4-Tetrachlorobenzene ...... ~...... 100 2.4.1 1. Photocyanation of 1,4-Dicyanobenzene and 1,2,4,5-Tetracyanobenzene. .. 10 1 2.4.12. Photocyanation of Decachlorobiphenyl ...... 103 2.4.12.1. Photolysis of Decachlorobiphenyl in the Presence of Sodium Cyanide for Various Penods of Time ...... 103 2.4.12.2. Measurement of Quantum Yield of Disappearance of Decachlorobiphenyl ...... 104 2.4.12.3. Quenching of Cyanation of DCB by Ferrocene ...... 105 2.4.12.4. Photocyanation of DCB at Various Concentrations of Water ...... 1O6 2.4.12.5. Photolysis of Decachlorobiphenyl at Various Concentrations of Sodium Cyanide ...... 106 . 2.4.12.6. Sensitization of DCB Photocyanation by Acetone, Acetophenone and Oxygen Rernoval...... 107 2.4.12.7. Competitive Photolysis of Three Hexachlorobiphenyls ...... 108 2.4.12.8. Photolysis of 2,3,4,5, 6-Pentachlorobiphenyl in the Presence of Sodium Cyanide for Various Periods of Time ...... 110
List of Figures
Figure 1. Charge Distribution in Ground and Singlet Excited States . of 4-NitrocatechoI ...... 8 Figure 2 . Para-Directing Effect of Nitro Group in the Photo-Smiles Rearrangernent ...... -9 Figure 3 . Net charges and the H orbital coefficients of the Tl state and net charges of the radical anion of 4-nitrophenol ...... 10 Figure 4 . Pathways of Photocyanation of LMethoxynaphthalene ...... 19 Figure 5 . Mechanism of Nucleophilic Substitution of Haloanisoles in Aqueous t-BuOH ...... 21 Figure 6 . Mechanism of Photocyanation of Halonaisoles Involving the Radical Ion Pairs ...... 22 Figure 7. Mechanism of P hotocyanation of Aromatic Compounds in the Presence of Electron Acceptors ...... -25 Figure 8. Mechanism for the Photocyanation of Naphthalene in the Presence of 1,4-Dicyanobenzene...... -27 Figure 9 . Mechanism of Aromatic Cyanation by CuCN in Ground State...... 30 Figure 10. Mechanism of Cyanation of Aryl Halides by tris(Tripheny1phosphine)Nickel and NaCN ...... 31 Figure 11 . Mechanism of Palladium Complex CataIyzed Cyanation of Iodobenzene ..... 32 Figure 12. An ORTEP Diagram of a Phenolate Ion of 27 ...... -41 Figure 13. A View of Part of the Layer Structure of Crystal Asymmetnc Unit of 27 ...... 41 Figure 14. Resonance Structures of the Phenoxide Ion of 27 ...... 42 Figure 15. Dark vs. Photochernical Cyanation of Pentachlorobenzonit~leeee...... 43 Figure 16. Determination of Isosbestic Point in the UV Spectra of Photocyanation of HCB ...... -44 Figure 17. Dark Reactions of 1,4-Dichloro-2,3,5,6-tetracyanobenzene 29 ...... -45 Figure 18. Estimation of a Quantum YieId for Disappearance of HCB ...... 48 Figure 19. Pseudo-first order Kinetics in Photocyanation of HCB and . Pentachlorobenzonitrile ...... -48 Figure 20 . Double Reciprocal Plot of qhSversus PCB] ...... 50
Figure 21 . Double Reciprocal Plot of q5dïSs of HCB ver- [NaCN] ...... 51 Figure 22. A Stem-Volmer Plot of Quenching of HCB Photocyanation by Ferrocene .... 52 Figure 23 . Photocyanation of Tetra-, Tri- and Dichlorobenzenes ...... 56 Figure 24 . Measwement of Pseudo-first Order Rate Constants for Photocyanation of HCB, Pentachlorobenzene and 1,2,3,4-Tetrachlorobenzene in Solutions with Equals Absorbancies ...... 57
Figure 25 . Photocyanation of 1,CdicyanO- and 1,2,4,5-tetracyanobenzenes...... -58 Figure 26 . Possible Mechanism of Photocyanation of 1.2.4.5.tetracyanobenzene ...... -59 Figure 27 . Photo verms "Dark" Cyanation of 1.2.4.5-tetracyanobenzene . Pseudo-first Order Plots ...... -61 Figure 28 . Pseudo-first Order Kinetics of Photolysi s of Decachlorobip henyl in the Presence and Absence of Sodium Cyanide ...... 63 Figure 29 . Determination of Quantum Yield of Disappearance of Decachlorobiphenyl ...... 64 Figure 30 Quenching of Photocyanation of DCB by Ferrocene ...... -65 Figure 3 1 A Double Reciprocal Plot of #d, of DCB verms WaCN] ...... 66 Figure 32 influence of Water on the Quantum Yield of Disappearance of DCB ...... 66 Figure 33 Psuedo-first Order Plots of Photocyanation of Hexachlorobiphenyls 35, 36 and 37 ...... 69 Figure 34 Measurement of Pseudo-first Order Rate Constants of Photocyanation for 4-hydroxy-2', 3 ',4', 5 '-tetrachlorobiphenyl (38) and 3-hydroxy-2,4, 6- trichlorobiphenyl (40) ...... 72 Figure 3 5. COSY Spectrum of Compound 42 ...... 76 Figure 3 6 . HMBC Spectrum of Compound 42 ...... 77 Figure 3 7. Photolysis of OCDD in the Presence and absence of NaCN . Pseudo-first Order Plots ...... -79 Figure 3 8. Photolysis of OCN in the Presence and absence of NaCN . Pseudo-first Order Plots ...... 79 Figure 39 . Attempts to Quench Photocyanation of OCN and OCDD b y Ferrocene ...... -80 Figure 40 . A Double Reciprocal Relationship of the Quantum Yield of Disappearance of DCB and the Initial Concentration of DCB ...... 80 Figure 4 1. Photolysis of Hexafluorobenzene in the presence of NaCN at 254 nm ...... -81
vii Figure 42. A Scheme of the Disposai of the Dioxin Laboratos, Wastes ...... 143 Figure 43. Apparatus for Photolysis of Dioxin-containing Liquid Laboratoiy Waste.. .-146 Figure 44. Disappearance of 1,2,3,4-TCDD, 1,3,6,8-TCDD, OCDD,and DCB During the Photolysis at 254 nm ...... ~...... ~...... 148 Figure 45. First Order Plots for disappearance of 1,2,3,4-TCDD, 1,3,6,8-TCDD, OCDD, and DCB During the Photolysis at 254 nrn ...... 148 Figure 46. Photolysis of 1,3,6,8-TCDD in Hexane/toluene Mixtures at 254 nm ...... 150 Figure 47. Conversion of the Components of the Waste Mixture (Standards) in Photolysis at 254 nrn ...... 151 Figure 48. Change in Peak Areas (HRMS) of PCDD (a) and PCDF (b) During the Photofysis of the Waste with Low Concentration of Contaminants at 254 nm ...... ~...~~..~...... 153 Figure 49. Photodegradation of 2,3,7,8-TCDD by Reductive Dechlorination...... 15 5 Figure 50. Formation and Disappearance of Less Chlorinated Dioxins During the Photolysis of OCDD in Hexane at 254 nm ...... 157 Figure 5 1. Proposed Formation of Carbene Intermediates in the Photolysis of PCDD .. 158 Figure 52. HPLC Chromatograms of 3~-containingProducts of Photolysis of 1,6-[3a-2,3,7,8-~~~~ in Hexane at 300 nm ...... 16 1 Figure 53. ES-MS Chromatogram of Products of Photolysis of 2,3,7,8-TCDD ...... 163 Figure 54. A Mechanism of Photolysis of Dibenzo-p-Dioxin ...... 167 Figure 5 5. The Relationship between Percent Specific Binding of 1,6-13JiIJ-2,3 ,7,8-TCDD and the Conversion of 2,3,7,8-TCDD in Photolysis ...... 170 Figure 56. Change in Radionctivity Associated with Estrogen Binding Afinity afier Photolysis of 2,3,7,8-TCDD ...... 172 List of Schemes
Scheme 1. Proposed Mechanism of Photocyanation of HCB ...... 53
~cheme2. Mechanism of Photocyanation of Decachlorobiphenyl ...... -67
Scheme 3. General Mechanism of Reductive Dechlorination by Homolysis ...... 154 List of Tables
Table 1. Observed and Calculated 13cNMR Chernical Shifts of 27 ...... 40 Table 2 . Influence of Concentration of Water on Product Formation in Photolysis of HCB ...... 44 Table 3 . Product Formation in Ground State Reactions of Chlorinated Polycyanobenzenes ...... -46 Table 4 . Infiuence of Concentration of Water on Percent Conversion of HCB at Equal . . Irradiation Times ...... 51 Table 5 : Products of Photocyanation of Penta- and Less-Chlonnated Benzenes ...... -55 Table 6 . Products of Photocyanation of Decachlorobiphenyl ...... 62 Table 7. Measurement of Quantum Yield of Disappearance of Hexachlorobenzene...... -86 Table 8 . Photolysis of HCB in the Presence of Sodium Cyanide for Various Periods of Time ...... 87 Table 9 . Sensitization of HCB Photocyanation by Acetone, Acetophenone and Oxygen Rernoval ...... 88 Table 10 . Quenching of Cyanation of DCB by Ferrocene ...... 89 Table 11. Photolysis of HCB at Various Concentrations of Sodium Cyanide ...... 90 Table 12 . Photocyanation of HCB at Various Concentrations of Water ...... 90 Table 13 . Products of Photocyanation of HCB at Various Concentrations of Water ...... -91 Table 14. Dependence of Quantum Yield of Cyanation of HCB on the Initial Concentration of HCB ...... 93 Table 15. Sumrnary of Crystal Data, Data Collection, Structure Solution and Refinement Details ...... 95 Table 16. Reactions of 2,3,4,5, 6-Pentachlorobenzonitrile with Sodium Cyanide ...... 98 Table 17 . Cornpetitive Photocyanation of HCB, Pentachlorobenzene and 1,2,3,4-Tetrachlorobenzene...... 100 Table 18. Photochemical and Ground State Reactions of 1, 4-Dicyano benzene with NaCN ...... 102 Table 19. Photochemical and Ground State Reactions of 1,2,4, 5-Tetracyanobenzene with NaCN ...... 103 Table 20 . Photolysis of DCB in the Presence of Sodium Cyanide for Various
Table 40. Attempt to Quench Cyanation of OCDD by Ferrocene ...... 12 1 Table 4 1. Photolysis of Hexatluorobenzene in the Presence of NaCN at 254 nm ...... -122 Table 42. Determination of the Yields of Dechlorination Products in Photolysis of 2,3,7,8-TCDD ...... ~...... ~...... 158 Table 43. Measured Radioactivity of the Samples in the Estradiol Assay on the Products of Photolysis of 2,3,7,8-TCDD...... 172 Table 44. The Calibration Curves of Dibenzo-p-dioxin, 2-MCDD, 2,3-DiCDD, 2,3,7- TrCDD, 1,2,3,4-TCDD, 1,3,6,8-TCDD, 2,3,7,8-TCDD, OCDD, and DCB .-178 Table 45. Photoiysis of 1,2,3,4-TCDD, 1,3,6,8-TCDD, OCDD, and DCB in Hexane ... 179 Table 46. Photolysis of 1,3,6,8-TCDD in Toluene at 254 and 300 m...... 18 1 Table 47. Photolysis of DCB in Toluene at 254 nm ...... 182 Table 48. Photolysis of 1,3,6,8-TCDD at 254 nm in Hexane Containing 1, 3 and 10 Percent Toluene ...... 182 Table 49. Photolysis of the Dioxin Waste fiom Wellington Laboratories (Standards) -183 Table 50. Photolysis of the Dioxin Waste Eom Wellington Laboratories Gxtracts) .... 184 Table 5 1. Determination of the Yields of Dechlorination Products in Photolysis of 2,3,7,8-TCDD ...... 185 Table 52. Separation of the Products of the Photolysis of 1,6-['HI-2,3,7,8-~~~~on HPLC ...... 188 Table 53. Measured and Calculated Concentrations of 2,3,7,8-TCDD During Photolysis at 300 nm ...... 190 Table 54. HAP Assay with the Products of Photolysis of 2,3,7,8-TCDD...... 19 1 Table 55. Estradiol Assay with the Products of Photolysis of 2,3,7,8-TCDD ...... 193
xii Glossary of Abbreviations
1,4-DCB 1,4-dicyanobenzene DCB Decachlorobiphenyl DCDD Dichlorodibenzo-pdioxin DD Dibenzo-p-dioxin DPM Disintegrations Per Minute ES-MS Electrospray Mass Spectroscopy GC Gas Chromatograp hy HAP Hydroxy lapatite HCB Hexachlorobenzene HeCDD Hexachlorodibenzo-p-dioxin HMBC Heteronuclear Multiple Bond Correlation HpCDD Heptachlorodibenzo-p-dioxin HPLC High Performance Liquid Chromatography HRMS High Resolution Mass Spectroscopy HSQC Heteronuclear Single Quanta Correlation MCDD Monochlorodibenzo-p-dioxin OCDD Octachlorodibenzo-p-dioxin OCN Octachloronaphthalene PCB Pol ychlorobip heny f PCDD Polychlorinated Dibenzo-p-dioxins PCDF Polychlorinated Dibemfurans PeCDD Pentachlorodibenzo-p-dioxin TCDD Tetrachlorodibenzo-p-dioxin TrCDD Trichlorodibenzo-p-dioxin
... Xlll CHAPTER 1.
STATEMENT OF PURPOSE
At the beginning of the PhD program I considered four different projects related to the photochemistry of chlorinated aromatic compounds: research of the reactions of haiogenated aromatic compounds with electron donors; reactions of chlorinated aromatics with electron acceptors; photonucleophilic substitution reactions of chloroarornatics; the mechanistic investigation of photolysis of polychlorinated dibenzo- p-dioxins.
In the study of the reactions of halogenated arornatic compounds with electron donors we wanted to examine dependence of the leaving group abilities of different types of halogens on the nature of the electron transfer (ground state versus photochernical versus electrochernical). The project was abandoned when we found that a similar study was already published.
We found that electron acceptors such as 1,Cdicyano- and 1,2,4,5- tetracyanobenzene quench fluorescence of 2-chloro- and 2,2'-dichlorobiphenyls by an electron transfer nom a biphenyl to the cyanobenzene. We attempted to use the above phenornenon in order to initiate the photosubstitution of an aromatic chlorine by cyanide ion in the presence of electron acceptors. Photolysis in the presence of NaCN resulted in a slow disappearance of both 2-chlorobiphenyl and 1,4-dicyanobenzene. No cyanation products were detected, but an adduct of 2-chlorobiphenyl and 1,4-dicyanobenzene was isolated. We did not pursue this project since the photoaddition reactions of analogous benzonitrile 1,2,4,~-tetrac~anobenzene~*~had been studied in great detail. During the experirnents with 1,4-dicyanobenzene in one of the control samples we
observed an unusual photoreaction of hexachlorobenzene and sodium cyanide. Our
preliminary studies demonstrated that this reaction was common for various classes of
highly chlorinated aromatic compounds and, what appeared to a successive
photocyanation reaction became my primary research interest. The objectives of this
research were to identiQ the products of the reaction and to study the photochemical and
photophysicai aspects of the reaction mechanism and to evaluate its possible synthetic
potential. The major part of this thesis is devoted to the photochemical cyanation
reactions of chlorinated aromatic compounds.
During my MSc studies on the photolysis of octachlorodibenzo-p-dioxin 1 had
encountered the phenornenon of missing material in dioxin photoreactions. At that tirne we were unable to resolve this problem. 1 had the opportunity to continue this study in
my PhD program when a local company requested our laboratory to develop a method for
on-site destruction of small quantities of dioxin-containing iiquid waste by W. The goals of this project were more cornplex. It was necessary to develop an inexpensive apparatus capable of dioxin removal below the detection limits of the analytical instruments of the company and to test the apparatus on the actual waste samples. We had to identiQ the products of photolysis of dioxins; this was important fkom both scientific and practical points of view. For the prospective application of the destruction method it was necessary to determine whether the products of photolysis retained the type and level of toxicity associated with dioxins. Accounts of this project are presented in Chapter ID of the thesis. Biblioaaphv for Cha~ter1
1. Yarnada, S.; Kimura, Y.; Sakurai, H. J Am. Chem. Soc., 1977,99,667.
2. Pac, C.; Nakanose, A.; Sakurai, H. J: Am. Chem. Soc., 1977,99, 5806. CHAPTER n.
PHOTOCHEMICAL CYANATION OF CHLORINATED AEWMATIC COMPOUNDS
2.1.Introduction
The reactions of chlorinated aromatic compounds with sodium cyanide under UV
light, which are described in this chapter, were discovered by the author in the course of
another study. These multiple cyanations were found to be a general reaction for various
classes or polychlorinated aromatic compounds, such as benzenes, biphenyls,
naphthalenes and dibenzo-p-dioxins. These polycyanation reactions do not yet have any
synthetic value, since the practical applications of the reaction products, namely polycyano-(-polychloro-)-hydroxy-aromatic compounds are quite limited (dthough the produas of photocyanation are properly called -nitriles 1 will often refer to them as cyano- compounds in order to emphasis the origin of the products). Before discussing the results of this study 1 shall review some aspects of photochernical aromatic nucleophilic substitution, the photochernical cyanation of aromatic cornpounds, methods for cyanation of halogenated aromatics in the ground state, and synthesis and applications of polycyano-(-polychloro-)-hydroxy-aromatic compounds. The emphasis of this introduction will be on the photochemical/photoinduced cyanation of aromatic compounds.
2.1.1 ; Photochemical Aromatic Nucleophilic Substitution
Photonucleophilic aromatic substitution has been extensively studied since its discovery in 1956.' The reaction is usually described as S& and was found to occur with a variety of aromatic esters, halogenated aromatics, polycyclic and heterocyclic compounds. The reaction often proceeds to a high degree of conversion accompanied by good yklds of easily isolated produas. The photosubstitution is usually carried out in polar solvents: water, 1-butanovwater, acetonitrile, or tetrahydrofuradwater mifiures. A wide variety of both anionic and neutral nucleophiles was shown to participate in SN&*, such as: OH, OR-, CN-,CNO-, CNS-, Cr, ~03~;CH3COO-, N02; H, HzO, ROY NH3, primary, secondary and tertiary amines and pyridines. The most cornmon leaving groups in photochernical nucleophilic substitution are: OPO~H'-,OSO~~-, OR-, NO<, F, Cl', Br-,
T, H,N2, soZ2-,~03~; and CN." Three mechanisms for the substitution reaction have been identitied: 1) direct displacement, SNZA~';2) photoinduced activation of substrate through an electron transfer fiom aromatic compound to an acceptor, followed by attack of nucleophile on the aromatic radical-cation, SR+NIAT*;3) substitution through electron transfer fkom nucleophile to substrate, SR-^ I d3
In the sN2~r8reaction a Meisenheirner cornplex is formed in the interaction between the excited substrate and the nucleophile, followed by the departure of a substituent from the complex. The SR+NIAT* mechanism is mostly followed for the nucleophilic substitutions of the excited molecules activated by electron donating groups. The initial ionization of excited arenes is possible via ejection of an electron upon absorption of a photon (Equations 1-6) or through the interrnediacy of the excimer (Equations 7- 10)
ArX hv+ 'A~X*+ ?4rx* (1)
3~r~*-t Arp + e, (2)
Arp + e, + ArX (3
~rx'.+ N + A.rXN' (4)
AN+- + x- (5)
Arp + eaq+ ArN (6) 'A~x* {ArpArX"} (8)
+ N + substitution product (9)
Arr-+ reduction product (10)
The most notable feature of photochernical nucleophilic substitution is the reversa1 of the orientation effects of the ring substituents compared with the ground state reaction.
Thus electron-withdrawing groups such as nitro or acetyl activate the meta position towards nucleophilic photosubstitution, while electron donating substituents such as methoxy or amino direct ortho and para nucleophilic atta~k.~
Although the meta orientation of the nitro group clearly holds for cyanide and hydroxide ions as a nucleophiles, when amines are used the regioselectivity depends on the structure of the amine. For example, 2-methoxy-5-nitrophenol (1) is the only product in reaction of 4-nitroveratrole with potassium hydroxide under irradiation fiom a medium
1 II + OH- - 1 II
No, pressure mercury larnpY3but when primary amines or pyridine are used instead of hydroxide the para products (2) predominate.
It was shown that only amines with low ionization potential prefer the orthu/para substitution; in contrast, methylamine and ammonia, which have higher ionization potentials, were shown to attack meta to the nitro group in 4-nitro~eratrole.~*'It is believed that that the meta products arise fiom the s~2A.r'mechanism, while the electron transfer fiom the amine to the aromatic leads to formation of the ortholpma products fkom the aromatic radical anion via the &-NI Ar* pathway. Van Eijk et ai. have shown by laser spectroscopy that the photosubstitution of 4-nitroveratrole with OH occurs only from the lowest reactive triplet state of the aromatic; in the case of amines, the meta substitution with methylamine is a singlet state reaction, but a triplet state reaction with various amines starts with a discrete electron transfer and formation of the radical anion of 4-nitroveratroie. Subsequent attack by the radical-cation of the amine leads to the para produd.
Similady, Bunce et al.? have shown that primary amines cause replacement of the nitro group upon photosubstitution of 1-methoxy-4-nitronaphthalene (3), while secondary amines displace the methoxy substituent in 3. The radical-cation of 3 was observed by flash photolysis; its yield was found to depend on the ease of the amine oxidation, and follows the order tertiary > secondary > primary. It was concluded that for the amine nucleophiles the s~2A.r'process is in competition with the electron transfer process, and occurs when the electron transfer is not energetically favored, e.g. with primary amines as nucleophiles. In the sN& mechanism, when the substitution mefa to the nitro group is not possible, it occurs ipso at the nitro group.7 3
NHR Several theones have been developed to explain the orientation effects of the finctional groups in photochemical nucleophilic substitution. Initially, Comelisse and
Havinga correlated the meta orientation by nitro substituents and the para orientation of methoxy with the charge distribution in the x + x* excited states of nitroanisoles and
4-nitroveratrole.' The positions preferentially attacked correspond to those bearing the least negative charge in the lowest singlet, and sometimes the lowest x + $triplet state
(Figure 1). However this theory failed to explain ortho orientation observed with electron donating substituents and 'reversed' effects of some amine nucleophiles.
Figure 1. Charge Distribution in Ground and Singiet Excited States of4-Nitrocatechol Based on their study of the photo-Smiles remangement reactions (Figure 2) Mutai et
suggested that in several systems the course of nucleophilic substitution could be predicted fiom an examination of the fhntier molecular orbitals of the aromatic substrate.1° According to the FM0 theory, as proposed by Mutai, the regioselectivity of s~2A.r'reaction is govemed by the interaction between the nucleophile "HOW and the orbital corresponding to the aromatic substrate's ground state "HOMO"; the nucleophile attacks the aromatic substrate at the positions where the molecular orbital coeEcient is highest. In reactions proceeding via electron transfer, the interaction between the
"LUMO" of the aromatic substrate and the "HOMO" of the nucleophile controls the regioselectivity; if an electron is transferred from the nucleophile to the substrate the nucleophile radical cation will then attack at the position bearing the highest LUMO coefficient. Mutai's rules were successfully applied to the photosubstitution of meta- alkoxynitrobenzenes, 3,4-diaikoxynitrobenzene and chloronitrobenzenes. 'O
Fi yre 2. Para-Directing Effect of Nitro Group in the Photo-Srniles Rearrangement Mutai's theory was criticized by Cantos et al." based on the substitution of
4-nitroanisole by pnmary amines. AM1 semiempirical calculations confirmed that the fkontier orbi& considerations can justiS. the experimental regioselectivity observed in the SN~A~'mechanism. However, according to the authors, neither the fiontier orbital coefficients nor the net charge can explain the regioselectivity of the electron transfer substitutions.ll Figure 3 shows the net charges of the TI state and radical anion of 4- nitrophenol as wel1 as the H orbital coefficients for the Tl state of this molecule. The largest coefficient of H orbital is on the carbon bearing the nitro group. Substitution of the nitro group indeed occurs with ethyl glucinate as the nucleophile (s~2A.r' mechanism). Analysis of the net charges of the Tl shows that the values for both carbon atoms bearing the substituents are similar, so the regioselectivity of the substitution by ethyl glucinate has little dependence on the net charges. In the radical anion of Cnitrophenol the most negative carbon atom is bonded to the nitro group, but the photoreaction between 4-nitroanisole and n-hexylamine leads to the displacement of the methoxy goup.l1 Cantos et al. suggest that the regioselectivity of nucleop hilic
Net Charges H-Orbital Coefficient:
Figure 3. Net charges and the H orbital coefficients of the TIstate and net charges of the radical anion of 4-nitrophenol substitution under electron trader conditions is controlled by the relative stability of the intermediate a-complexes formed upon collapse of the radical-ion pair, although several other explanations are possible.
2.1.2. Photochemical Cvanation of Aromatic Compounds
The preparation of aromatic nitriles is usually carried out by ground state substitution of aromatic halides, sulphonates and diazonium sdts with inorganic cyanides or by dehydrations of oximes and amides.'* Substitution of aromatic cornpounds with cyanide ion under W light is the most cornmon photochemical method to introduce a cyano group into an aromatic rnolec~le.~
Although high isolated yields and high regioselectivity were demonstrated in several studies on the nucleophilic aromatic substitution, to date photochemical cyanation has had very limited synthetic applications. Photocyanation has played an important role in shaping our understanding of nucleophilic aromatic photosubstitution. It was studied
16.17 extensively for methoxy-substituted benzenes 13.14.15 , aromatic hydrocarbons and some heteroaromatic compounds18.19 . Both direct and electron-acceptor assisted photocyanations are possible. Depending on the nature of substituents and the reaction conditions, cyanation rnay proceed via different mechanisms, including s~2A.r'and
SR+NIAT* .2,15
2.1.2.1.Products of Photochemical Aromatic Cvanation
Photochernical cyanation follows the orientation rules for photochemical nucleophilic substitution described in Section 2.1.1. with the exception that the cyanide ion is a strong enough nucleophile to substitute the aromatic hydrogen. When 3-fluoroanisole (4) is irradiated in the presence of potassium cyanide in t-butanol-water, the major product is 3-fluoro-4-cyanoanisole with only traces of 3-cyanoanisole f~rmed.~'Even though fluoride is the best halogen leaving group in photonucleophilic aromatic substitution reactions, the ortho/para directing effect of the methoxy group of 4 favours hydride displacement. In 1,3-dimethoxybenzene (5) substitution of the hydrogen
trace CN
atom ortho/pma to methoxy is preferred over replacement of the methoxy group in the meta position; this phenornenon has also been observed in 1,3,5-t1irnethox~benzene.~~On the other hand, when hydroxide is used instead of cyanide the only product is 3-
Photocyanation was studied for some heteroaromatic compounds2 1.22 -
Photocyanations in flavin (6) proceed readily and are of great synthetic value for isoalloxazine ~hernistry~~The reaction was found to proceed though the intermediacy of cyanodihydro flavins, the latter being oxidized b y oxygen to cyanoflavins.
Although carbonyl compounds such as benzophenone are themally unreactive towards cyanide under typical cyanation condition^,'^ the displacement of the acetyl group by cyanide in certain acetophenones was found to proceed with modest (ca. 50%) yields under UV lightm2' The reaction was found to follow the orfho/paa direction rules for ailsoxy substituents, e.g. 2-methoxybenzonitrile (7) was formed fiom
2-methoxyacetophenone, but if the acetyl is positioned meta to methoxy, the methoxy group was displaced with formation of3-cyanoacetophenone (8) in 70-80% yield. Unlike many reported photocyanations, the acetyl displacement is inhibited by the presence of air or water. It was suggested that water reduces the nucleophilicity of the cyanide ion thus inhibiting direct ~~anation.~'
CN', hv -CH3CN, N2
OCHB
2.1 2.2. Effects of Solvents
Very polar solvents are required for dissolution of alkali cyanide salts and also for the occurrence of photochemical electron transfer. Several solvent systems give satisfactory results in photocyanation. Havinga et used aqueous t-butyl alcohol and KCN as the cyanide source in reactions of unsubstituted arenes. Pure acetonitrile is often
used for phase-transfer catalyzed aliphatic nitrile synthesis in the ground state. In this
solvent cyanide ion is weakly solvated, therefore its nucleop hilicity is enhanced.
Beugelmans et reported that in ethanol no photocyanation takes place for
unsubstituted aromatic compounds. The authors explained this fact by strong solvation of
CN- by EtOH. Although methanol is a good solvent for dissolution of sodium cyanide,
studies of the photocyanation of phenanthrene and naphthalene have shown that the
yields are lower in methanol and methanol-water cornparecf with DM.-water or
acetonitrile-water mixtures.28The quantum efficiency of photocyanation of biphenyl and
naphthalene was much lower in dry methanol than in aqueous acetonitrile although
addition of water to methanol greatly enhanced the reaction." A similar effect has been
noted in the photocyanation of anisole de ri vat ive^.^' It is unclear why addition of water
enhances photocyanations since MeOH as well as EtOH are also hydroxylic solvents.
Another factor to consider is that water cm produce hydroxide ion in equilibrium with
cyanide salts. Possibly OK is indirectly involved in the reaction (see Section 2.2.1.2).
The quantum yields for photocyanation of haloanisoles have shown a very strong
dependence on water concentration; for 4-fluoroanisole it increases fiom 0.02 to 0.69 on
going fiom 64% to 99??(mole fraction) water. Since the quantum yield of intersystern
crossing of 4-fluoroanisole is 0.75~~nearly every triplet molecule that is formed leads to the substitution product in highly aqueous media.
In photocyanations of unsubstituted aromatics excess water sometimes may Lead to lower product yields. Photocyanation of naphthalene and biphenyl with akali metal cyanides in the presence of 1,4-dicyanobenzene in acetonitnle-water (1 :1 v/v) gave carbonitnle products in low isolated yields (12-25%) and with a poor material balance of
reaction. 30 Similar expenments by Yasuda et in acetonitnle containing 10% water
resulted in an increase of the isolated yieId for cyanonaphthalenes to 44%. The quantum
yield of the cyanation of phenanthrene in the presence of 13-dicyanobenzene increases
with increased concentration of water in the presence of oxygen, but decreases when the
photolysis is done under nitr~~en.~'Attempts to use DMF-water in assisted cyanations of
unsubstituted arornatics have shown that dark cyanation reactions take place to a
considerable e~tent.~~
When photocyanations are carried out in aqueous mixtures, photohydrolysis can, in principle, compete with photocyanation. In practice, this is rarely a problem. With unsubstituted arenes, photohydrolysis does not occur; in other systems hydroxide is a weaker nucleophile than cyanide. For example, the maximum quantum yields of photohydrolysis of Cfluoroanisole and 4-chloroanisole in t-butanol-water at 280 nm are
- 0.5, the same as the maximum quantum yields of photocyanation. This suggests that at sufficiently high nucleophile concentration, both OH and CN trap the same proportion of excited states. However, photohydrolysis is negligible when [CN] > 0.1 mol L-',~' showing that CNcan out-compete the solvent for the substrates' excited states.
Several authors have found that photosubstitution is more efficient in aprotic organic solvents than in protic organic solvents when the reaction takes place in the presence of phase transfer agents, such as 18-crown-6.32 In some instances replacement of
18-crown-6 by polyethylene glycol leads to additional enhancement of the rea~tion.'~
Polyethylene glycol and crown ether both activate the cyanide ion in non-polar aprotic solvents by solvating the counter ion, leaving "naked" CN. 2.1 -2.3. Effects of the Additives and Air on Aromatic Photocyanation
The quantum yields of photocyanation of unsubstituted aromatics in the absence of the electron acceptors are usually low because the leaving group is an unactivated hydrogen atom. The quantum yields of cyanation of naphthalene and biphenyl were found to increase with increasing concentration of 1,4-di~~anobenzene.~~1,4-
Dicy anobenzene also facilitates cyanation of anthracene, pyrene, henanthrene,28 and 9- phenylanthracene.17Other electron acceptors, such as O- and m-dicyanobenzene, methyl- p-cyanobenzoate, 1-cyanonaphthalene, 9-cyanophenantherene, 9,l ~-dic~anoanthracene,~~ di~~anoeth~lene,'~and persulphate iod0 have been successfilIy used to enhance the photoc$mation. The assisted photocyanation is successful for the reaction systerns where the fiee energy changes for the photochernical electron transfer fiom arenes to electron acceptors are negative, in other words, the reduction potential of the acceptor has to be higher then the oxidation potential of the aromatic donor for the reaction to occur. Thus,
1,Cdicyanobenzene (Ein = -1.64~~~is a more efficient acceptor than 1- cyanonaphthalene (El12 = -1.56~)'~ in the photocyanation of anthra~ene*~and naphthalene.35 Since the products of the cyanation of unsubstituted aromatic cornpounds are usually good electron acceptors themselves, autocatalytic effects are often observed.
For example, photocyanation of phenanthrene in the absence of electron acceptors gradually accelerated with an increase in con~ersion.~~Addition of 9-cyanophenanthrene to the reaction mixture increased the rate of reaction, and a kinetic analysis of the dependence of the observed reaction rate on the concentration of 9-phenanthrene showed that the latter product assists the photocyanation.36 The rates of disappearance of naphthalene and biphenyl in photoreactions with methanolic sodium cyanide increased
with progress of reaction, also due to the promotion of the reaction by products.30
The presence of an oxidizing agent is required when hydrogen is replaced by
cyanide in arenes. In the absence of oxidizing agents the hydrocyanation products prevail,
but their yields significantiy diminish in favour of cyanation products when oxidants are
present. The photocyanation of phenanthrene under Nz in aqueous acetonitrile in the
presence of 1,4-dicyanobenzene gave 9,lO-dihydrophenanthrene-9-carbonitrile(9) and
9-cyanophenanthrene (10) in 55 and 16% yields re~~ectivel~.~~Under aerobic conditions,
10 was the only product, isolated in 78% yield. In dry acetonitrile, in the absence of
oxygen, formation of 10 was cornpletely suppressed. l6
9 IO
One of the characteristic features of nucleophilic aromatic photosubstitution is the
so called a-effectS2This was introduced as a general term by Cornelisse and ~avin~a;'it
refers to preference for substitution on position 1 in naphthalene and amlene, positions 2
and 4 in biphenyl, and position 9 in phenanthrene. Under aerobic conditions in the
presence of CN; 1-methoxynaphthalene (11) gives photosubstitution of an a-hydrogen at
C-4 @ara) and not at C-2 (B, ortho), and 2-methoxynaphthalene (12) shows photosubstitution at the a-hydrogen on position C-1. In 2,3-dimethoqmaphtha1ene (13) the hydrogen atom at C-l (a,orthohneta) is replaced by cyanide ion instead of the beîter methoxy leaving group at the P position. However, under N2,methoxy substitution
17 appears to be the main reaction path for these compounds and the hydrogen displacement becomes in~i~nificant.~~
11 CN Aero bic: 1% 99%
Anaerobic: 23% 11%
-CN' ~-BuOH-H~O moCH3OC H3 0CH3 OC H3 13 Aerobic: 71 % - Anaero bic: 9% 19%
These observations rnay be explained by kinetic analysis of a possible reaction scheme, show in Figure 4. Since the para substitution products prevail in the presence of oxygen, kz is larger than kl.However, in the absence of oxidant the intermediate lla almost always reverses to the starting material (k4and 1-cyanonaphthalene is formed via ipso substitution. Formation of 11% of 1-methoxy-4-cyanonaphthalene from 11 is probably due to incomplete oxygen removal. + CN' OMe OMe I zq I H CN CN Ila
Figure 4. Pathways of Photocyanation of 1-Methoquaphthalene
It should be noted that the presence of OÎ is beneficial for photocyanation only at low oxygen concentrations. Molecular oxygen hinders formation of dihydrocyanated products by oxidation of intermediate O-cornplex, similar to lla to the cyanoaromatics and the peroxide anion, HO^.'^ Flash photolysis studies by Lemmetyinen et al. 15.3 1,3226 have shown that photochernical cyanation of several aromatic compounds involves formation of two photoinduced transients besides the excited state, both of which are quenched by oxygen. Interaction of oxygen with the second transient, which is believed to be a 0-cornplex, leads to products, while the oxygen quenching of the initially formed intermediate appears unproductive. Given that the rate of the latter process is faster, increasing the concentration of 02will decrease the rate of reaction. Recently, the role of oxygen in the mechanism of assisted photocyanation of unsubstituted aromat ics was re~ised?~the discussion of this topic follows in Section 2.1.2.5.
Although many reagents may act as oxidizers not al1 combinations lead to substitution when CN- is present. For example, both H20t and di-t-butyl peroxide quench naphthalene fluorescence; but in the presence of CN, hydrogen peroxide causes rapid destruction of naphthalene and di-t-butyl peroxide retards the photoreaction of
naphthalene with CW." Nitrous oxide and (NH&S208 were found to be efficient
oxidizing agents in photocyanation of anisole and its meta-substituted derivatives, and
also of benzene.20When photocyanation is carried out in aqueous media, m)&08 is
preferred over oxygen because of its higher solubility. Nitrous oxide was found to behave
similarly to oxygen; a maximum rate of cyanation was obtained at low NzO
concentrations and the rate decreased with increasing N2O concentration^.^^
2.1.2.4. Mechanisms of Unassisted Photocyanations
Depending on the presence or absence of substituents on arornatic ring and electron
acceptors, photocyanation may involve severai different reaction rnechanisms.
Photocyanations of 3- and 4-nitroanis~le~~'~~4-11itroveratrole~~ and 1-methoxy-4-
nitr~na~hthalefie~~are believed to proceed via the SNZA~'mechanism. Although the
SN~A~'mechanism was suggested for the unassisted photochemical cyanations of halo an isole^,^^ sirnilarities of the yields of different products in the photochemical and anodic substit~tions~~have suggested the involvement of a more complex SR+NAT* reaction mechanism.
In aromatic systems carrying electron donating groups, such as rnethoxy, the initial reaction step is probably photoionization to a radical cation. These reactions were classified as SR+NA~'.' Formation of a solvated electron was observed in laser spectroscopic studies of the photocyanation of Cchloroanisole and 4-fluoroanisole. l5 Den
Heijer et dzosuggested the following mechanism for nucleophilic substitution of 4- haloanisoles (Figure 5). Ionization of the triplet anisole molecule lads to the formation of the radical cation of anisole and a solvated electron. The attack on the radical cation by cyanide ion is followed by departure of the leaving group; back electron transfer eom solution gives the substitution product.
Figure 5. Mechanism of Nucleophiiic Substitution of Eaioanisoles in Aqueous t-33~0~~B"d""L
eiined. Lernmetyinen et al." have found that in acetonitrile-water mixtures formation of the radical-cation of 4-chloroanisole is linearly proportional to its ground state concentration; in other words, the interaction between the excited and the ground state molecule produces the ion pair which dissociates into radical cation and radical anion of
4-chloroanisole (Figure 6). Attack of the nucleophile on radical cation yields neutral radicals of type ArXCN, which were observed by laser spectroscopy. The authors'' argue that the next step is the depamire of the halogen atom rather than the halide ion, since no radical cations of the type ArCw were observed by flash photolysis and the measured
Iifetime of solvated electron is too short to allow for the back electron transfer. The released halogen atom X yields the halogen anion upon recornbination with the negative species formed in the reaction of water and radical anion of 4-fluoroanisole. It should be noted that at low substrate concentrations the formation of the excimer is unlikely and the mechanism shown in Figure 5 should prevail. - 0CH3 OCH I 1 --hv isc a
Figure 6. Mechanism of Photocyanation of Halonaisoles Involving the Radical Ion Pairs
Consistent with the previously discussed mechanisms, in reactions where substituents like halogens and methoxy are displaced by cyanide, oxidizing agents like oxyge- NzO or (Nl&)2S20s do not influence the eficiency of phoioreaction.20
Photosubstitution in the absence of the electron acceptors proceeds via the triplet excited state for ani~ole,~~halo aniline^,^' 4-fl~oroanisole~~~~~and 4-chloroanisoIe.1"20 The triplet energies of these substrates are around 335 l~.J/rnol,~~and the photoreactions were sensitized by acetone (ET = 343 mol)^^ and quenched by tram- 1,3-pentadiene
(Er= 247 kJ/rn01).~~The order of reactivity in photosubstitution of halogen by cyanide ion (F = Cl > Br > 1) is similar to that of a ground state aromatic nuclephilic substitution.
ut hors^' argue that this order of reactivity is determined by the shorter lifetime of the triplet molecule as a consequence of increased intersystem crossing to the ground state; however it may also depend on the relative electronegativity of the fdogens, as in the
ground state reaction.
Dependence of the quantum yield on the substrate concentration has been noticed in
photocyanations of several unsubstituted aromatic compounds. Bunce et aL2' have found that the quantum yield of cyanation of naphthalene or biphenyl-in the absence of electron
acceptors depends on the concentration of the starting material. Triplet sensitizers failed to increase the quantum yield and quenching experiments were inconclusive, so the
authors suggested that the photochernical cyanation of naphthalene and biphenyl proceeds via formation of the singlet excimer of the arene which then dissociates to the radical ion pair. Formation of a dihydrocyano product in the unassisted photocyanation was rationalized by disproportionation of a pair of cyanohydroaryl radicals (Equation 12) formed by attack of cyanide on the radical cation of the aromatics (Equation 11).
Cid&' + CN+ C1oHsCN. (1 1)
2Cl&CN- -+ CioH7CN + CloHgCN (12)
This mechanism allows no exchange of hydrogen with the solvent; indeed, when photocyanation was carried out in CH3CN-D20there was no incorporation of deuteriurn into the products.29 On the basis of analysis of Stem-Volmer quenching for the reaction between the excited singlet state of nap hthalene and electron acceptor ~ernrnet~inen~~ suggested that the unassisted photocyanation of naphthaiene is a triplet state process, however, no direct expenmental evidence of the tnpiet reaction was submitted. Vink et al. observed that photocyanation of biphenyl could be sensitized by benzophenone and proposed a triplet state rnechani~rn.~~ The photocyanation of phenanthrene was studied in great detail.31, 32, 43, 44 It takes place via a triplet excited state in the absence of electron acceptors. This was demonstrated by sensitization and flash photolysis met hods.) ' AIthough it was suggested that the radical cation of phenanthrene is formed by electron transfer to a cyanide ion32 the observations that CN- does not directly quench either the singlet excited state of phenanthrene or its triplet stateW suggests that the radical cation of phenanthrene is formed either through the excimer or by electron photoejection. Flash photolysis experiments have shown that a new transient is formed at high concentrations of phenanthrene and its yield increases with the increased concentration of the ~ubstrate.~'
This observation suggests the involvement of the excimer of phenanthrene, sirnilar to that of naphthalene, in the photoreaction.
2.1 -2.5. Mechanism of Photocvanations in the Presence of Electron Acceptors
Although there is still some uncertainty on the rnultiplicity of the unassisted photocyanation, a vast amount of experimental data shows that the reaction is a singlet state process when the electron donors such as 1,4-dicyanobenzene are present. p-
Dicyanobenzene strongly quenched the fluorescence of naphthalene and biphenyl; kinetic experiments with naphthalene have shown that the reaction is first order in CN and p- dicyanobenzene, but not dependent on the concentration of naphthalene under conditions when al1 incident light was absorbed by naphthalene.30Suzuki et al. reported that in the presence of 1,Cdicyanobenzene photocyanation of anisole proceeds via a singlet excited
tat te.^^ A singlet state reaction in the presence of electron acceptors has also been suggested for anthra~ene,~~methoxy-substituted naphthalenes,28phenanthrene, ZS,~ 132 and
9-phenylanthracene.17A generalized mechanism for photocyanation in the presence of 1,Cdicyanobenzene is shown in Figure 7. Electron transfer fiom the singlet state aromatics to the electron acceptor leads to formation of radical ion pair. Cyanide ion adds to the aromatic radical cation forming ArH-CN, which then may disproportionate, resulting in dihydrocyanation and cyanation produas, transfer the electron back to the radical anion of 1,4-dicyanobenzene and add a proton, or fom the cyanation product by oxidation with oxygen or other oxidizing agent. The first two reactions are important under anaerobic conditions; the latter pathway becomes predominant in the presence of oxidizing agents fike Oz or NtO.
ArH 'A&
'A~H'+ DCB + ArH'. + DCB'
A? + CN+ ArH-CN
2 ArH-CN -+ H-ArH-CN + ArCN
AïH-CN + DCB" + ArH-CW + DCB
ArIH-CN' + H' + H-AïH-CN
ArH-CN + Ox + ArCN + OxH'
Figure 7. Mechanism of Photocyanation of Aromatic Compounds in the Presence of Electron Acceptors @CB is 1,4-dicyanobenzene, Ox is oxidaat)
The experimental data of Yasuda et show that recovery of the electron acceptors after reaction is relatively low: between 9 and 71% for 1,Cdicyanobenzene and between 50 and 70% for 9-cyanophenanthrene. A possible explanation for the consurnption of the electron acceptors during photocyanation has been given in the study of photochernical cyanation and hydroxylation of naphthalene by Niranen et al. l6 The authors suggest that interaction between the neutrd radical A.Hand the DCB"radica1
25 anion results in proton transfer and formation of hydroxynaphthalene radical anion
(Equation 13). Because the oxygen is needed for the reaction product, the final step is the oxygenation reaction (Equation 14).16 Electrochemical studies have shown that proton capture and dimerization reactions are typical for the decay of the radical anion of 1,4- dicyanobenzene.45.46
CloHsOH' + DCB' + CloH70W + DCBH (13)
C loH70IF' + 02+ C loH7OH + 0; (14)
By-analogy, in photocyanations 1,Cdicyanobenzene may be consumed through the proton transfer fiom ArH-CN (Equation 15)
ArH-CN + DCB" + ArCN' + DCBH (15)
Recent studies of deuterium isotope effects on the photocyanation of naphthalene by
Cornelisse's group35 gave some new insights on the rnechanism of assisted photocyanation. Irradiation of naphthalene in non-degassed acetonitrile in the presence of various electron acceptors resulted in the formation of 1- and 2-cyanonaphthalene, in the ratio of 7 tol. The ratio is in line with calcuiated charge distribution of naphthalene radical cation; the calculated charge at position 1 is +0.09,while at position 2 is -0.05.~~
Attack at position 1 is also favored since the resulting ArHCN radical has a lower energy compared to 2-cyano-2-H-naphthyl radical (AHr =393 versus 408 kJ/rnol) due to more extensive resonance. Although the quantum yield of disappearance of naphthalene was dependent on the electron acceptor (and reached the maximum for 1,4-dicyanobenzene), the ratio of 1- and 2-cyanonaphthalenes was constant. Photocyanation did not proceed in the absence of 1,Cdicyanobenzene. According to Cornelisse et al, when used in catalytic amount, 1,Cdicyanobenzene
was not consumed dunng the photocyanation of naphthalene. The authors3' suggested that 1,4-dicyanobenzene is the initial electron accepter, and that oxygen picks up an
electron from DCB radical anion, yielding the oxygen radical anion O$-(Figure 8). At low concentrations of oxygen in solution a primary isotope effect with perdeuteronaphthalene as a substrate was observed (TE changed from 1-6 to 2.5). Under these conditions the hydrogen abstraction step is rate limiting. Therefore oxygen or
CN- 1 ciiffusion
DCB'+ O2 - DCB + 02
Figure 8. Mechanism for the Photocyanation of Naphthalene in the Presence of 1,4-Di~~anobenzene~~ oxygen radical anion, but not l,Cdicyanobenzene, act as hydrogen atom abstractors. The calculated barrier for hydrogen atom transfer tiom 1-cyano- 1-H-naphthyl radical to oxygen radical anion is less than 8.4 kl/mol, while that to oxygen is 117 k~/rnol.~~This amounts to a relative rate of 10" at equal concentrations of oxygen and oxygen radical anion. The transfer of the hydrogen atom from ArHCN will occur almost exclusively to oxygen radical anion, despite is low concentration relative to oxygen in solution.
This short review demonstrates that a variety of reaction mechanisms may be involved in photocyanation, depending on the substrate and the reaction conditions.
Although photocyanation has been studied extensively not al1 the aspects of the reaction are fully understood, especially for the reactions involving the electron transfer.
2.1.3. Methods for Cyanation of Haloeenated Aromatics in the Grotind State
Conversion of aromatic or heterocyclic halides into the corresponding nitriles is called cyanation but has also been referred to as nitrilation and cyanodehalogenation. The classical procedure involves treatment of aryl halide with copper (I)cyanide at high temperature in an appropriate solvent and is known as the Rosenmund-von Braun reaction.12 Aromatic iodides, bromides, chlorides and fluorides can be converted to nitriles, with iodides being the most reactive. Although the reaction mechanism is different, the difference in the reactivity of halogens towards nucleophiles follows the order typical for nucleophilic aromatic substitution4'. The difference in reactivity is sufficient to allow preferentiat cyanation of the iodide in the presence of chloride; for example, in the synthesis of 2,4,6-trichlorobenzonitrile fkom 1-iodo-2,4,6- tnchloroben~ene.~~The palladium catalyzed cyanation of 3-chloroiodobenzene with potassium cyanide in hexamethyl phosphoric tnamide (HMPT) resulted in a mixture of
70% of 3-chlorobenzonitrile and 21% of 1,3-di~~anobenzene.~~A similar selectivity was demonstrated for trimethylsilyl cyanide toward 4-bromochl~robenzene~~and for CuCN cyanation of pentafluorobromobenzene.51Exarnples of cyano substitution on fluorine are rare, although pentafluoropyridine gives 273,576-tetrafluoropyridine-4-carbonitnlein 33% yield on reaction with copper cyanide at O'C in DMF.~~
Substrates with more than one halogen, such as polyiodides48,polybromides 5334 and polychlorides", usually react to give polycarbonitriles; sometimes the polycyanation is incompleteS5 or accompanied by a loss of a halogen? For example, 1,3,5-tribrorno-
2,4,6,-trïrnethylbenzene yields the corresponding tricarbonitrile in high yield54.Treatrnent of polymeric bromotriphenyls (14, R = Br) with copperO cyanide in quinoline at 237'~ results i.n cyano-polymers (R = CN) of high thermal stabilitye5'
Certain substituents may have pronounced effects on the progress of cyanation.
Electron withdrawing substituents such as nitro and carboxy groups in the 2- and 4- positions of aromatic ring increase the reaction rate.58 Some selectivity in cyanation of polyhalogenated aromatics is also possible. For example,
2,3-dichloronitrobemene is converted into 2-chloro-6- L - nitrobenzonitrile (15)'~ and 7-bromo-5-nitrobem-2- thia-1,3-diazole-4-carbonitrile (16) is the only product of cyanation of 4,7-dibromo-5-nitrobenzo-2- thia- 1,3-dia~ole.~~ Ary 1 halides carrying the following substituents have been converted to the
corresponding nitriles: w1, alkyloxy, alqlthio, hydroxy, acyl, fomyl, carboxyl, carboxyl ester, nitro, primary amino and tertiary am in^.^' The primary arnino substituent was shown
to exert an unusual effect on the reaction; thus 4iodoaniline failed to react with potassium
cyanide-palladium (II) acetate in HMPT," while Cbromoaniline gave a good yield of
nitrile under similar ~onditions.'~
The mechanism of cyanation by copper (I)cyanide was extensively studied by several
researchers6' though it is not completely understood. Kinetic studies of homogeneous
cyanation of 1-ha~o~enona~hthalenes~~and iodobenzenesS8 showed that the second order
rate constant increased as the rdonprogresse& suggesting autocatalytic effect by the
products, aryl nitriles (a sirniiar effect in photocyanations was described previously). For
example, addition of traces of product accelerated the cyanation of Zbromobenzoic acid!'
Nucleophilic substitution of aryl halides is not affecteci by radical traps; addition of excess
CW as KCN or halide ion (as potassium salt) decreases the rate of cyanation, and addition
of a large excess of copper (I)cyanide makes the reaction pseudo-est order in aryl
halideeg ~heseobservations eliminate the radical or dissociative rnechanisms for
cyanation. The mechanism of cyanation proposed by Couture and paineS8 involves the formation of a x-complexed organocuprate f?om copper (I)cyanide and haloarene.
Intramolecular attack of the CNion to give a tetrahedral intermediate is followed by a rate determining loss of halide and rapid formation of aryl nitrile (Figure 9).
Figure 9. Mechanism of Aromatic Cyanation by CuCN in Ground State This mechanism explains the retarding effects of potassium cyanide and
potassium iodide and is supported by similar SN reactions with stable transition rnetal
~orn~~exes.~~
During the past two decades the cyanation of aryl halides in the presence of
transition metals other than copper has been investigated. The metals that have been
successfilly used are ni~ke1,~~*~~*~~~~~palladium 49.68,69 and cobak7* The advantage of
using nickel tris(tnpheny1 phosphine) complexes and sodium cyanide are low
temperatures, in the range of 50-60~~,and high yields of cyanation." However, this
method requires anaerobic conditions and the cost of reagents is high. Cyanation in the
presence of nickel complexes involves two steps: an oxidative reaction between aryl
halide and tris(tripheny1phosphine)nickel followed by displacement of the halide
complex (17, Figure 10) by the cyanide ion. This method has some other limitations:
halogenated aromatics with nitro substituents cannot be cyanated because of the
interaction between the nitro group and the catalyst, and cyano or chloro substituents in
ortho positions significantly reduce the rate of cyanation. Addition of phase-transfer
catalysts such as 18-crown-6 ether7l or replacement of sodium cyanide by acetone
cyanohydrina allows for cyanation under very mild conditions. For example,
17 L Figure 10. Mechanism of Cyanation of Aryl Halides by trisflripheny1phosphine)Nickel and NaCN bromobenzene is cyanated in 98% yield in acetone under nitrogen at 50'~ for 40 min in the presence of acetone cymohydrin and tram-chioro(1-naphthy1)-bis(tripheny1- phosphine)nickel-triphenylphosphinetnethy- amine ~orn~lex.~'
High yields of nitriles fiom aryl iodides are obtained at temperatures of 100-
120'~ when cyanation is catalyzed by palladium (II) salts, especially palladium acetate in the presence of HMPT.Potassium hydroxide or sodium ethoxide, added in small amounts as co-catalysts, enable the reaction to proceed at even lower ternperatures.49 Aryl brornides are also reactive under these conditions; for exarnple, bromobenzene, palladium
@) acetate, potassium cyanide, triphenylphosphine, and calcium hydroxide in DMF at
100'~ for 20 min gave 93% yield of ben~onitrile.~~Experimental findings showed that zero-valent palladium is necessary for cyanation to proceed. The CN ion is trapped by palladium (II)acetate to form a salt [P~~'(cN-)L] (where L is a ligand such acetate ion or solvent) which does not initiate cyanation. Palladium acetate is reduced to ~d'by bases; pd0 reacts with halobenzene to give an oxidative-addition cornplex, which then reacts with cyanide to fom the aryl nitrile (Figure 1l)." Therefore, an excess of palladium (II) acetate over KCN is necessary for the progress of reaction; ~d'activates the halobenzene and pd2' promotes the dissolution of potassium cyanide.
1 base 1 I KCN + L I P~(OAC)~---- P~*+(cN-)L
Figure 11. Mechanism of Palladium Cornplex Catalyzed Cyanation of Iodobenzene Recent modifications73,74 of cyanation overcome this limitation. When zinc
cyanide is used instead of potassium cyanide, ody 4 mol% of the catalyst is required because the covalent nature of zinc - cyanide bonds limits the dissociation of ZnCN.
Cyanation of various aryl bromides under these conditions gives 90-95% yields under mild reaction conditions." The other approach to keep the concentration of CN- low and the yields high is to use phase transfer catalysts such as the crowned phosphine cornplex (18) in combination with PdClt 18 and Zn or the water soluble catalysts of the type PdClzpPh2(m-C&OsNa)]2 with zinc chloride or sodium borohydride and sodium anid ide.^^
When trimethylsilylcyanide in the presence of (PPh3)$d is used instead of KCN the substituent effeas on cyanation are negligible; good yields were obtained in cyanation of bromoaryls whether electron-withdrawing or electron-donating groups were present.50
Cobalt catalyzed cyanations are less well studied. Cyanation by cobalt (II) chloride with potassium cyanide involves formation of tetracyanocobaltate ion
[co(cN)~]~', which forrns the complex K~+[A~CO(CN)~~~-.Displacement of the halogen in this cornplex by CW gives pentacyanocobaltate, which breaks down to benzonitrile and regenerated [Co(CN)4] 3+. 70
Cyanation of sorne aryl haiides by potassium cyanide may be induced electrochernically using aqueous potassium brornide as electrolyte and a hanging mercury drop cathode.75 Tt was show that the electrochemical cyanation of 4- bromobenzophenone at -38'~ proceeds via the formation of radical anions:
[4-PhCOC&&r]'- + 4-PhCOC& + Bi -t [4-PhCOC&&CN]'-
33 Similar mechanistic findings were reported by Yoshida for regiocontrolled anodic
cyanations of pyrroles and in do le^.^^
Specific structurai features of some substrates allow the displacement of halogen by cyano group in the absence of CuCN or any other catalyst. In the preparation of 1,4-
dip henyl- 1,2,3 -triazole-5-carbonitrile (19) relatively unreactive chlorine undergoes
substitution in the absence of copper (I)cyanide due to the effect of a pair of pyridine- type nitrogens in a 1,2,3-tnazole ring.77
A sirnilar effect was observed dunng conversion of 2-fluoropyrazine into the corresponding nitrile (20) .78 Activation b y cyano groups permits cyanation of 2,4,6- trifluorobenzene-1,3,5-tricarbonitrile(21) under remarkably mild condition^.^^ Tetramethylarnmonium cyanide, when reacted with structurally activated halides
at ambient or slightly higher temperature, gives good yields of nitriles; for example, the
formation of 3-ethoxyquinoxaline-2-carbonitrile (22)."
+- M-SO, 50' Me4N CN 70% uxc'+- 03'" OEt 22 OEt
2.1 -4. Svnthesis and Applications of Polvcvano-(-Polvchloro-1-Hvdroq-Aromatic
Compounds
Since the hydroxy group deactivates an aromatic ring in nucleophilic aromatic
substitution, synthesis of polycyano@olychloro) aromatics by this route is unpractical,
although 4-cyanophenol rnay be obtained from Ciodophenol under harsh condition^.^^
Polycyanophenols are not easil y accessible. 2,6-Dicyanophenols and their
halogenated derivatives were prepared from the corresponding diformyl compounds by
aeration with sodium hydroxide and urea in the presence of copper catalyd2A senes of highly fûnctionalized cyanophenols was obtained from thermolysis of 4-allcynyl-3-azido-
1,2-benzoquinones (23). 83,84 Pentacyanophenol was prepared by Friedrich in a multistep synthesis starting from 1,3,5-tnmethyl benzene (24). 85-86 Me Me CHCb CHOI 12,ccl~~ciC12,h~Gci ___jC"Gci - 205~~\ '.- \ '. Me Me Me Me CbHC CHCb OHC CHO
t
NC CN NC CHNOH ICN- CI
Chi OC2H5 OH NC@CN + NC@CN -EtoH, 78'~ NC@CN \ \ \ NC CN NC CN NC CN CN CN CN 1 Et*-. -20°c. 30 min t
To date practical uses for these cornpounds are limited. p-Cyanophenol is an
intermediate for herbicides and insecticides.*' Several compounds such as sulphonates
and carbonates of 2,6-diiodo(dich1oro)-4-cyanophenols, 87,88,89 2,6-dichloro-4-
thiocyanophenol and their derivatives were found to exhibit herbicidal, fungicidal and
insecticidal properties. Diesters similar to 25, 90.9 1 and laterally fluorinated Ccyanophenyl
and Ccyanobi p heny 1 benzoatesg2have large negative dielectric ani sotropy and may be
useful as nematic liquid crystals. Since cyanophenols are unusually acidic, they are often used as mode1 substances in the field of physical chemistry. Cyanophenols were subjects of the study on the influence of photoexcitation on a~idit~.'~Reactivity of a nitrile function towards benzonitrile oxide cycloadditions was found to be enhanced by hydrogen bonding in oriho-cyanophenols.g4
This short review on photochernical and ground state cyanations of aromatic compounds is followed by the discussion of our findings in this field. 2.2. Results and Discussion
2.2.1. Photolvsis of Hexachlorobenzene in AcetonitriIe in the Presence of Sodium
Cvanide
Most of the findings discussed in this Section have already been published.95
2.2.1.1. Product Identification
Irradiation of HCB at 300 nm in aqueous acetonitrile in the presence of excess
NaCN resulted in rapid disappearance of the title compound and the formation of a highly
polar (TLC) yellow-green fluorescent product mixture. The greenish color was first
observed within 5 minutes of irradiation and aRer 50 minutes a 90% conversion of HCB
(Co= 7.5~10~mol L-') was reached. The crude product showed IR stretching bands at v
= 2203 and 2176 cm-' characteristic for the cyano group, but 'H NMR showed no signals
apart from residual solvent. The products could not be analyzed by GC or GC-MS, but
were successfully separated and analyzed by negative ion electrospray mass spectroscopy
(ES-MS). Three tentative parent ions (M-H)- were found at dz218 (no chlorine), 227
(one chlorine) and 236 (two chlonnes). MSMS experiments on each of the tentative
parent ions showed loss of Cl for dz227 and consecutive loss of two chlorines for m/r
236; loss of CO (M-H-28), characteristic for phenolic compoundsg6,was observed for al1
three parent ions. The possibility that one or more acetonitrile moieties might be present
in the products was eliminated by showing that the product composition was unchanged when propionitrile replaced acetonitrile in the solvent. Furthemore, photolysis of HCB
in the presence of KC'% led to the mass spectral peaks at m/z 218, 227, and 236 being displaced to dz223, 23 1 and 239, corresponding to the incorporation of 5, 4, and 3 CN groups respectively. The foregoing evidence suggested that the three products were pentacyanophenol, a tetracyanochlorophenol and a tricyanodichlorophenol respectively.
Thus compound 26, dz218, was pentacyanophenol. The substance having dz227 was isolated, and after crystallization with much dificulty, was identified by X-ray dimadon as the sodium salt of 4-chloro-2,3,5,6-tetracyanophenol, 27. Cornpound 28, having m/r
236, was isolated, but could not be crystallized.
The structure and composition of 27 was established by X-ray analysis, carried out by Dr. G. ~er~uson,~'as the 4-chloro-2,3,5,6-tetracyano isomer. During the preparation of the crystals, the phenol was present as the corresponding phenolate although this was not known at the time. The crystal asymmetric unit was shown to contain three phenolate anions, one in a general position and two lying on crystallographic two-fold axes; a view of one of the phenolate ions which lies on a two- fold axis is in Figure 12. Also present in the crystal asymmetric unit were three sodium cations (one in a general position and two on two-fold axes), four water molecules (one on a two-fold axis) and two propanol molecules (each present with 0.5 occupancy and disordered). The crystal lattice was made up f?om three sheet-like Iayers of ions and molecules with the sheets stacked dong the long b-direction of the unit cell. The sheets at y = 0, 0.5, 1.0 had disordered propanols of solvation and the sheets at y = 0.25, 0.75 had the water molecules and the sodium cation in a general position. In between these two sheets the sodium cations on two-fold axes and the phenolate anions were uniformly 39 stacked with their phenolate ends directed towards the sodiurdwater sheets and the
4-chloro ends directed to the propanol layers. A view of part of the layer structure is shown in Figure 13.
Molecular dimensions are unexceptional and serve to establish the structure; additionai details can be found in Table 15 (Section 2.4.5.9). The sodium cations (Nal,
Na2) on the two-fold axes each form four Na ...N bonds to two phenolate anions and coordination around these sodiums is completed by propanol O atoms (to Nal) and by water O atoms (to Na2); dimensions are:- Nal ...N 2.491(7) - 2.578(8), Nal ...O 2.53(1) -
2.62(1)-A; Na2 ...N 2.391(6) - 2.704(6) NaZ.0 2.372(6) A. The sodium (Na9 in the general position 3 is coordinated to one N atom ffrom the phenolate in a general position
(Na3 ...N 2.520(7) A) and to five O atoms (fiom phenolate and water molecules) (Na3. ..O
2.356(5) to 2.716(9) A).
Analysis of the I3c NMR spectnim of 27 revealed six peaks (Section 2.4.5.9).
Table 1 shows the tentative assignments of the observed and calculatedg6chernical shifis of 27. There is a reasonable agreement between the observed and calculated signals on the positions 1, 2/6, and 3/5 of the ring. Carbon 1, ipso to 0- displays a downfield signal at 171.46 ppm. A similar feature was observed in 13cNMR spectra of 28 (Section
Table 1. Observed and Calculated 'J~NMR Chernical Shifts of 27
1 4 Position 2,6 3,s 7, 10 8, 10 174.5 113.8 123.1 129.7 118.7 118.7 Calculateci 6, ppm NC 9 Observed 6, ppm 171.5 114.0 122.5 117.8 113.5 115.8 N5@i:CI CL1 @
Figure lZAn ORTEP Diagram of a Phenolate Ion of 27
Figure 13. A View of Part of the Layer Structure of Crystal Asymmetric Unit of 27 Two molecules of propanols of solvation are shown as C64-C63452-061 and C54-C53-C52-051. 2.4.5.9). This value is close to chemicai shifts typical of carbonyl carbons; the phenornenon is probably caused by the resonance between the phenoxide and the carbonyl cyclohexadiene forms of the cyano phenol (Figure 14). The nitriles at positions 7 and 8 can not be assigned due to the lack of theoretical data on the chemical shifts of cyano groups on multi-substituted aromatic rings. The observed signal for carbon 4, pma to the phenoxide oxygen, is more upfield than the calculated value. The correction values in the tables of substituent effects on chernical shifi may not be unambiguously used to calculate shifts on çuch highly functionalized moleailes; however, the shift at C-4 may directly depend on the amount of negative charge put into the aromatic ring by the phenoxide ion.
Figure 14. IResonance Structures of the Phenoxide Ion of 27
Nine signals were found in the 13cNMR spectrum of the dichloro-tricyanophenol
28. Attempts to establish its structure by correlating the calculated and the observed chemical shifts failed because of large deviation of the values. From the analysis of the spectnim the only conclusion about the structure of 28 that can be drawn is that the compound is unsymmetrical since nine rather than six carbon signals were found. 2.2.1 -2.Reaction Mechanism
It is reasonable to assume that the products are formed by successive replacement of -Cl by -CN. We explored the possibility that the photochemical reaction is limited to formation of pentachlorobenzonitrile followed by successive thermal cyanation O- and p- to the cyano group(s). Dark controls with pentachlorobenzonitrile showed that this was not the case (Figure 15); the reaction involves successive photochemical steps. The absence of an isosbestic point in the UV absorption spectra of the photoreaction in progress shows that more than one intermediate is involved (Figure 16). Only traces of a trichloro-dicyanophenol (m/r 245) were found in ES-MS spectra of the HCB photocyanation produas. These observations suggest that when three or more cyano groups have been introduced into the product, hydrolysis can compete with further substitution by CN i.e., replacement of -Cl by -OH.Alternatively, the introduction of the phenolic group may involve the replacement of -CN by -OHafter at least four cyano
0 Dark Cyanation X Photocyanation
O 50 100 150 Time, min
Figure 15. Dark vs. Photochernical Cyanation of Pentachlorobenzonitrile 500 DN.)
Figure 16. Determination of Isosbestic Point in the UV Spectra of Photocyanation of HCB.
groups are present. A weak precedent for the latter pathway is the hydrolysis of
hexacyanobenzene to pentacyanophenol, but this is slow, even at elevated temperaturesg8.
Either of these explanations is compatible with the observation that raising the
concentration of water in the solvent system increases the concentration of 27 and 28 in
the produa mixture relative to 26 (Table 2); the phenolic products are resistant to further
substitution of chlonne, because the phenol is immediately converted to its conjugate
base. Consistent with this postulate, pentachlorophenol did not undergo any reaction even
after a 24 hour irradiation at 300 nm in the presence of sodium cyanide.
Table 2. Influence of Concentration of Water on Product Formation in Photolysis of HCB
Phenolic Produas by ES-MS, % [water], % 26 27 28 In an attempt to distinguish between CI substitution and CN substitution in the
formation of the phenols, we exarnined the hydrolysis of 1,4-dichloro-2,3,5,6- tetracyanobenzene (29). The dark reaction of OK (2.5% NaOH in acetonitrile-water) with 29 resulted in the formation of 2,5-dichloro-3,4,6-tricyanophenol thus demonstrating that -CN, but not -CI, was substituted by hydroxide (Figure 17).
Figure 17. Dark Reactions of 1,4-Dichloro-2,3,!j,6-tetraryanobenzene (29)
However, in acetonitrile/I 0% water, with no hydroxide ion added, hydrolysis of 29 took the alternative route (replacement of Cl by water), and 93% of
27 was detected in the product mixture. When 29 was treated with CN- in acetonitrildwater mixture (9: pentacyanophenoi (26) was a major product, and only 18% of 27 was detected in the product mixture. Similar trends were observed when dichlorotncyanophenol was reacted with water, OK and CN (Table 3). Table 3. Product Formation in Ground State Reactions of Chlorinated Polycyanobenzenes Reagent Water NaOH NaCN Reactant c9cl3N3 29 G~c13N3 29 c9cl3N3 29 Products (%):
a. This compound may be 28 or its isomer.
The material balance of the preparative sale photolysis of HCB in the presence
of cyanide ion (Section 2.459) was ca. 80°/0, showing that photocyanation followed by
hydrolysis is a major reaction route. Although the rnissing matenal (20%) may be
attributed to losses during the workup procedures, we explored the possibility of
photochemical homolysis as a side-reaction. Since the acetonitrile-water mixture is a
fairly poor source of hydrogen atoms, and the W light at 300 nm has barely enough
energy to bnng about carbon-chlorine bond dissociation, the conditions for the reductive
dechlonnation of HCB are unfavorable. A product of dechlorination
(pentachlorobenzene) was detected as a very rninor reaction product. Its concentration
increased on adding 10% isopropyl alcohol to the reaction mixture (ratio of penta- to
hexa-chlorobenzene in the product mixture changed fiom 1250 to 1:8 upon addition of
isopropyl alcohol to acetonitrWwater; however, the same change (firom 1:300 to 1:9)
was observed in control samples without NaCN, indicating that dechlorination occurs through a minor competing pathway of C-Cl homolysis of HCB, or possibly fiom the
decomposition of an excimer of HCB".
2.2.1.3.Reaction Kinetics
The overall disappearance of HCB follows pseudo-fist order kinetics
(ln{[HCB]Jm]roc t, please see notelw) with apparent rate constants ranging from -3 -1 1.1~1O s to 6.7~1O5 S-' depending on the concentrations of water and sodium cyanide.
The absolute quantum yield for the disappearance of HCB was difficult to determine
because the products are much more strongly absorbing than the starting material.'O1 In
addition, sodium chloride, a reaction by-product, forms a cloudy suspension, which
causes substantial light scattering. Extrapolation of the apparent #diss (the quantum yield
of disappearance of HCB) to zero time at PaCN = 0.324 mol L-' and PCB] =
3.88~10~mol L*'in 955 acetonitnle: water mixture gave a limiting quantum yield of
0.14 at 313 nm (Figure 18). A notable feature of the reaction is that the subsequent
photocyanation steps must each proceed with q5djss > 0.14, because no intermediate
produas such as pentachlorobenzonitrile or tetrachlorodicyanobenzene are detected in the product mixture. We detennined the quantum yield of the secondary cyanation by measuring the rate constant of disappearance of pentachlorobenzonitrile (Figure 19) and cornparhg it to the one for HCB, with the correction for the difYerent light absorbance.
The expenmental conditions were chosen such that the initial concentrations of HCB
(A3i3=0:054) and pentachlorobenzonitrile (Ag13=0.278),the concentration of sodium cyanide and percentage of water in acetonitrile were equal.
The following considerations enabled us to calculate the quantum yield of disappearance of pentachlorobenzonitrile. By definition, the quantum yield of disappearance of a chernical substance may be expressed as the ratio of the change in concentration (-dc/dt) and the amount of absorbed light IA (Equation 16). The amount of absorbed light is dependent on the incident light (Io), concentration and the extinction coefficient(&), Equation 17. Mathematical manipulations of the equations 16 Figure 18. Estimation of a Quantum Yield for Disappearance of HCB The solvent was acetonitrile containing 5% water
O 500 Io00 1500 2000 2500 3000 3500 Time, sec
Figure 19. Pseudo-first order Kinetics in P hotocyanation of HCB and Peatachlorobenzonitrile and 17 resulted in Equation 18, which was integrated using the Table of Standard
Int ega1.s 'O2 (Equations 19 and 20). Since the concentrations at time t cm be
determined for given compounds from the reaction rate constants k, (Equation 21) and
their extinction coefficients were calculated from the absorption spectra, the quantum
yield of disappearance pentachlorobenzonitrile #s can be calculated (Equation 22).
Substituting the value of the limiting 4d,, for HCB (0.14) into Equation 22 resufts in a
value of 0.51 as the quantum yield of disappearance of pentachlorobenzonitrile (b).
Therefore the total photon efficiency of photocyanation of HCB is very high. This is an
example of auto-acceleration reaction, previously observed in photocyanations (Section
2.1.2.3). We cannot Say if the following step, cyanation of a tetrachlorodibenzonitrile is a
photochernical or a ground state reaction since Our attempts to obtain this compound were unsuccess~l. As was stated in the introduction to this Chapter (Section 2-1-23},
photocyanations in the absence of added electron acceptors can take place through the
intemediacy of an excimer,'" such that attack on the excimer by nucleophile permits
charge to be removed nom the excimer in the form of the anion radical. In order to
investigate whether an HCB excimer was involved in product formation, we measured
#djSs as a fùnction of HCB concentration. The plot of @diS;' vs. (initial FCB])-' eigure
20) had a negative dope (-4.1 5*O.34 x 1O-' mol L-'), indicating that although an excimer of HCB is indeed formed upon irradiation, it is unreactive. A precedent for an unreactive excimer exists in the homolysis of chlorobenzene, the quantum yield of which decreases with increasing ~h~l].~~
6 -
5 -
8 -4 - \a - A 3- + 30.7 min A '1 0 20.4 min 1 - A 12.5 min
O 8 O 1OOOO 20000 30000 40000 50000 IqHCB], moi L"
Figure 20. Double Recip rocal Plot of #& versus [HCB] The three sets of Lines represent data points taken at dinerent stages of the reaction; the apparent 4, changes with percent conversion as shown in Figure 19,
The reaction between photoexcited HCB and CNwas, as expected, first order in 2.986 +0.796wa~~]-',2 = 0.983) when WCBJ = 3.74~10~mol L-'. Increasing the
water content of the aqueous acetonitnle at constant [ClTl significantly reduced +diss, on
account of the decrease in the ratio CN-:OH- (Table 4, also see Section 2.1 -2.2).
0.00 5.00 10.00 15.00 20.00 25.00 1(NaCN], mor' L
Figure 21. Double Reciprocal Plot of 4m, of HCB versus WaCN]
Table 4. Influence of Concentration of Water on Percent Conversion of HCB at Equal 'Irradiation Times
[water], % %Conversion of ECBa
14.6 24 0.06 a. Initial conceritmtion of HCB was 1.67~104M, wilh FJaCNJ = 0.13 1 M b. Apparent &,, not extrapolated to zero conversion Photocyanations of aromatic hydrocarbons in acetonitrile solution are known to
proceed via the singlet excited state in the presence of electron acceptors, but via triplet
excited state excimers in the absence of electron acceptors3'. The reported quantum yield
of intersystem crossing for HCB is 0.5'. consistent with either singlet or triplet state
reactivity. The following observations indicate a triplet state reaction. The rate of
photocyanation was two-fold faster in degassed than in aerated solution, and sensitization
of the HCB/CW reaction (ETof HCB is 307 k~/rnol)~~~could be achieved by acetone
(Ef332 k~/rnol)'~~or acetophenone (Ey3 11 k~/rnol)'~~.A triplet was shown to be the
only reactive excited state in the process of interest, since a Stem-Volmer plot of the
quenching of HCB by ferrocene el59kJ/rno~)'~~ was linear (Figure 22): = 1-27 +
O.OE+OO 5.0504 1 .OE-03 1 .SE-03 2.0603 2.5E-03 3.0E-03 3.5E-03 [Fe], Mol L"
Figure 22, A Stem-Volmer Plot of Quenching of HCB Photocyanation by Ferrocene The foregoing expenments led us to write Scheme 1 as the tentative mechanism
for photocyanation of HCB. The products are proposed to form fiom the triplet excited
state of HCB,most likely via the SN2(Ar*) mechanism, with excimer formation acting as
an unproductive side reaction. The rninor amount of homolysis has not been included in
the Scheme.
Scheme 1. Proposed Mechanism of Photocyanation of HCB
~AKI *:'KI * 5 3~1*+~1
k4 'Arc1 * +CN - + products
Steady state analysis of Scheme 1 led us to Equation 23, which gives the rate of
disappearance of HCB, and Equation 24, which is the relationship between @&' and
w~cN]''.Equation 23 takes account of the conditions under which these experiments were carried out, namely low light absorption by HCB. Consequently, the rate of produa formation increases with WCB], even though there is not a reactive excimer, simply because more light is absorbed. It was therefore necessary to make a careful determination of light absorbed at each concentration of HCB before computing Note that a reaction scheme in which the excimer lies on the route to product gives the relationship &[' = [HCBI-', with the dope being positive (cf: Figure 20). The parameter InterceptlSlope from Equation 24 had the value 4.72; this parameter is k4/(k3+ k@CB]), frorn which the ratio of 'HCB molecules that are trapped by CN can be calculated for any concentration of CN-and HCB. More usefully, the fiaction of 3~~~ molecules that are trapped by CN is equal to 4.72[CN-]/(1 + 4.72[CN']). At PaCw =
0.324 mol L" and [HCB] = 2.56~10~mol L-', this fraction is 0.6.
The results of our study of the mechanism of the HCB photocyanation are most readily accomrnodated by an SN2Ar*process. It seems unlikely that an electron-deficient compound such as HCB would form a radical cation, and no obvious electron donors are present to produce a radical anion. Moreover, the excimer of HCB, though formed, appears to be unproductive. There is a possibility that the subsequent photocyanation step(s) proceed(s) via the excimer or rather exciplex mechanism. The increase in the disappearance quantum yield in the second step may be attributed to the autocatalytic effects of the formed polycyanated products, or the introduction of each CN group makes the substrate more electrophilic. Polycyanochlorobenzenes are potentially good electron acceptors and may cause a switch in the reaction mechanism to SR+N~AT.;the precedents of the autocatalytic effects in photochernical and ground state cyanations were described in the Introduction to this Chapter.
Additional support for the hypothesis of the initial SN2Ar*chlorine displacement, followed by the SR+N 1Ar* p hotosubstitution and the ground state cyanatiodhydro 1ysis, cornes fkom Our study on the photocyanation of lower chlorinated benzenes, which are discussed next.
2.2.2. Photocyanation of Penta- and Lower Chlorinated Benzenes
Irradiation of the lower chiorinated benzenes for 20 hours in the presence of sodium 'cyanide resulted in the disappearance of the staring materials and formation of the polar products, visually similar (yellow-greenish solutions) to those formed in photocyanation of HCB. ES-MS analysis of the products of photocyanation revealed that pentacyanophenol is the major product for al1 congeners (Figure 23) except pentachlorobenzene, for which a tetracyanochlorophenol predominates (Table 5).
The importance of the observed results is that in penta- and less-chionnated benzenes a hydrogen atom substitution by cyanide occurs in the absence of added electron acceptors. Since the H substitution usually occurs by SR+N 1Ar* mechanism
(Introduction, Sections 2.1.1 and 2.1.2) and the excimer in photocyanation of HCB, although formed, is unreactive, the hydrogen displacement may occur via an exciplex, formed between the substrate and the cyano products of the chlorine displacement.
Table 5. Products of Photocyanation of Penta- and Les-Chlorinated Benzenes Starting : 1,2-Dichiorobenzene 1,3-Dichlorobenzene 1,4-Dichlorobenzene 1,2,3 -Trichlorobenzene Product: Relative Percent Formation (based on ES-MSanalysis) (232N40- 11% 15% 13% 0% Figure 23. Photocyanation of Tetra-, Tri- and Dichlorobenzenes
A simple experiment aliowed us to determine the influence of the degree of benzene chlorination on photoefficiency of photocyanation. Solutions of HCB, pentachlorobenzene and 1,2,3,4-tetrachlorobenzene with the substrate concentrations adjusted for equal light absorbance, having equal PaCN] and were irradiated at
300 m. Under such experimental conditions the ratios of measured pseudo-first order rate constants (Figure 24) were proportionai to the ratios of the corresponding #d,, Since for HCB under these conditions = 0-14, the quantum yields of disappearance of pentachlorobenzene and 1,2,3,4-tetrachlorobemene were ca. 0.07 and 1.4~10" respectively. This trend demonstrated that in the bemene series the quantum efficiency of photocyanation increased with the number of chlorine substituents on the aromatic ring of a substrate (cf: p.55 - substrates with more electron withdrawing groups are more reactive towards cyanide ion) 56 Time, sec
Figure 24. Measurement of Pseudo-first Order Rate Constants for Photocyanation of HCB, Pentachlorobenzene and 1,2,3,4-Tetrachlorobenzene in Solutions with Equals Absorbancies
2.2.3. Photocvanation of Polycyanobenzenes
Irradiation of 1,Cdicyanobenzene in the presence of NaCN resulted in the formation of pentacyanophenoxide as a major product (Figure 25). ES-MS analysis of the reaction products at 95% conversion also revealed the presence of small amounts of produas with dz193 and 195 in the ratio of 1:2.5. These produas were tentatively identified as a tetracyanophenoxide (CioHN40-) and a dihydrotetracyanophenoxide
(~10~~~40').The formation of a dihydro product, as was discussed eadier (Section
2.1.2.3) is typical for replacement of hydrogen by CN in the absence of oxidants. CN Figure 25. Photocyanation of 1,4-dicyano- and 1,2,4,Stetracyanobenzenes *- possible structures; products not isolated
Photocyanation of 1,2,4,5-tetracyanobenzene produceci simi lar products, with the exception that a dihydro product was formed in a higher yield. (Figure 25).
A possible mechanism for the photocyanation of 1,2,4,5-tmacyanobenzene which rationalises the formation of the observed produas is depicted in Figure 26. Interaction of photoexcited 1,2,4,5-tetracyanobenzene with a ground state molecule results in an excimer or radical-ion pair. Attack by hydroxide nucleophile on the radical cation 29 results in hydrophenoxy radical 30. This species may disproportionate transfer) giving a tetracyanophenol (m/z 193) and a dihydrotetracyanophenol (m/z 195), or accept an electron (not shown) hom the radical-anion counterpart of 29, forming a o-cornplex, also leading to m/z 195 product upon protonation. Nucleophilic addition of CN to 29 results in a hydropentacyanoaryl radical 31. Disproportionation of 31 by H transfer (not shown) would lead to pentacyanobenzene 32 and a dihydropentacyanobenzene, which by losing
HCN will revert to the starting matenal. Alternatively, an electron transfer may occur, giving a+ and o- complexes 33 and 34. Loss of cyanide ion form 34 leads back tc 1,2,4,5-tetracyanobenzene,while the OHattack on 33 would ultirnately result
58 Figure 26. Possible Mechanisrn of Photocyanation of 1,2,43-tetracyanobenzene
59 in formation of m/t 193 product and pentacyanobenzene. The latter product could not be detected by mass spectroscopy since it should be more reactive towards cyanation/hydrolysis than the starting material and is easily converted to pentacyanophenol26.
Disappearance of the starting material followed pseudo-first order kinetics for both cyanated benzenes. At PaCV = 0.125 mol L-' the observed rate constant was
6.86~10"S-' (If = 0.986) for 1,4-dicyanobenzene, and the value of 1.86xlo5 S-l (2 =
0.997) was obtained for 1,2,4,5-tetracyanobenzene. Although 1,2,4,5-tetracyanobenzene is considered a better electron acceptor than I,4-dicyanobenzene, to our surprise, the calculated quantum yield of disappearance of the former compound (0.03) is much lower than that of the latter (0.23). We can assume that this quantum inefficiency, aiong with a higher Yield of a dihydro product in case of 1,2,4,5-tetracynobenzene is caused by the formation of an unproductive intermediate, similar to the one below, which easily reverts to the starting material.
Although no "dark" reaction between the cyanide and 1,Cdicyanobenzene was detected, addition of 0.1 mol L-' solution of NaCN in acetonitrile - 10% water to the solution of 1,2,4,5-tetracyanobenzene in acetonitrile in the absence of light caused rapid development of a deep purple colour and disappearance of the starting material (Figure 27). Analysis of the reaction mixture fkom the "dark" cyanation of 1,2,4,5- tetracyanobenzene by ES-MS revealed the formation of products of high molecular mass: m,z 384 (71%), m/z 406 (26%) and m/z 455 (3%). These products could not be identified. No tetracyano or pentacyanophenoxide but only a trace amount of a dihydrotetracyanophenoxide was found in the ES-MS spectra.
Tirne, sec xQ Groundmoto -- State . -. - .
Figure 27. Photo versus "Da&" Cyanation of I,2,4,Stetracyanobenzene Pseudo-first Order Plots
The photochernical cyanation of polychiorinated compounds was not limited only to the polychlorinated benzenes. In the following Sections 1 will describe the research on successive photocyanation of highly chlorinated polycyclic aromatics, such as biphenyls, naphthalenes and dioxins. 2.2.4. Photocvanation of Polvchlorinated Biphenvls
Photolysis of the title compound in aqueous acetonitrile in the presence of sodium cyanide led to formation of a complex mixture of hydroxybiphenyls with various degrees of cyano substitution. Eight groups of congeners, bearing 50m one to eight cyano groups were identified by ES-MS (Table 6). The following conclusions arise from analysis of the product composition:
Introduction of a hydroxy group and possibly termination of firther cyanation may
occur at early stages of reaction, when just one or two chlorines are substituted by
CN-.
Cyano substitution may occur on both rings of decachlorobiphenyl, since biphenyls
bearing five and more CN substituents were detected.
~imbltaneoushydroxylation of both arornatic rings on highly cyanated biphenyls is
unlikely, since no di-hydroxy products were found (please see notelo5).
TabIe 6. Products of Photocyanation of Decachlorobiphenyl
Relative Abundance, 7 12 20 24 16 8 7 6
Isolation of the individual products for characterisation and structure determination was not practical. However, we used decachlorobiphenyl as a mode1 compound to study photophysical aspects of the successive photocyanation of polychlorinated biphenyls. 2.2.4.2. ~hot6~h~sicalAspects of Photocvanation of Decachlorobi~henvl
Similar to HCB, successive photocyanation of decachlorobiphenyl followed pseudo-fist order kinetics (Figure 28). The observed pseudo-first order rate constant was high, and 90% conversion was reached within 15 minutes of irradiation in the presence of
NaCN. However, significant amounts of reductive dechlorination products were found by
GC-MS.. From the ratio of the rate constants for the disappearance of DCB in the absence and the presence of sodium cyanide it was estimated that ca 13% of the title compound is consumed by reductive dechlorination. This dechlorination probably occurs via simple homolysis; although ET of decachlorobiphenyl is unknown, the triplet energies of several chlorinated biphenyls were found to be high enough to allow C-Cl bond cleavage energetically. 99,106 Measurements of the quantum yield of disappearance of decachlorobiphenyl gave the limiting value of 0.26 for @disS. The quantum yield is higher
O NaCN Present x No NaCN 2.50 -
2.00 - h Ci
c 1.O0 - y = 4.24x104x r2 = 0.882 X
O 200 400 600 800 1O00 lime, sec
Figure 28. Pseuddirst Order Kinetics of Photolysis of Decachlorobiphenyl in the Presence and Absence of Sodium Cyanide than that of HCB (0.14); as in the former case, the limiting value was obtained by
extrapolation of apparent to time zero (Figure 29).
O 20 40 60 80 100 120 lime, min
Figure 29. Determination of Quantum Yield of Disappearance of Deeachlorobiphenyl (Co= 2.4~IO-' mol L"; Badon Quadmplicate Samples).
Results of quenching and sensitization experiments allowed us to conclude that photocyanation of decachlorobiphenyl is a triplet state process. The #d,, value increased in the presence of acetone and acetophenone, and when oxygen was removed by freeze- purnp-thaw cycles (Section 2.4.12.6, Table 25). A Stem-Volmer plot of the quenching of decachlorobiphenyl by ferrocene had a positive slope of 784 mol-1 L with 3 = 0.975
(Figure -30).
The low solubility of decachlorobiphenyl in acetonitrile precluded a test of the dependence of q&iss on the concentration of DCS; however, an excimer involvement in the photocyanation mechanism in this case is unlikely due to low initial concentration of the reactant (usually, 2.5x IO-' mol L"). O.OE+OO 4.OE-04 8.OE-04 1.2E-03 1.6E-03 [Ferrocene], mol 1-'
Figure 30. Quenching of Photocyanation of DCB by Ferrocene
Photolysis of DCB at WaCN] ranging fiom 0.285 to 0.020 mol L-' resulted in decrease of the limiting q& from 0.29 to 0.07 (Section 2.4.12.5, Table 24) and, as for
HCB, the relationship l/#d,ss cc l/paCN] was approximately linear (Figure 3 l), demonstrating that the photocyanation is first order in both excited DCB and sodium cyani de.
Increasing the water content in the solvent mixture up to 26% decreased the quantum yield of disappearance of DCB (Figure 32j. This effect was much less profound than that for HCB (see Table 2), where increasing m20] fiom 1.4 to 14% caused a three fold drop in (bdiss. Since 4i, o [&O] is linear (2 = 0.973), then, using the intercept of the plot in Figure 32, when m20] O &iSs is 0.27 at PaCV = 2.55x10-~mol L-' and
PCB]' = 2.63 x 1O-* mol L". Figure 31. A Double Reciprocal Plot of 4- of DCB versus [NaCNJ A dashed line and the equation represent the linear fit
Figure 32. Influence of Water on the Quantum Yield of Disappearance of DCB Scheme 2 depias a tentative mechanism for photolysis of DCB in the presence of sodium cyanide. The products of photocyanation are formed fiom the triplet excited state; the products of homolysis are also suggested to arise from the triplet, based on available literature data.99,106
Scheme 2. Mechanism of Photocyanation of Decachlorobiphenyl
k4 3 ArCl * +CN - -, cyanation products 4 Arc1 * + dechlorination products
Steady state analysis of Scheme 2 allowed us to determine the relationship between the quantum yield of disappearance of DCB (+diSs ), or the quantum yield of photocyanation of DCB (kN),and the concentration of sodium cyanide. Equation 25 demonstrates that when the reductive dissociation cornpetes with photocyanation, the plot l/(6diss CE l/waCN] should be non-hear at low concentrations of NaCN. According to
Equation 26 the double reciprocal relationship of the quantum yield of photocyanation of
DCB (kN)with no homolysis and the concentration of NaCN should be linear. The plot on Figure 31 may have some curvative character. Although a slow
homolysis was observed in control experiment (Figure 28), the linear regression on the
data yielded 2 = 0.978, hence the plot may be interpreted as linear. The kinetic analysis
of Scheme 2 implies that in the photocyanation of DCB the reductive dissociation is
negligible and q$d,, = hNwhen WaCN] is in the range of 0.285 to 0.020 mol L-'. When
WaCN] 3 m the intercept for the plot in Figure 3 1 is equal to (l+kl/k2), which is equal to
&&', su the quaotum yield of intersystem crossing for DCB is 0.46, compatible with the
reported value of 0.5'" for hexachlorobenzene. The obtained value of the quantum yield
of intersystem crossing is uncharacteristically low for highly chlorinated aromatic
compounds, since for such compounds & is usually close to the unity due to the heavy
atom effect.
2.2.4.3. The a-effect in Successive Photocvanation of Polvchlorobiphenyls
The title phenornenon, characteristic for nucleophilic aromatic substitution in
polycyclic aromatic compounds, was described in Section 1.2.3. To investigate whether the relative position of chlonnes on the aromatic ring of biphenyl affects the quantum efficiency of photocyanation, the following hexachlorobiphenyls were photolyzed in the presence of
NaCN. Pseudo-first order rate constants of 2,2',4,4',6,6'-hexachlorobiphenyl (35), with four chlorines in position ortho to the biphenyl bond, 2,2',4,4',5,Sr-hexachlorobiphenyl (36), having two orfho chiorines, and 3,3',4,4',5,5'-hexachlorobiphenyl (37) with no ortho chiorines were measured in the experiment where the initial concentration of al1 three substrates, the concentration of NaCN and the &action of water in acetonitrile were equal
(Figure 33).
O 500 Io00 1500 2000 2500 3000 3500 Time, sec
Figure 33, Psuedo-first Order Plots of Photocyanation of Hexachlorobiphenyls 35,36 and 37
At the initial concentrations of 1.0~10~mol L", the measured rate constants for
35-37 did not show any significant difference or structure dependence, they were in the range of 4.27~10~to 2.31~10~ c'. However, when the relative quantum yields were calculated using Equation 22, taking into account the extinction coefEcients of these compounds at 300 nm, with #di, for 35 adjusted to unity, the ratio of the quantum yields
of disappearance for 35, 36, and 37 was 1:0.45:0.05. This trend is consistent with the
a-effect for biphenyls, e.g. the isomer, having more chlorines in ortho position is more
reactive towards substitution; in this particular case, the difference in reactivity between
the hexachlorobiphenyl with four ortho chlorines and its isomer with no ortho CI was
twenty times. The reductive dechlorination has long been known to proceed faster (with
higher @diss) for chlorobiphenyls with orfho chlorines due to the higher triplet energy. 'O7
To rule out the possibility that the observed dependence of +dia,, is due to homolysis rather than cyanation biphenyls 36, 37, and 38 were irradiated in the absence of sodium cyanide
(under the same conditions). The measured rate constants for homolysis were an order of magnitude smaller (4.47~IO'.' to 1.26~10~'s") than those for photocyanation of 35, 36, and 37. Thus, during the irradiations when sodium cyanide was present the homolysis was almost negligible and the observed ratio of the was dictated by the ortho effect characteristic for the photocyanation, rather than homolysis.
The products of the photocyanation of 35, 36, and 37 were subjected to ES-MS analysis. For al1 the three substrates cumplex mixtures of products were observed, with the major products being a substance with m/r 371 followed by the one with m/z 380. The former produa displayed a three-chlorine pattern and was tentatively identified as a tetracyanotrichlorohydroxybiphenyl (sodium salt) with the molecular formula of
CL&C~~N~OU~+;the substance with m/z 3 80 (a four-chlorine pattern) had the molecular formula of c~&cw~oN~'.Since the isolation of these products was unpractical and the exact structures are unknown we cannot determine whether the substitution of chlonnes and hydrogens in 35, 36, and 37 by the cyanide occurred on the single ring, or whether the chlorines fiom both aromatic rings of biphenyl were displaced.
2.2.4.4. Photocvanation of Polychlorohvdroxvbi~henvls
ES-MS analyses of the reaction mixtures of hexachlorobenzene photocyanation have shown that the product ratio does not change with time for samples taken after 5, 50 and 120 min of the irradiation (Section 2.2.1). This observation led us to the conclusion that for HCB, the &rther cyanation is teminated when the hydroxyl is introduced in the product due to deactivation of the rnolecule towards nucleophile by the ionization of the phenolic OH. On the other hand, although pentachlorophenol failed to undergo photocyanation (Section 2.2.1.2), p-chlorophenol has been reported to give p- cyanophenol in 75% yield on irradiation in aqueous cyanide solution. *O8
When we photolysed 4-hydroxy-2',3 ',4 ',5 '-tetrachlorobip henyl (38) in the presence of sodium cyanide, the starting material rapidly disappeared (90% conversion in
25 minutes), and a mixture of hydroxybiphenyls with various degrees of substitution was formed. The major product was a hydroxy-dicyano-dichlorobiphenyl (m/z 287, not isolated); it may have a structure similar to 39, based on the higher reactivity of chlorines at positions 2 and 4 (Section 2.1.2).
We assumed that 4-hydroxy-2 ', 3 ',4 ', 5 '-tetrachlorobip henyl undenvent photocyanation because the deactivating effect of hydroxyl group, observed for pentachlorophenol, does not "translate" to the adjacent ring in 38. To investigate this possibility 3-hydroxy-2,4,6-trichlorobiphenyl40was irradiated in the presence of NaCN. Three products were found by ES-MS: m/z 219, mh244, and m/z 253 (1 chlonne pattern), in the ratio of 3:1:2.5.They were identified as a hydroxy- tricyanobip heny l (presumably 3-hydroxy-2,4,6-tricyanobiphenyl 4 l), a hydroxy- dicyanobiphenyl, and a hydroxy-chloro-tricyanobiphenyl. Measurement of pseudo-fist order rate constants of photocyanation for biphenyls 38 and 40 (Figure 34) and estimation of the relative quantum yields of disappearance by the method described earlier showed that compound 40 is much more labile towards photochernical cyanation than 38, since
40 38 #dis : = 411. Wecan rationalize these results ody by assuming that the hydroxy group in 40 acts as an ortholpara activator for cyano photosubstitution; this effect was described in Section 2.1.1 for the methoxy group.
O 200 400 600 800 1000 1200 Time, sec
Figure 34. Measurement of Pseudo-first Ordcr Rate Constants of Photocyanation for 4-hydroxy- 2'gr,4'$'-tetrachIombiphenyl (38) and 3-hydroxy-2,4,6-ttichlorobipheoyl(40) 2.2-4.5.Photocvanation of 2.3 -4.5.6-~entachlorobiphenvl
One of the major difiiculties in the study of photocyanation for polychlorinated
biphenyls was poor product selectivity; e.g. photocyanation of DCB and compounds 35,
36, and 37 resulted in the formation of such cornplex mixtures that the separation of the
major products was unfeasible. On the contrary, when 2,3,4,5,6-pentachiorobiphenyl, a compound with one ring fiilly chlorinated and the other unsubstituted, was subjected to irradiation in the presence of NaCN for 2 h, the resulting product mixture contained 80%
(ES-MS) of a single substance with m/z 303. The pseudo molecular ion (dz 303) displayed a one-chlorine pattern and was identified as C1&CIN40* (42). The minor products included m/z 337 (a two-chlorine pattern, C1&C12N40-) and m/r 294 (no chlorines, Ci7H4NsCF). Continuous irradiation of the product mixture for 48 hours resulted in the decrease of relative intensity of the ion m/z 303 and formation of new products (dz2 18, ml. 193 and m/r 195) which were tentatively identified as compounds
26, a tetracyanophenoxide (Ciorno-) and a dihydrotetacyanophenoxide (C ioH3N40-) respectively.
The compound 42 was isolated in 59% yield as a sodium salt upon a large-scale photocyanation of 2,3,4,5,6-pentachlorobiphenyl. Attempts to protonate this salt by treatment with concentrated hydrochloric acid (20%) resulted in decomposition, while less concentrated acid was too weak to cause protonation (based on the observation of the
UV spectrum of the product). Our attempts to crystallize this Salt out of the variety of solvents or to use ~a~~ as a counter ion to promote crystallization were so far unsuccessful, possibly due to the presence of several isomers of the product in the mixture. Proton NMR (400 MHz) for Cl&ClN40- was of Me help in the structure elucidation due to a cornplex coupling pattem (Appendix A).
However, the partial structure of the compound was deduced fi-om ES-MS, MS-
MS (electrospray) and 2D NMR data. Examination of the molecular formula of 42 suggested that during the photocyanation of 2,3,4,5,6-pentachlorobiphenyl one hydrogen was substituted by CW: although the starting material had only five chlorines, six substituints are present in Ci&CNO-. A fragment with ~I'z201 corresponding to
C&ClO'- was found in the MS-MS spectrum of 42; this was presumably formed by fragmentation of the biphenyl bond in 42, leaving a cyanobenzene radical as the counterpart.
The substitution pattem on the non-chlorinated aromatic ring was established by a
COSY experiment (Figure 3 5). The signals at 6 = 7.28 and 7.3 1 were due to impunties; they did not display the characteristic coupling 'square' with any other proton signals.
From the COSY spectra it appeared that the proton at 6 = 7.41 ppm (HA,doublet of doublets) is coupled to the proton at 7.49 ppm (He, appears as a doublet of triplets). The proton at 7.52 pprn WC,appears as a doublet of triplets) is coupled to Hg and HD (7.60 ppm, doublet of doublets). The coupling constants are typical for pnmary coupling in the aromatic ring: JA-B is 7.4 Hz and Jc-D is 8 Hi. Since the coupling JA~is only slightly different fkom Jac (7.4Hz vs. 7.6 Hz) the observed signal fiom proton B at 7.485 pprn is technically a doublet of doublet of doublets, but appears visually as a doublet of triplets.
The same considerations apply to Hc; unambiguous J values could only be determined for HA and HD. Long range meta couphg was observed between protons A and C (4~~-c
= 2 Hz) and between protons B and D (4~~~= 1.4 Hz). The above data were compatible only with a 3,4,5,6 pattern of C-H bonds hence the CN group had been incorporated at
the 2 position.
Heteronuclear co~ectivitywas established from HSQC NMR (Ap pendix B). The
chemical shifts of the proton-bearïng wbons are given in the Expenmental (Section
2.4.13.9). With the aid of a long-range proton to carbon coupling experiment (HMBC,
Figure 36) chemical shifts for carbon 1'(6 = 151.3 ppm, downfield due to the phenoxide
substituent) and carbon 2' (6 = 125.7 ppm, upfield due to the CN substituent) were
detexmined. Cross-signals between HAand Cl ', and between HDand C6' indicate that HA
is on the position 3' and Ho is on the position 6' of the biphenyl ring.
The data from ID 13cNMR experiment alone were not sufficient to determine
the placement of the substituents in the second aromatic ring, and 2D 13cexperiments
like DQF-PS COSY or INADEQUATE were ünfeasible due to insufficient quantity of
42. At present the complete structure of Ci&ClN40'is therefore unknown.
Dunng the photocyanation of 2,3,4,5,6-pentachlorobiphenyl a small amount of
product with m/z 337 (C1&t3C12N40-)was formed, in which two hydrogens are
substituted by the cyano group. Possibly, the direct precursor to Cl&Cl~N40- has a
structure similar to 43 or 44, where the hydrogens on positions 2', 4', or 6' are substituted
in favour for the chlorines at the positions 3 and 5; the product (Ci&ClzN40') is then formed by the replacement of one of the cyano groups by OIT. Figure 35. COSY Spectrum of Compound 42 Figure 36. HMBC Spectmm of Compound 42 It should be noted that the matenal balance in the large-scale photocyanation of
2,3,4,S ,6-pentachlorobiphenyl was relatively high (73%), showing that formation of the photocyanation products is the major reaction pathway.
2.25 Photocyanation of Octachloronaphthalene and Octachlorodibenzo-p-Dioxin
phot61yses of octachloronaphthalene (OCN) and octachlorodibenzo-p-dioxin
(OCDD) in the presence of NaCN were studied to demonstrate the versatility of the photocyanation reaction for different classes of highly chlonnated aromatic compounds.
For both OCN and OCDD the product mixtures were complex, and no attempts to isolate the cyanated products were made. It is worth mentioning that, according to ES-MS data, a tetrachloro-tricyanohydroxynaphthalene (z 354) and a pentachlorodicyano- naphthalene (mh 363) predominate in the product mixture of OCN and only traces of higher cyanated naphthalenes are found. Funher photocyanation may be hindered by the low reactivity of P-chlorines in naphthalene towards the photosubstitution by cyanide.
For both title compounds, reductive dechlorination competed with photocyanation
(Figures 37 and 38); in the case of OCDD the pseudo first order rate constant for the
NaCN-assisted photolysis was only twice as large as the rate constant for the irradiation in the absence of NaCN. The quantum yields of disappearance of OCN and OCDD were rather small, 2x 104 and 0.001 respectively. Similar to HCB and DCB, the && for both compounds decreased with the increasing water content from 2 to 10% in the solvent mixture (Sections 2.4.14.4 and 2.4.15.4, Experimental).
The results of sensitization experïments with acetophenone and acetone as the sensitizers suggest that photocyanation of OCN and OCDD is also a triplet state reaction.
Attempts to quench photolysis of OCN or OCDD in the presence of NaCN by ferrocene Figure 37. Photolysis of OCDD in the Presence Figure 38. Photolysis of OCN in the Presence and absence of NaCN. Pseudo-first Order Plots and absence of NaCN. Pseudo-first Order Plots failed; increasing the concentration of ferrocene resulted in the enhancement of the reaction in both cases (Figure 39). Examination of the half-wave reduction potentials for OCN (Ein = 1.29~)'~~and ferrocene (Ela = 1.33V)110revealed that the electron transfer frorn ferrocene to OCN is possible. Although the half-wave reduction potential for OCDD is not reported in the literature, the electron transfer is likely to occur, since the addition of electron donors significantly enhances photolysis of OCDD in hydrocarbon solvents."' It shouId be noted that in previously described quenching of
DCB no interfenng electron transfer from ferrocene to substrate was possible since for
DCB Eln = 1.76v.lo9
Relatively high solubility of OCN in aqueous acetonitrile allowed us to investigate the dependence of the quantum yield of disappearance on the starting concentration of OCN. The results of the experiment, presented in Figure 40, suggest that increasing the initial concentration of OCN results in higher #dissi in other wordq the photocyanation of OCN may proceed via the intermediacy of an excimer. Figure 39. Attempts to Quench Photocyanation of OCN and OCDD by Ferrocene
2000 - o 30 min hv 1000 - 50 min hv Linear (50 min hv) O 1 I I 1 O.OE+OO 5.OE+04 1 .OE+05 1.5€+05 2.0€+05 2.5E+05
l/[OCN], mol L-1
Figure 40. A Double Reciprocal Relationship of the Quantum Yield of Disappearance of OCN and the Initial Concentration of OCN 2.2.6. Photochernical Cvanation of Hexafluorobenzene
Fluorine is considered the best leavhg group among halogens in photochernical nucleophilic substitution (see Section 2.1 -2.1). Bearing that in mind, we atternpted to induce the photosubstitution of fluorine in hexafluorobenzene. At 300 nm, where Cas is transparent, photolysis was slow (55% conversion in 47 hours). Irradiation of the
8.75~1o4 mol L-' solution of C6F6in the presence of 0.05 moi L-'N~CNat 254 nm (&fi
= 0.136) resulted in rapid disappearance of the starting material. The reaction did not follow pseudo-first order kinetics (Figure 41), and 85% conversion was reached within 5 minutes of the irradiation. Analysis of the reaction mixture by ES-MS revealed that the major product was a difluorotetracyanobenzene (dz214), with trace arnounts of 26 (dz
21 8), a difluorotricyanophenoxide (dz204), a tetrafluorocyanophenoxide (m/r 190), and pentafluorophenoxide (m/l 183).
O 100 200 300 400 Time, sec
Figure 41. Photolysis of Hexafiuorobenzene in the presence of NaCN at 254 nm A dashed line and the equation represent the linear fit 2.3. Conclusions
The novel aspects of the successive photocyanation of highly chorinated aromatic compounds are the discovery of the first photocyanation without the assistance of an added electron acceptor, and the observation of multiple photosubstitution steps, each proceeding with high efficiency. The reaction reported here seems to be cornmon to a wide variety of polyhaloaromatic compounds, such as polychlorinated benzenes, highly chlorinated polycyclic aromatics, such as biphenyls, naphthalenes and dioxins, and polyfluonnated aromatic compounds.
The synthetic potential of the successive photocyanation is low. Although for some particular compounds some selectivity in product formation and high yields were observed, the products of the photocyanation, due to their high acidity, can only be isolated in their conjugate base fonn as salts. The quantum yield of photocyanation varies fiom 0.26 to 0.0004 for different classes of polyhaloaromatic compounds and it depends on the concentration of sodium cyanide in solution and the percentage of water in the solvent system. For HCB, 4diSs decreased with the increasing concentration of the starhg material due to the formation of unproductive excimer, while photocyanation of DCB and
OCN may proceed via an excimer at high initial concentrations. The photocyanation proceeds through a triplet state for al1 compounds that were studied. 2.4. Experimental
2.4.1. General
HPLC grade acetonitnle, rnethanol and ethyl acetate were obtained nom Fisher. All solvents were used as received. Potassium cyanide (99%f isotopic purity) was purchased from Isotech Knc.
2.4-2, Irradiations
Solutions were photolyzed in 8 mm 0.d. Pyrex0 test tubes in a Rayonet RPR photochernical reactor equipped with 16 RPR-3000 low pressure rnercury lamps which emit a band at 300 nm with a bandwidth of 25 nm at a haif-intensity. When necessary, samples were degassed and sealed on a vacuum manifold using four freeze-pump-thaw cycles (las two cycles using an oil-ciiffision pump, P < 0.01 torr). Preparative photocyanations were done using up to four 30mm 0.d. Pyrex@ test tubes sealed with rubber septa and placed in a
Rayonet merry-go-round p hotoreactor.
2.4.3. AnaIysis
The course of reactions was monitored by HPLC equipped with Waters U6K injecter, Waters 600E system controller and Waters 486 hrnable absorbame detector and
Waters p-Bondapack Ci* 39x300 mm column, using acetonitrile/water (9/1) as a mobile phase. Calibration mesfor analytes (t%.99) were built using Millemiurn II software.
Absorption spectra of analytes and transmittance spectra of optical filters were acquired on a doub le-beam Shimadm UV-VIS recording spectrophotometer (Mode1 UV-16OA). Carbon
Nuclear Magnetic Resonance (13c NMR) spectra were recorded at 100.6 MHz on
Avance400 Bruker spectrometer, using a~etonitn1e-d~as a solvent. GC-MS analyses were perforrned using a Varian Sam3 Ion Trap MS coupled with Varian Star 3400CX gas chromatograph, the latter being equipped with SPI injector and DB-5MS capillary colurnn
(30m x 0.25mm x 0.25 pm). The canier gas was Ultra High Purity helium @OC, Guelph,
Ont.), the sarnples were injected in hexane, with the volumes of injection not exceeding 1.0
a.~he acquired spectra were analyzed using Satum 4D software Version 5.2.
2.4.4. Mass Spectrometry
Samples were prepared for electrospray mass spectrometry by diluting the reaction mixtures in a solution of 50% acetonitrile and 50% water containhg either 0.1% ammonium hydroxide (for negative ion spectra) or 0.1% formic acid (for positive ion spectra) (Caledon, Georgetown, Ont., CAN) to approximately 10 pg/rnL. The sample solution was then introduced into the mass spectrometer via an injection valve
(Rheodynem mode1 7010) with a 10 pL capacity. The mobile phase consisted of 50%
NANOpure water (Barnstead, Dubuque, Iowq USA) and 50% acetonitrile (Caledon,
Georgetown, Ont., CAN), and was delivered using a binary LC pump (Hewlett Packard
1090 Series II/L, Pa10 Alto, CA, USA) at a flow rate of 15 pL/rnin.
The mass spectra were obtained on a VG Quattro II (Fisons UK Ltd., Altrincham,
UK) triple quadrupole mass spectrometer equipped with an atmospheric pressure ion source. Instrument control, data acquisition and data processing were carried out using a
~ass~~nx~"software package.
Liquid Nz (Praxair, Kitchener, Ont., CAN) was used as drying and ES nebulising gas at flow rates of 450 and 20 Lh, respectively. The ion source temperature was 65'C.
In the negative ion mode the instrument was operated at a capillary needle potential of -
4.20 kV and a cone voltage of 20V. The mass spectra were acquired in multi channel analysis (MCA) data acquisition mode by scanning the first quadrupole in 0.1-u increments nom dz100 to 800 in 1.2s. Mass spectra were averaged over at least 8 scans.
Product-ion tandem rnass spectra were acquired by transmitting the individual w-H]-ions through the first quadrupole and into the second quadrupole, where the ions underwent collision-induced dissociation (CID) with ultra pure argon (Praxair, Kitchener,
Ont., CAN) gas. The fragment ions arising from the selected W-Hl- ion were detected by scanning the third quadrupole fiom m/z 45 to 800 in 1-2 S. The argon gas pressure was 2.0 x 104 mbar, and the collision energy was varied between 20-50 eV to optimize fragmentation of the individual FI-Hl' ions. The mass spectra were obtained in MCA data acquisition mode, which were averaged over a minimum of 5 scans.
2.4.5. Photochernical Cvanation of Hexachlorobenzene
Hexachlorobenzene (Aldrich) was twice recry stallized from chloroforrnhnethan01 before use.
2.4.5.1. Measurernent of Ouantum Yield of Disap~earanceof HCB
Quantum yields of substrate disappearance were determined by potassium femoxalate actinometry112using solutions of potassium chromate and nickel sulphatel" as an optical filter. The actinometer in water and the test solution in acetonitrile, containing 5% water, were placed in 8 mm 0.d. test Pyrex@ tubes which were positioned coaxially in 25 mm 0.d. Pyrex@ test tubes filled with optical filter. The filter transmitted in the 285-335 nm region, with maximum transmittance of approximately 20% at 3 13 m. The actinometer and test solutions were irradiateci simultaneously in the photoreactor using the merry-go-round to assure equal light intensity incident on al1 solutions. Al1 the light was absorbed by the actinometer and the photon intensity absorbed by the ahorneter &=7.09x 1O-'&~.S~X 1 o4
mol min') was calculated based on hlrl-24 for potassium ferrioxalatel 14.
The light intensity absorbeci by the test solution IA was calculated using the
following equation:
IA (mol) = Io (mol min-')x time (min) x (1- 1O-*)
Solutions of HCB (3.88~10~mol L-', A313= 0.01) and NaCN (0.183mol L-') in
acetonitrildwater (5% &O) were degassed, sealed under vacuum and irradiated for various time intervals, ranging fiorn 10 min to 2 h. Mer irradiation, the ampules were opened and
disappearance of HCB was determined by HPLC analysis (averaged results of triplicate
experiments are presented in Table 7, cf: Figure 19). The apparent quantum yields were
calculated and plotted against Lk.
Table 7. Measurement of Quantum Yield of Disappearance of Hexachlorobenzene
Time, min Conversion, % ~ebmol 1, mol &b 2.4.5.2. Photolysis of Hexachlorobenzene in the Presence of NaCN for Various Periods of
Tirne
Eight samples were prepared by adding 0.5 rnL aliquots of 7.49~10~mol L-'
solution of HCB in acetonitde and 0.5mL aliquots of 2-851 x 10-l mol L-l solution of
NaCN in acetonitnle-water (10% v/v HzO)to 8 mm Pyrex@ tubes. Seven sample tubes
were degassed using three freeze-pump-thaw cycles and sealed under vacuum (0.01
Torr); one tube was capped with a rubber septum. Six degassed and one non-degassed
sample tubes were placed in a merry-go-round photoreactor and imdiated at 300 nrn for
various time intervals, one degassed tube was wrapped in aluminum foi1 and placed in a
cupboard as a dark control sample for the duration of the reaction. At given time intervals
a degassed tube was taken out of the reactor and opened for analysis; at the same time a
10pL aliquot was taken from the non-degassed sample and analyzed by HPLC. The
mobile phase was methanol with 10% water. The retention time of HCB was 4.18
minutes, retention time of products was 0.85 minutes. Calibration curve: WCB], mol L-l
Table 8. Photolysis of HCB in the Presence of Sodium Cyanide for Vanous Periods of Time
Concentration, mol L" The, Aerated ~e~assed- aerated degassed aerated degassed sec = Peak Area x 2.206~10'~,$=0.999. The concentration of HCB in solution, percent conversion and In(cdcJ values were calculated (Table 8, cf: Figure 18).
2.4.5.3. Sensitization of HCB Photocyanation bv Acetone and Acetoohenone and Oqgen
Removal
Concentrations of acetone and acetophenone were chosen such that most of the incident Light was absorbed by the seasitizers. Eight solutions of HCB (2.73~lo4 mol L*') in acetonitnldwater (5% H20),containing NaCN (8.06~1O-' mol L-') were prepared, with two having acetone (0.125 mol c', AIoo= 0.78) and the other two having acetophenone
(1.12~10" mol L-', A3rn= 0.57) added. The samples, except the two without sensitizers, were degassed by four freae-pump-thaw cycles, sealed under vacuum and irradiated at
300 nm for 7 min in the rnerry-go-round photoreactor. nie results of HPLC analysis of duplicate samples are presented in Table 9.
Table 9. Sensitization of HCB Photocyanation by Acetone, Acetophenone and Oxygen Removal
Sample: Aerated Degassed Acetone Acetophenone
k-mol L' 6.47~10%5.6~ 104 1.52~lo4fi.6x lad 2.42~104i3.6x lo4 2.67~l0~f6.63~ lod
Conversion, % 25 54 87 99
*-A value of &i&,is equal to cS-d / CO -a,where S and O represent samples with and without sensitizer 2.4.5.4. Ouenchine; of HCB Photocvanation bv Ferrocene
~hotol~seswere done in acetonitrile, containing 5% water, with the initial
concentrations of HCB and sodium cyanide equal to 1.67~10~mol L-' and 0.13 1 mol L"
respectively. Degassed samples were irradiated for 20 min and analyzed by HPLC (Table
10). The Stem-Volmer plot (Figure 22) was obtained by plotting the ratio of quantum yields
of reaction in the absence of the quencher to that in the presence of the quencher (MW)
versus the concentration of quencher.
Table 10. Quenching of Cyanation of DCB by Ferrocene
~errocene],mol L-' Conversion, % moi hhb
2.4.5.5. Photolysis of HCB at Various Concentrations of Sodium Cvanide
Non-degassed solutions of HCB (2.56~lo4 mol L-') in acetonitde/water (5%),
having various concentrations of sodium cyanide, were irradiated for 30, 50 and 80
minutes at 300 nm. Values for &ss were cdculated for each run and averaged (Table 11, cJ Figure 2 1). Table 11, Photolysis of HCB at Various Concentrations of Sodium Cyaoide
Conversion, %- NaCN, mol L" &~SS 3Omin hv SOmin hv 80& h
2.4.5 -6.Influence of Water on Photocvanation of HCB
Four 8 mm Pyrex@ tubes containing 1.0 mL aliquots of acetonitnle solutions
having constant concentrations of HCB and NaCN but various concentrations of water
were placed in the merry-go-round photoreactor and irradiated at 300 nm for 1 hour 40
min. The initial concentration of HCB was 1.86~10~mol L-', the concentration of NaCN
was 3.97x10-* mol L-l. Mer irradiation, samples were analyzed by HPLC (Table 12).
The function of vs. percent water was linear, with a negative dope of -0.0092 and
intercept of 0.186 (3= 0.98 1).
Table 12. Photocyanation of HCB at Various Concentrations of Water
[water], % kmol L-' Conversion, % hi= Irradiated samples were subjected to analysis by ES-MS. Areas of base peaks of products with m/Z of 2 18,227 and 236 were calculated by adding the signal intensities in the
2 17.2-217.9, 226.0-226.9 and 23 5.1-23 5.9 m/z ranges respectively. Peak abundances and relative percentages of the cyanation products are presented in Table 13.
Table 13. Products of Photocyanation of HCB at Various Concentrations of Water --- . Peak Abundance, au, (Relative Percentage) [water], % m/t 218 m/z 227 dz236
2.4.5.7.Determination of Isosbestic Point in Cyanation of Hexachlorobenzene
A 1.5mL solution of 2.53x 1o4 mol L-' HCB in acetonitde and 1.5 mL of
1.59x10-~mol L-' NaCN in acetonitrile/lO%water were placed in a Ixlcm square quartz cuvette. Absorbance spectra of the mixture were acquired in the 200 to 650 m region with the higher absorbance Iimit set to 2.5 and the lower set to -0.1 units. The baseline was corrected against acetonitrile with 3.3% water (v/v). The cuvette was secured in the meny-go-round photoreactor by means of a rubber septum and the solution was irradiated at 300 nm. Mer 1, 2.5, 5, 10 and 20 minutes the cuvette was taken out of the reactor and the absorption spectrum was recorded. Acquired spectra were scanned at resolution of 300 dpi in grayscale mode using Microtek Scanmaker IIEX The images were imported into Adobe@ Photoshop 4.0.1 and placed into individual layers. White background was removed using 'select color range' command with the 'highlight' option 'on' and then deleting the selections on each layer. The spectra were aligned by snapping to grid and Iabeled using type tool. The image was flattened and saved in Targa format
(.tg&). The resulted isosbestic plot was converted fiom raster to vector graphics
(Encapsulated Postscript - .eps) using Adobe Streamline 3 .O (Figure 16).
2.4.5.8. Deoendance of Quantum Yield of Cyanation of HCB on the Initial Concentration of HCB
Six 8 mm Pyrex@ tubes containing 1.0 mL solutions of HCB in acetonitde (5% water) with concentrations ranging fiom 2.56~10~mol L" to 2-41x 10-~ mol L*' and
7.70~10"mol L-l NaCN were placed in the merry-go-round photoreactor and irradiated at 300 nm. Aliquots (10 pL) were taken fkom each tube after 12, 20 and 30 min periods and subjected to analysis by HPLC (Table 14). The resulting double reciprocal function of [&J vs. WB]was linear for al1 three data sets with ? ranging from 0.987 to 0.933.
2.4.5.9. Photocvanation of HCB on Preparative Scale
In a typicd experiment 100 mg (0.35 mmol) of HCB and 450 mg (9.2 mmoi) of
NaCN were dissolved in 250 mL of acetonitrile, containing 5% water using a mode1 FS-3 sonicator (Fisher). The solution was ptaced into four 25 mm 0.d. Pyrex@ test tubes and irradiated for 8 h at 300 nm in a rnerry-go-round. The combined solutions were extracteci with hexane (3 x 70 mL) and stirred in a well-ventilated fume cabinet at room temperature in the presence of of 5.6 g of Dowex 50W (50x4-100,4.8 mmoVg Meg equivalent, Sigma).
The solution was stirred for an additional 30 min at 50°C to remove dissolved hydrogen cyanide: Upon cooling to rmm temperature, the resin was removed by vacuum filtration.
The solution was dried overnight over activated 4A molecula.sieves. Mer drying and Table 14. Dependence of Quantum YieId of Cyanation of HCB on the Initial Concentration of HCB WCB],,, molL" mol^-' Conversion, %
12 min Irradiation:
20 min Irradiation:
- 30 min bdiation
2.97~10-~ 2.19~lo5 91 0.22 Absorbante of HCB in solutions ranged from 0.06 to 0.0006 and followed the Beer's Law. filtering, acetonitnle was removed on a rotary evaporatoq the residue was taken up on 5 g
of silica gel (60-200 mesh), which was then placed ont0 a bed of pre-equilibrated silica gel
and eluted with ethyl acetate-acetonitrile mixture (411). Collected hctions (typically 15
fiactions of 10-12 mL) were subjected to analysis by TLC and combined, if similar. The
purity of isolated products was checked by ES-MS. The hexane extracts were shown to
contain only unreacted HCB with a typical recovery of ca 40 mg of the starkg matenal.
The products 27 and 28 were isolated with 60% raw yield based on reacted HCB.
2J,4,5,6-Pen tacyanop henol (not isolated). Mass spectmm, m/z (%) 2 1 8 (M-H) (1 OO), 2 19
(13.91, 220 (1 -9); MS/MS, Daughters of m/r 218: dz(%) 190 (46), 164 (5. l), 138 (2.6), 114
(46.2).
4-Chloro-2,3,5,6-tetracyanophenol(27) (sodium salt). Approximately 20 mg (3 8% yield,
based on reacted HCB) was obtained with 93% purity (ES-MS) as a yellow solid tuniing
fluorescent-green in solution, m.p. 38 1-3 84OC (decornp.), 13c NMR (CbCN): 6 113-25,
114.08, 114.66*, 115.50*, 115.83, 11612*, 117.85, 122.19*, 122.48, 171.46 (*- irnpurity
fiom 28); Mass spectnim, m/r (%) 227 (M-l3) (100), 228 (12), 229 (34.1), 230 (4.2).
MS/MS, Daughters of m/z 227 (M-H): m/z (%) 199 (16.7), 192 (1 5.2), 164 (49.9), 15 1 (1.9),
138 (9.6), 123 (6.6).
X-Rqy CrystaIZogrqhy.The crystals were obtained b y slow evaporation of saturated
solution of 27 in propanoVwater (98/2), kept at room temperature. Some details of the X-
ray ana~~sis"~are given in Table 15. Table 15. Summary of Crystnl Data, Data Collection, Shucture Solution and Refuement Details
(a) Crystd Data formula 2~a+,2(~&m40-),3 -5(H20)$3H80 molar rnass 624.3 1 color, habit yellow, plate crystal size, mm 0.42 x 0.42 x 0.05 crystal system monoclinic a, A 12.6884(17) 44.369(3) 1l.3289(17) 90 110.567(14) 90 597 1.3(13) space group C2/m z 8 F(OO0) 2536 1.389 0.299
@) Data acquisition11s temp, K unit-cell reflcns (&range0) 1, 0.9975
(c) Structure Solution and ~efinernent~ refinement on ~2 solution method direct methods H-atom treatment C-Hriding, O-H not allowed for no. of variables in L.S. 426 weights: k in w = l/(&~c? + k) (O. 1429P)Z [P = @ + 2~c~)/3] R &gof 0.080,0.212,0.97 density range in f~ial-map, e AJ -0.300,0.40 1 final shiftlerror ratio 0.00 1
A dichlorotricyanophenol (28) (sodium salt). Approximately 15 mg (25% yield) was obtained with 90% purity (ES-MS), as a yellow solid, stays yellow in solution, m.p.192- 194°C(decomp.), 13cNMR (Cam:6 107.69, 114.10 (impurity fiom 27), 114-60, 115-78,
116.13, 119.29, 122.19, 131.81, 141.50, 171.02; Mass spectrum, m/z (%) 236 (hd-H) (100),
237 (12.4), 238 (63.5), 239 (73,240 (1 1.1); MS/MS, Daughters of dz236 (M-H): dz(%)
208 (6.5), 201 (6.2), 173 (16.5), 166 (16.1), 138 (54.7).
2.4.6. Svnthesis of 2.3.4.5.6-pentachlorobenzonitrile
2,3,4,5,6-Pentachlorobenzonitnle was obtained by cyanation of
pentachloroiodobenzene, the latter was synthesised by iododeamination of 4-
chloroaniline followed by polychlorination of 1-iodo-4-chlorobenzene.
2.4.6.1. Iodination of 4-chloroaniline
The procedure was adapted ffom Vogel. Il6 To a mixture of 2.76 g (2.15~10'~ mol) of Cchloroaniline (Aldrich), 5.5 g of concentrated hydrochloric acid and 5.5 g of water, cooled in ice-water bath to 1°c was slowly added a cold solution of sodium nitrite
(1.6 g, 2.33~10"mol) in 4.0 rnL of water. The mixture was stired for 5 min and 3.6 g
(2.17x10"mol) of potassium iodide in 4.0 rnL of water was added dropwise. The reaction mixture was left overnight, then heated for Ih at 60'~. The mixture was steam-distilled.
Yield 5.02 g (97%), rn.p.~3~~(lit.l16 55'~)~purity 98% (GCMS).
2.4.6.2. Polvchlorination of 1-iodo4chlorobenzene
The procedure was adopted from Crarnlyn and Gronjc. Il7 One gram (4.2 mmol) of l-iodo-4-chlorobenzene dissolved in 1.5 mL of chlorosulfonic acid was heated to 60'~ for 24 h. The progress of reaction was monitored by GCMS. Mer cooling, the reaction mixture was washed with 10% sodium bicarbonate and the product was isolated by flash chromatography with hexandchloroform (5: 1) as an eluent. Yield 0.63 g (40%) of pale- yellow crystals, m.p. 53 OC mp 53-54 OC), purity 98% (GCMS).
96 2.4.6.3 .Cvanation of Pentachloroiodobenzene
A mixture of 0.28g (7.4~10~ mol) of pentachloroiodobenzene and 0.07 g (7.6~1o4
mol) of aiprous cyanide (Fisher) in 2.0 mL of HMPA was stirred under argon at 80'~ for 3
h. The reaction mixture was allowed to cool to room temperature and then was poured on
8.0 rnL of a saturated solution of FeC13 in water. The resulting precipitate was vacuum
filtered and dissolved in 10.0 rnL of CH2C12.The solution was washed with dilute sodium
bisdphate then water, and dried under anhydrous sodium sulphate. The product was isolated
by flash' chromatography using hexane as an eluent to recuver the starting material (30 mg,
90% conversion), followed by hexanehethylene chloride (4: 1) to obtain 110 mg (54%) of
2,3,4,5,6-pentachlorobenzonitrile (95% pure by GC/MS) as pale-yellow crystals, m.p. 209
OC (lit. mp 2 10 OC), Mass spectrum 0,m/z (%) 279 (21), 277 (67),275(100), 273 (43),
240 (1 8), 23 8(15), 205 (8), 203 (8) 168 (7), 133 (20), 118 (1 O), 109 (1 1).
2.4.7. Reactions of 2.3.4.5.6-Pentachlorobenzonitrile with Sodium Cvanide
To 1.0 rnL of a 2.40~10-~ mol L-'solution of 2,3,4,5,6-pentachlorobemonitrile in
acetonitrile was added 1.0 rnL of a 0.125 mol L-' solution of NaCN in
acetonitrile/lO%water. The 8 mm Pyrex0 tube was wrapped in aluminium foil and
placed in a merry-go-round photoreactor dong with the control sample, made up of
1.0 mL of a 5.49~10~mol L-' solution of 2,3,4,5,6-pentachlorobenzonitrile in acetonitrile
and 1.0 mL of a 0.125 mol L*' solution of NaCN in acetonitrile/lO%water but not foil- wrapped. During the course of reaction 10 pL aliquots were taken fiom each sample at various periods of time and analyzed by HPLC'. Concentration of 2,3,4,5,6-
- The mobile phase was pure acetonitrile. The retention time of 2,3,4,5,6-pentachlorobenzon.it1iiewas 1.82 minutes, Calïbraton niive: [C7HClsNO], mol L-'= Peak Area x 3.63~10*'~,h.999. 97 pentachlorobenzonitnle in solution, percent conversion and In(cJcJ values were calculated (Table 16, c$ Figure 19).
Table 16. Reactions of 2,3,4$,d-PentachIorobenzonitrile with Sodium Cyanide
Non-lrradiated SampIe
Tirne,sec O 900 1800 3600 7200 10800
Peak Area 3.26% 106 3.46~10~ 3.24~106 3.28~106 3.16~106 3.43~106
Concentration, mol L-' 1.19~10" 1.26~10'~ 1.18~10'~ 1.19~105 1.15~10" 1.25~loJ
Irradiated Sample
Tirne, sec Concentration, mol L-' %Conversion wcdcd
2.4.8. Calculation of Ouantum Yield of Disappearance of Pentachlorobenzonitrile
The extinction coefficients at 313 nm and the rernaining concentrations at ~300 sec for HCB and pentachlorobenzonitrile were 197 and 1015 L mol-' cm-', and 2.36~1o4 and 1-71x10-' mol L-' respectively. The quantum yield of disappearance of pentacyanobenzonitrile was calculated using a custom transform fùnction created in
SigmaPlot 4.0: CrH=CoH-CtH
ClS=col(4) cts=coi(s)
CrS=ClS-cts
eH=col(6)
eS=col(7)
al=CrH-(?/eH)*LdifH
Ldif H=ln((l -exp(-2.303*eH*CtH))/(i -exp(-2.303*eH*CoH)))
LdifS=ln((l -exp(-2.303*eS*CtS))/(1-exp(-2.303*eS*CI S)))
- b2=CrS-(l/eS)*LdifS
coI(8)=O.l4*bZal where CH and C,S are concentrations of HCB and pentachlorobenzonitrile at /=O and
300 seconds and eH and es are the extinction coefficients of the above compounds.
2.4.9. Svnthesis and Ground-State Reactions of 1.4-Dichloro-2.356-Tetracvanobenzene and a Trichloro-Tricvanobenzene
To a mixture of dichlorotricyanophenol28 (3.1 mg) and phosphorus oxychioride
(0.1 mL) at room temperature was added 14 p.L of fieshly distilled pyridine. The reaction mixture was heated at 80°C for 7 min, cooled to 25°C and poured ont0 icdwater mixture
(2.0 mL). Upon standing (45 min) a precipitate was formed, and was filtered under vacuum, washed with a small amount of cold water and dried. The product was identified as a trichlorotricyanobenzene by GC-MS,it appeared as white powder, mp 287 OC. Yield
1.8 mg. Mass spectnim FI),m/z (%): 259 (3 l), 257 (100) 255 (84), 220 (6), 124 (12), 111
(9), 109 (14). Chlorination of 1.8 mg of 27 by the same procedure yielded ca. 0.8 mg of
1,4-dichloro-2,3,5,6-tetracyano benzene 29 as colorless crystals, mp 325 OC (lit.120 mp 328 OC). Mass spectrum (EI), m/r (%): 248 (661, 246 (1 OO), 185 (6), 135 (6), 124 (14), 11 1
(6), 109 (10).
The chlorination products were dissolved in acetonitrile (10.0 mL) and 1.0 mL aliquots were mixed with 1.0 rnL solutions of NaCN in acetonitddwater (0.262 M,
10%H20), NaOH in water (2.5%), and deionized water respectively at room temperature.
The progress of reactions was monitored by HPLC and the mixtures were subjected to
LC-MS analysis after 1 h.
24-10. ' Competitive Photocvanation of HCB. Pentachiorobenzene and 1.2.3.4-
Tetrachlorobenzene
Concentrations of stock solutions of HCB, pentachlorobenzene and 1,2,3,4- tetrachlorobenzene in acetonitrile were adjusted for equal absorbance at 300 nm
(A~M)= 0.017). For each compound duplicate samples were prepared by mixing 1.0 mL of the corresponding stock solution and 1.0 mL of 0.125 mol L-~solution of NaCN in
CH3CN/lO%H20 in 8 mm Pyrex@ tubes. The tubes were sealed with mbber septum and placed in the merry-go-round photoreactor. The solutions were irradiated at 300 nm; at set intervals 10 pL samples were taken for analysis by HPLC (Table 17, cf: Figure 24).
Table 17. Competitive Photocyanation of HCB, Pentachlorobenzene and l,2,3,4-Tetrachlorobenzene HCB
- Time, sec Concentration, mol L" Table 1 7 Continued
Tirne, sec Concentration, mol L-' Wonversion wcdc3
- - Time, sec Concentration, mol L"
2.4.1 1. Photocvanation of 1.4-Dicvanobenzene and 1.2.4.5-Tetracvanobenzene
The solutions were prepared by addition of 1 .O rnL of a 0.125 mol L" solution of
NaCN in acetonitnIe/IO%water to 1.O rnL of 5.49~1 o4 mol L-' acetonitrile solution of a corresponding cyanobenzene in 8 mm Pyrex@ tube. Two solutions were prepared for each cyanobenzene; with one solution mixed in Pyrex@ tube wrapped in aluminum foi1 and irnmediately subjected to HPLC anaiysis. The other solution was placed in a merry- go-round photoreactor and irradiated at 300 nm. Dunng the irradiation 10 pL aliquots were taken fiom each sample at various periods of time and analyzed by HPLC.
Concentrations in solution, percent conversion and In(cJcc) values were calculated for
1,Cdicyanobenzene (Table 18) and 1,2,4,5-tetracyanobenzene (Table 19).
Table 18. Photochernical and Ground State Reactions of 1,J-Dicyanobenzene with NaCN
Photocyanation
Tirne, sec Concentration, mol L-' O/oConversion fn@dc d
Ground State Reaction
The, sec Concentration mol L-' Table 19. Photochernical and Ground State Reactions of 1,2,4,!!LTetracyanobenzene with NaCN
Photocyanation
-- Time, sec Concentration, mol L-' %Conversion @O&)
Ground State
2.4.12. Photocvanation of Decachlorobiphenvl
2.4.12.1. Photolysis of Decachlorobiohenyl in the Presence of Sodium Cvanide for
Various Periods of Time
Four 8 mm Pyrex0 tubes were placed in a merry-go-round photoreactor and irradiated at 300 nrn. Each tube contained 1.O0 mL of a 1.92~10~mol L-' s01ution of
DCB in acetonitrile; 1.O0 mL of a 2.851x10-' mol L-' solution of NaCN in acetonitrile- water (5%v/v HtO) was added to two sample tubes while 1.00 ml of acetonitrile-water
(5%v/v H20) was added to the other two. During the course of reaction 10 pL aliquots were taken at various periods of time and analyzed by HPLC+ concentration of DCB in solution, percent conversion and
I.(c& values were calculated (Table 20, cj Figure 28).
Tablc 20. Pbotolysis of DCB in the Presence of Sodium Cyanide for Various Periods of Time
Concentration, mol L-' Wonversion (&'d Time, sec CN No CET CN No CN CET No CN
2.4.12.2. Measurement of Ouantum Yield of Disa~~earanceof Decachlorobiphenyl
The quantum yield of disappearance of DCB (+&) was measured by the method described in 2.2.1. The amount of reacted DCB was detennined at five different time intervals using quadruplicate samples. Volume of aliquots was 1.0 rnL. The initial concentration of DCB was 2.40~ 10.' mol L-', concentration of NaCN was 0.142 mol L-', the solvent was acetonitrile with 5.0% water (vlv). Absorbante of the 2.40~10"mol L-' solution of DCB in acetonitnldwater (97.5:2.5) at 3 13 nm was 0.0 1. The results of the experiment (average) are presented in Table 21.
The mobile phase was pure acetoniûile. Retention time of DCB was 4.46 minutes, retention time of products was 1.06 minutes. Calîration cwe: [DCB],mol L.' = Peak Area x 4.46~10-'~,h.998. 104 Table 21. Measurement of Quantum Yield of Disappearance of Decachlorobiphenyl
The, min Conversion, % c,.-& mol IA mol #h
2.4.12.3- Ouenchinn of Cvanation of DCB bv Ferrocene
The experimental setup was similar to that described in Section 2.4.5.4. Initial concentration of DCB was 2.54~lo5 mol L-', the concentration of NaCN was 0.251 mol L", the solvent was acetonitrile/2.5%water (v/v). Eight sample tubes, each containing 1.0 mL aliquots with various concentrations of ferrocene (absorbance of ferrocene at maximum concentration was less than 0.1 at 300 nm) were irradiated for 4 min in a merry-go-round photoreactor and analyzed by HPLC with pure acetonitrile as a mobile phase. The acquired averaged data (duplicate) are presented in Table 22.
Table 22. Quenching of Cyanation of DCB by Ferrocene
[Ferrocene], mol L" Conversion, %- Llobmol M&2 2.4.12.4, Photocvanation of DCB at Various Concentrations of Water
Six 8 mm Pyrex0 tubes containing 1.0 mL aliquots of acetonitrile solutions having constant concentration of DCB and NaCN but various concentrations of water were placed in the rnerry-go-round photoreactor and irradiated at 300 nm for 5 min. The initial concentration of DCB was 2.63~IO-' mol L-', the concentration of NaCN was
2.55~10"mol L-'. After irradiation samples were analyzed by KPLC (Table 23, c$
Figure 32). The function of percent conversion vs. percent water was linear, with a negative dope of -1.076 (2 = 0.954).
Table 23: Photocyanation of DCB at Various Concentrations of Water
Sample # [water], 5% bmol Conversion, %
2.4.12.5. Photolysis of Decachlorobiphenvl at Various Concentrations of Sodium
Cyanide
Eight 8 mm Pyrex0 tubes containing 1.0 mL aliquots of 2.5 1x10" mol solutions of DCB in acetonitrilehater (95/5 vlv) and various concentrations of
NaCN were placed in the merry-go-round photoreactor and irradiated at 300 nm for 8 min. After irradiation samples were analyzed by HPLC (Table 24, cJ Figure 3 1). The function of the reciprocal of quantum yield of disappearance of DCB (hi,) vs the
reciprocal of WaCW was linear, with a slope of 0.117 mol-' L (2 = 0.978).
Table 24. Photolysis of Decachlorobiphenyl at Various Concentrations of Sodium Cyanide
2.4.12.6. Sensitization of DCB Photocvanation by Acetone. Acetophenone and Oxv~en
Removal
Concentrations of acetone and acetophenone were chosen such that most of the incident light waç absorbed by the sensitizers. Eight solutions of DCB (2.47~10'~mol L") in acetonltnlehvater (5% &O), containing NaCN (6.26~10-2 mol L-') were prepared, with two having acetone (0.125 mol L-', A3rn= 0.78) and the other two having acetophenone
(1.12~1O-' mol L-', A3~=0.57) added. Four samples containing acetondacetophenone and two without sensitizers were degassed using four freeze-pump-thaw cycles and sealed under vacuum. The duplicate samples were irradiated at 300 nm for 7 min and analysed by HPLC
(Table 25) Table 25. Sensitizatîon of DCB Photocyanation by Acetone, Acetophenone and Oxygen Removal Sample: Aerated Degassed Acetone Acetophenone
hmol L' 5.49~10% 7.2~IO-' 1.38~~o-~I 2.8~10~ 1.90~ IO"* 8.28~IO-' 2.35~10"+1.31~ 10"
Conversion, % 22- 56I1 76s 95I1
4440 1 .O0 2.5 1 3.45 4.28
A value of w40 is equal taSiad/ c'~where S and O represent samP1& with and without sensitizer
2.4.12.7. Cornpetitive Photolvsis of Three Hexachlorobiphenyls
Solutions of 2,2',4,4',6,6'-hexachlorobiphenyl(35), 2,2',4,4',5,5 '-
hexachlorobiphenyl(36) and 3,3 ',4,4',5,5 '-hexachlorobiphenyl(37) in acetonitrile
(2.00~10~mol L") were prepared using an ultrasonic bath. A 1.00 mL aliquot was taken
from each solution and placed in 8 mm Pyrex@ tube, containing 1.00 mL of the 0.125
mol L-' NaCN solution in acetonitnle/lO%water. Duplicate samples were photolysed at
300 nm in the meny-go-round photoreactor. At various time intervals 10 pL aliquots
were taken nom each sample and subjected to HPLC analysis (Table 26).
Table 261 Cornpetitive Photolysis of Three Hexachlorobiphenyls in the Presence of NaCN
Lbmol L*' Conversion, %
Tirne, sec 35 36 37 35 36 37 35 36 37 Pseudo-first order rate constants of photocyanation for hexachlorobiphenyls were
determined fiom plots In(CJCt) vs. time. They were 2-53x lo4 s-' (h.961),
4.28~10 4 s-1 (?=0.963) and 2.3 lx 10-4 s 1(?=0.966) for 35,36 and 37 respectivefy.
The results of the control experiment are presented in Table 27. Al1 the conditions
were kept as above with the exception that 1.O mL of acetonitrile/lO%water was added to
the biphenyl stock solutions in place of the NaCN solution.
Table 27. Cornpetitive Photoiysis of Three Herachiorobiphenyls in the Absence of NaCN
- - & mol L-' Conversion, %
Time, sec 35 36 37 35 36 37 35 36 37
First order rate constants of photocyanation for hexachlorobiphenyls were
determined nom pIots h(CJCt) vs. time. They were 2.73~10~~s-' (?=0.880),
5 1 4.47~10' s (?=0.97 1) and 2. Mx 1U5 S-' (?=0.976) for 35, 36 and 37 respectively.
The absorbantes of 1.0~10~mol L-' solutions of 35, 36, and 37 in acetonittile at
300 nm were 0.005, 0.018 and 0.087 respectively. The relative quantum yields of disappearance for the above compounds were calculated using a custom transfonn function created in SigmaPlot 4.0: CtS=wl(S) CrS=Cl S-CtS eH=col(G) eS=wl(7) a 1=CrH-(1 1eH)'Ldifi-l LdifH=ln((l-exp(-2.303*eH*CtH))/(1 -exp(-2.303*eH*CoH))) LdifS=ln((l-exp(-2.303*eS*CtS))/(1 -exp(-2.303*eS*C1 S))) bP=CrS-(1/eS)*LdifS co1(8)=b2/al
where CiH and C,S are the concentrations of 35 and 36 (or 37) at /=O and 1500 seconds
and eH and es are the extinction coeEcients of the above compounds.
2.4.12.8. Photolysis of 2.3.4.5.6-Pentachlorobi~henvl in the Presence of Sodium Cvanide for Various Periods of Time
The initial concentration of 2,3,4,5,6-pentachlorobiphenyl was 5.58~lo4 mol L", the concentration of NaCN was 0.125 mol L-', the solvent was acetonitrile with
5% water. During the course of reaction 10 pL aliquots were taken at various periods of time and analyzed by HPLC? The concentration of 2,3,4,5,6-pentachlorobiphenyl in solution, percent conversion and In(cJcJ values were calculated (Table 28).
Table 28. Photolysis of 2,3,4,5,6-Pentachlorobiphenyi in the Presence of Sodium Cyanide for Various periods of Time
Time,sec Concentration, mol L-' Wonversion Wdcd
The mobile phase was pure acetonihile. The retention time of 2,3,4,5,6-pentachlorobiphenyi was 2.25 minutes. Calibration me:[2,3,4,5,6-pentac~orobiphenyl], mol L-'= Peak Area x 5.85~10-'O, h.984. 110 2.4.12.9. Large-Scale Photocvanation of 2.3.4.5.6-Pentachlorobiphenvl
Forty one milligrams (1.25~10~mol) of 2,3,4,5,6-pentachlorobiphenyl
(Accustandard Inc, 99% pure) and 0.5 g (10 mmol) of NaCN were dissolved in 50.0 rnL of acetonitrile/5%water. The solution was placed in two 30 mm Pyrex@ tubes, sealed with rubber septa and irradiated at 300nm for 24 h. The resultant yellow liquid was treated with 3 -2 g Dowexm 50W and extracted with 3x20mL hexane. Approximately 30 mg of starting material, which was recovered fiom hexane, was redissolved in 50 mL of acetonitrile/5%water with 0.5 g of NaCN and irradiated for additional 48 h. Mer treatment with Dowexa 50W and hexane extraction 18 mg of the starting pentachlorobiphenyl was recovered. The combined acetonitrile layer was reduced in volume to ca 1 mL and then placed ont0 a bed of pre-equilibrated silica gel column and eluted with ethyl acetate- acetonitrile mixture (4:l). Collected fractions (13 fiactions of 20 mL) were subjected to analysis by TLC and combined, if similar. The material balance is presented in Table 29.
Tetracyanochlorohydroxybiphenyl 41. Yield 59% (based on reacted material), mass spectrum, dz(%) 303 (h4-H) (100), 266 (10.5), 241(9.5) 224 (5.9); MSMS, daughters of
Table 29- Material Balance of Large-Scale Photolysis of 2J,d J,6-Pentachlorobiphenyl
Fractions 1-3 12.5 mg m/z 303,95%pure by ES-MS 4.12~10" mol
Fractions 4-7 1.5 mg rn/z 303:m/z337=1:1 ca. 5. lx los mol
Fractions 8-13 2.0 mg Recovered 18 mg 5.51x10%1ol rn/z303:m/z337=1:10 ca. 7.8x10~mol
23 mg, Total: 5.03~10' mol 7.0~10" mol
I I 1 Materiai baiance accounts for 73% of reacted 2,3,4,5,6-pentachlorobiphenyl. dz303: m/s (%) 293 (142), 213(29), 201(14) 188 (26), 163 (10.5), 35 (56). 'H NMR: (6,
CRCN), 7.41, dd, J=7.4 Hz, 4~=2Hz, 1H (IB'); 7.49, ddd, F7.4 Hz,b7.6 Hz, 4~=1.4Hz,
1H (H43; 7.52, ddd, J=7.6 Hiq J=8.0 & 4~=2Hq 1H (H53;7.60, d4 J=8 Hz, 4~=2Hi, 1H
(H6'). "C NMR (partial spectmm): (6, CD3CN), 125.7 (C27, 128.6 (C43 13 1-1(C6r)7
131.4 (C3'), 132.6 (C5'), 151.3 (Cl').
2.4.12.10. Preparation of 3-Hvdroxv-2.4. 6-Trichlorobi~henyl
In a 10 mL round-bottom flask equipped with stirring bar were placed 0.5 g
(2.94 mmol) of 3-hydroxybiphenyl (Aldrich), 1.2 g (9 rnmol) of N-chlorosuccinirnide
(Aldrich), 1.O rnL of glacial acetic acid (Fisher) and 1 drop of concentrated HCI. The mixhre was stirred for 10 h under argon, then neutralised with 1û% NazCQ+in water and extracted with 3x5 rnL CH2Cl2-The combined extract was dried under anhydrous sodium sulphate and reduced in volume to 1mL. The producî, 3-hydroxy-2, 4, 6-trichlorobiphenyl was isolated by flash chromatography (60-200 mesh silica-gel, mobile phase: hexane:CH~Cl2:MeOH= 49:49:2) in 63% yield (502 mg) as colourless solid, mp 72 OC.
Purity was 99% by GC-MS. 'H NMR (6, CDCl,), 7.35, m, 2H; 7.40, s, 1H; 7.45, ni, 3H. 13c
NMR (6, CDCb), 123.0, 127.1, 127.9, 128.1, 128.3, 130.6, 131.8, 1411' 174.52. (*- coinciding signal~fiom Cl and Cl '). Mass spectrum (EI): 276 (33), 274 (100), 272 (97),
237 (6), 209 (8), 207 (14), 202 (19), 175 (18), 173 (50), 138 (24).
2.4.12.11. Cornpetitive Photocyanation of 3-hvdroxv-2.4.6-trichlorobiohenvl and 4- hvdroxv-2',3 '.4 '5'-tetrachlorobiphenv1
The initial concentration of both hydroxybiphenyls was 5.64~104mol L-', the initial concentration of NaCN was 0.050 mol L-', solvent system was acetonitrile containing
2.5%water. The absorbantes of 3-hydroxy-2,4,6-trichlorobiphenyl and 4hydroxy- 2 ',3 ',4', 5 '-tetrachlorobipheny 1 solutions of the above-mentioned concentration at 3 00 nm were 0.122 and 0.258 respectively. Pyrex0 tubes (8 mm) containing 1 .O mL aliquots were photolysed at 300 MI for various time intervals; the progress of reactions was monitored by
HPLC (Table 30, cf: Figure 34)
Pseudo-first order rate constants for photocyanation of hydroxybiphenyls were determined hm plots In(CJCt) vs. time. They were 2.60~10~s-' (>f=0.965) and
1-70x 1o4 s*' (?=O -953) for 3-hydroxy-2,4,6-trïchiorobipheny l and 4-hydroxy-2 -,3r,4 ', 5 *- tetrachlorobiphenyl respectively.
Table 30. Cornpetitive Photocyanation of 3-hydroxy-2,4,6-trichlorobiphenyland J-hydroq-2'3'P'$'-tetrachlom bip henyl Concentration, mol L-' Conversion, % WYCt)
Time, sec 3-OH 4'-OH 3-OH 4'-OH 3-OH 4'-OH
2.4.13. Photocvanation of Octachloronaphthalene
2.4.13.1. Photolvsis of Octachloronaphthalene in the Presence of Sodium Cyanide for
Various Periods of Time
The initial concentration of OCN was 1.73~10~mol L*', the concentration of
NaCN was 0.125 mol L*', the solvent was acetonitrile containing 5% water. Duplicate samples of 1.00 mL volume along with two control samples with no cyanide added were
113 placed in a merry-go-round photoreactor and irradiated at 300 nrn. During the course of
reaction 10 @ aliquots were taken at various periods of time and analyzed by HPLC.'
The concentration of OCN in solution, percent conversion and Zn(cdcJ values were
calculated (Table 3 1, cj: Figure 3 8).
Table 31. Photolysis of OCN in the Presence of Sodium Cyanide for Various periods of Time
Concentration, mol L-' %Conversion WdiJ
Time, CN- No CN- No No CN sec CN CN- CN
2.4.13.2. Measurement of Ouanhim Yield of Disamarance of Octachloronaphthalene
The quantum yield of disappearance of OCN (#ds) was rneasured by the method
described in Section 2.4.5.1. The amount of reacted OCN was determined at five
different tirne intervals using tnplicate samples. Volume of aliquots was 1.0 mL. The
initial concentration of OCN was 1.53~10~mol L-', the concentration of NaCN was
0.142 mol L", the solvent was acetonitnle with 5.0 % water (v/v). Absorbance of the
The mobile phase was pure acetoniûile. Retention tirne of OCN was 3.87 minutes. retention times of products were 0.67 and 2.50 minutes. Calibration curve: [OCN], M = Peak Area x 9.16~10-' l, ?=U.998.
114 1.53 x 1o4 mol L-l solution of OCN in acetonitrildwater (97.5:2.5) at 3 13 nm was 0.86.
The results of the experiment (averaged using method of standard deviation) are presented in Table 32.
Table 32. Measurement of Quantum Yield of Disappearance of Octachloronaphthalene
Tirne, min Conversion, % kmol IL mol #di=
1133 60.1 9.18~10~18.5~ 10-' 6.93~1om4 1.3~10% 1.2~lo4 The hinction of #& vs. IA was iinear, with intercept of 2.04~lo4 and h.878.
2.4.13.3: Attempt to Ouench Cvanation of OCN with Ferrocene
The experimental setup was similar to that described in Section 2.4.5.4. The initial concentration of OCN was 3.17x10"mol L-', the concentration of NaCN was 0.136 mol L-', the solvent was acetonitrile with 2.5% water (v/v). Eight 8 mm Pyrex@ tubes, each containing 1.O mL aliquots with various concentrations of ferrocene (absorbance of ferrocene at maximum concentration was Iess than 0.1 at 300nm) were irradiated for 106 minutes in a meny-go-round photoreactor and analyzed by HPLC with pure acetonitrile as a mobile phase. The experiment was done in duplicate and acquired data were averaged (Table 3 3). Table 33. Attempt to Quench Cyanation of OCN by Ferrocene
[Ferrocene], mol L-' Conversion, % Crraccod, mol L-' W@Q
2.4.13.4. Photocvanation of OCN at Various Concentrations of Water
Four 8 mm Pyrex@ tubes containing 1.0 mL aliquots of acetonitde solutions
containing 3.17~1o4 mol L-' of OCN and 9.71x 1O-* mol L-' NaCN and various
concentrations of water were placed in the merry-go-round photoreactor and irradiated at
300 nm for 2 h. After irradiation, the samples were analyzed by HPLC (Table 34). The
function of percent conversion vs. percent water was linear, with a negative slope of
- 0.026 (2= 0.998).
Table 34. Photocyanation of OCN at Various Concentrations of Water
Sarnple # [water], % b,mol Conversion, % 2.4.13.5. Dependance of Quantum Yield of Cyanation of OCN on the Initial
Concentration of OCN
Five stock solutions of OCN in acetonitrile ranging in concentration from
3 -61 x lo4 to 1.Ob 10-'rnol L-' were prepared using a dilution senes. Aliquots of 0.5 mL
volume of each solution along with 0.5 mL of 0.142 rnolL1 solution of NaCN in
acetonitrile/lO%water were placed in 8 mm Pyrex8 tubes and capped with rubber septa.
The samples were placed in the merry-go-round photoreactor and irradiated at 300 nrn.
Aliquots (10 pL) were taken from each tube after 3 1 and 5 1 min periods and subjected to
analysis by HPLC (Table 35). The resulting double reciprocal function of [4di,] vs.
WB]was linear for both tirne intervals with 3 of 0.985 and 0.988.
Table 35. Dependence of Quantum Yield of Cyanation of OCN on the Initial Concentration of OCN
[OCwo, mol L" mol L-' Conversion, % @diy
3 1 min Irradiation:
5 1 min Irradiation: 2.4.14. Photocvanation of Octachorodibenzo-pdioxin (OCDD)
2.4.14.1. Photolvsis of (OCDD) in the Presence ofNaCN in AcetonitrilefWater
Two 8 mm Pyrex@ tubes, containhg 1.00 rnL of a 5.00~lo4 mol L-' solution of
OCDD in acetonitrile; 1.00 ml of a 2.00x10-' mol L-' solution of NaCN in acetonitrile- water (5%v/v HzO), were placed in the merry-go-round photoreactor and irradiated at
300 nm. Two additional samples having the same concentration of OCDD but no NaCN added were present as wntrols.
During the course of reaction 10 pL aliquots were taken at various periods of time and analyzed by HPLC." Concentration of OCDD in solution, percent conversion and
In@& J values were calculated (Table 36).
Table 36; Photolysis of OCDD in the Pmsence of Sodium Cyanide for Various Periods of Time
Concentration, mol L-' %Conversion hl @dcJ
- - The, sec CN
.- The mobile phase was pure acetonitrile. The retention time of OCDD was 3.12 minutes, retention time of products was 0.71 minutes. Calibration curve: [OCDD],mol L-' = Peak Area / 8.70~10~~h.998.
118 2.4.14.2. Measurement of Ouantum Yield of Disa~~earanceof OCDD
The quantum yield of disappearance of OCDD (#&) was measured by the
method described in Section 2.4.5.1. The amount of reacted OCDD was detennined at
five different time intervals using triplicate samples. Volume of aliquots was 1.0 mL. The
initial concentration of OCDD was 1.00x10" mol L", the concentration of NaCN was
0.040 mol L-', the solvent was acetonitrile with 5.0% water (v/v). The absorbance of the
1.00~ mol IO solution of OCDD in acetonitnldwater (97.5:2.5)at 3 13 nm was 0.013.
The results of the experiment (averaged using the method of standard deviation) are
presented in Table 37.
TabIe 37. Measurement of Quantum Yield of Disappearance of OCDD
Time, min Conversion, % G~cdmol I,,, mol #diu
2.4.14.3. Sensitization of Photocvanation of OCDD bv Acetone and Acetophenone
The experimental set-up was similar to that described in Section 2.4.5.3. The
concentrations of acetone and acetophenone in solution were 8.0x10-~mol ~~'(~300= 0.49)
and 1.12~1 mol L-' (A3m= 0.57) respectively. The initial concentration of OCDD was
4-03x1 o5 mol L", the concentration of NaCN was 4.00x10-~ mol L-', the solvent was
acetonitrile/water (5% &O). The samples were imdiated at 300 nm for 60 min and
analysed by HPLC (Table 3 8). Table 38. Sensitization of OCDD Photocyanation by Acetone and Acetophenone
Sample: Control Acetone Acetop henone
mol L'
Conversion, % 14 3 1 97
bd#o 1.00 1.78 5.58 A value of #S/#O is equd to Creacced. S / O,where S and O represent samples with and wiîhout sensiîizer
2.4.14.4. Photocyanation of OCDD at Various Concentrations of Water
Four 8 mm Pyrex@ tubes containing 1.O rnL aliquots of acetonitrile solutions
containing 4.00~1o4 mol L-' of OCDD and 1.60x10~~molL-' NaCN and various concentrations of water were placed in the merry-go-round photoreactor and irradiated at
300 nm for 3 h. After irradiation the samples were anafyzed by HPLC (Table 39).
Table 39. Photocyanation of OCDD at Various Concentrations of Water
- -- -
Sample # [water], % GaGtcbmol Conversion, %
2.4.14.5. Attempt to Ouench Photocvanation of OCDD by Ferrocene
The experimental setup was sirnilar to that described in Section 2.4.5.4. The initial concentration of OCDD was 4.82~10'mol L-', the concentration of NaCN was 0.040 mol L-', the solvent was acetonitrile with 5% water (v/v). Seven sample tubes, each containing 2.5 rnL aliquots with various concentrations of ferrocene (absorbance of ferrocene at maximum concentration was less than O. 1 at 300 nm) were irradiated for 180
120 minutes in a merry-go-round photoreactor and analyzed by KPLC with pure acetonitnle
as a mobile phase (Table 40, cJ: Figure 39). The function of vs. Ferrocene] was
linear, with a negative dope of -295 (?=0.936).
Table 40. Attempt to Quench Cyanation of OCDD by Ferrocene
Ferrocene], rno1~-' Conversion, 0/o Crrsncd> rnoLL-' hhb
2.4.15. Photolysis of Hexafluorobenzene in the Presence of NaCN at 254 nrn
Two 8 mm quartz tubes, containing 1.00 mL of a 1.80~10~~ mol L-* solution of
OCDD ' in acetonitrile and 1.00 ml of a 2.00~10-' mol L-' solution of NaCN in acetonitrile- water (5%v/v HzO), were placed in the merry-go-round photoreactor and irradiated at 254 nm. During the course of reaction 10 aliquots were taken at various periods of time and analyzed by HPLC.~Concentration of hexafluorobenzene in solution, percent conversion and h(c& J values were calculated (Table 41). The product
The rnobiie phase was pure a,toni,ie. ,retenticm time of C6& was 3.12 minutes, retention theof products was 0.91 minutes. The detector wavelength was 227 nm Calibration curve: [Ca, mol L" = Peak
Area * 1.029~10",+0.998. 121 mixture was concentrated and subjected to ES-MS analysis. Three parent ions were found in ES-MS spectrum (m/r (M-H): 218 (CllN50; 14.2 %), 214 (CIOF~N~O-,78.4 %), 204
Table 41, Photolysis of Hexafluorobenzene in the Presence of NaCN at 254 nm Tirne, sec Concentration, mol L" Wonversion ln (çdc3 Biblioma~hyfor Chapter TI
1. Cornelisse, J.; Havinga, E. Chem. Rev., 1975, 75, 353.
2. Havinga, E.; Cornelisse, J. Pure AppL Chem., 1976,47, 1.
3. Cantos, A.; Marquet, J.; Moreno-Maiias, M.; Castello, A. Tetrahedron, 1988, 14, 2607.
4. Gilbert, A.; Baggotî, J. Ersentiols of MoZemZur Photochemistry CRC Press, Boca
Raton, 199 1, p 386.
5. Cantos, A.; Marquet, J. ; Moreno-Maiias, M. Teirahedron Lett., 1987, 36, 4 195.
6. Van Eijk, M. J. A-; Huizer, H.; Varma, C. A.G.O.; Marquet, .J. J. Am. Chem. Soc., 1989,
111, 88.
7. Bunce, N.J.; Cater, SR-;Scaiano, J-C.; Johnston, L. J. J. Or,.Chem., 1987, 52, 42 14.
8. Mutai, K.; Yokoyama, K.; Kamo, S. Kobayashi, K. Bull. C'hem. Soc. Jpn., 1982,5.5,
1112.
9. Mutai, K.; Kobayashi, K.; Yokoyama, K. Tefrahedron, 1984, 40, 1755.
10. Mutai, K.; Nakagaki, R.; Takeda, H. Bull. Ch.Soc. Jpn., 1985,58, 2066.
1 1. Cantos, A.; Marquet, J.; Moreno-Mahas, M.;Gonzilez-Lafont, A.; Lluch, J.M.;
Bertriin, J. J. Org. Ch.,1990,55,3303.
12. Friedrich, K.; Wallenfels, K. In The Chemistry of the Cymo Group; Rappoport, 2.
Ed.; Interscience, London, 1970; p 67.
13. Kuzmic, P.; Soucek, M. Collect. Czech. Chem. Commun. 1987,52,980.
14. Bunce, N. J.; Stephenson, K. L. Cm. J. Chem. 1989, 67,220.
15. Lemmetyinen, H.; Konijnenberg, J.; Cornelisse, J.; Varma, C. A. G. O. J.
Photochem. 1985, 30, 3 15. 16. Niiranen, J.; Nieminen, K.; Lemmetyinen, H. J. Photochem. PhotobioL, A 1991, 56,
43.
17. Konuk, R-;Comelisse, T.; McGlynn, S. P. J. Chem. Phys., 1986,84, 6808.
18. Eberson, L.; Radner, F. Acta Chem. Scand, 1992,16,3 12.
19. Berci Filho, P.; Neumann, M. G.; Quina, F. H. J. Chem. Res., Synop., 1991, 70.
20. Den Heijer, J.; Shadid, O.B.;Cornelisse, J.; Havinga, E. Tetrahedron, 1977, 33, 779.
21 Eberson, L.; Radner, F. Acta Chem. Scand, 1992, 46, 3 12.
22 Berci Filho, P.; Neumann, M. G.;Quina, F. H. J. ChRes., Synop., 1991, 70.
23. Traber, R.; Kramer, H.E.A.;Knappe, W.; Hemmerich, P. Photochem. Photobiol.,
1981, 33, 807.
24. Goechner, N.A.; Snyder, H.R. J Org. Chem., 1973, 38, 48 1.
25. Colb, A.L. J. Am. Ch.Soc., 1979,101,3417.
26. Vink, J.A.J.; Verheijdt, P.L.; Cornelisse, J. Havinga, E. Terrahedron, 1972, 28, 5081.
27. Beugelmans, R.; Ginsburg, H.; Legas, A.; Le Goff, MX.; Roussi, G. Tetraheclon
Lett., 1978, 35, 327 1.
28. Yasuda, M.; Pac, C.; Sakurai, H.; J Chem. Soc. Perkin. Trmzs. 1,1981, 746.
29. Bunce, N.J.; Bergsma, J.P., J. Org. Chem., 1980, 45, 2083.
30. Bunce, N.J.; Bergsma, J.P.; Schmidt, T.L., J. Chem. Soc. Perkin Trans. IL, 1981, 713.
3 1. Lernmetyinen, H.; Koskikallio, J.; Niiranen, J.; Lilie, J. Acta Chem. Scand. A., 1984,
38, 765.
32. Lernmetyinen, H.; Koskikallio, J.; Lindblad, M.; Kuzmin, M.G. Acta Chem Scd
A., 1982, 36,391. 33. Suzuki, N.; Shimazu, K.; Ito, T.; Izawa, Y. J. Chem. Soc. CM.Commun., 1980,
1253.
34. Kavarnos, G.J.; Turro, NJ.; Chem. Rev., 1986, 86, 40 1.
35. Zuilhof, A.; van Gelderen, FA;Comelisse, J.; Lodder, G. J. Phys. Chem., 1998, 102,
5456.
36. Lemmetyinen, H.J. 1 ChSoc. Perkin Tram. 11,1983, 1269.
37. Letsinger, R.L.; McCain, J.H. J. Am. Chem. Soc., 1966, 88,2884.
38. Letsinger, RL.; McCain, J.H. J. Am. Ch.Soc., 1969, 91, 6425.
39. Letsinger, RL.; Hautala, R-R-;Tetrahedron Lett., 1969, 4205.
40. Nakagaki, R. Yuki Gosei Kagaku Kyokaishi, 1982,40,65 1; Chem. Abstr., 1982,
9 7,12661 Os.
41. Frolov, A.N.; Kulbitskaya, ON.; Eltsov, A.V. 1 Org. Chem. USSR, 1973, 9, 2335.
42. Murov, S.V. Hrn~dbookof Photochemistry Dekker, New York, 1973, p 34.
43. Muzino, K.; Pac, C.; Sakurai, H. J Chem. Soc., Chem. Commun.,1975, 553.
44. Lemmetyinen, H.; Koskikallio, J.; Ivanov, V.L.;Kuzmin, M.G. J. Photochem., 1983,
22, 11s.
45.Gemar0, A.; Maye, A.; Maran, E.; Vianello, E. Electrochim. Acta, 1988,33, 167.
46. Yildiz, A.; Sertel, M.; Gambert, A.; Baumgartel, H. Electrochim. Acta, 1988, 33, 169.
47. Sykes, P. A Guidebook to Mechanism in Organic Chemzsz?y 5th Ed.; London, 1981, p
18 1.
48. Suzuki, H.; Hanafusa, T. Synthesis, 1974,53. 49. Takagi, K.; Okamoto, T.; Sakakibara, Y.; Ohno, A.; Oka, S.; Hayama, N. Bull. Chem.
Soc. Jpx, 1975, 48, 3298.
50. Funabuki, T.; Nakamura, H.; Yoshida, S. J. Organomet. Chem., 1983,243,95.
5 1. Bdf, N.J.; Buxton, M.; Fuller, G. J Chem. Soc., 1965, 3372.
52. Banks, R.E.; Haszeldine, R. N.; Yong, Y.M. J. ChSoc. C, 1967, 2089.
53. Friedman, L.; Schechter, H. J. Org. Ch.,1961,26,2522.
54. Weis, C.D. J; Org. Chem., 1962,27,2964.
55. Krepelka, J.; Vancurova, 1.; Holubek, J. Col[. Czech. Chem. Commu~i.,1981, 46,
1523.
56. Ellis, J.P.; Hudson, H.V. J Ch.Res., Sm.,1985, 372.
57. Orshak, V.V.; Teplyakov, M.M.; Dmitrenko, A.V. Kakauridge, D.M. Polymer Sci.
USSR (Engl. Trad.), 1980,22,256.
58. Couture, C.; Paine, A.J. Cm. J. Chem., 1985, 63, 11 1.
59. Haken, B. Br. Paient 861 898, 1961, Chem. Abstr., 1961,55, 175812.
60. Pilgram, K.; Skiles, R.D. J Heterocycl. Chem., 1974, 11, 777.
6 1. Ellis, J.P.; Romney-Alexander, T.M. Chem. Rev., 1987,87, 779
62. Bacon, R.G.R.; Hill, H.A.O. J. Chem. Soc., 1964, 1097.
63. Ito, T. ; Wanabe, K. Bull. Chem. Soc. Jpn., 1968, 41,4 19.
64. Cassar, L.; Foa, M.; Montanari, F.; MarineIli, G.P. J. Organomel. Ch,1979, 173,
335.
65. Sakakibaba, Y., Yoshinori, 1.; Sasaki, K.; Uchino, N. Bull. Chem. Soc. Jpn., 1993, 66,
2776. 66. Grushin, V.V.; Alper, H. Chem. Rev. 1994,94, 1047.
67. Sasakibara, Y. Okuda, F.; Shimobayashi, A.; Kirino, K.; Sakai, M.; Uchino, N.;
Takagi, K. Bull. Ch.Soc. Jpa, 1988, 61, 1985.
68. Andersson, Y.; Langestorm, B. J. Chem. Soc., Perkin Tram I, 1994, 1395.
69. Takagi, K.; Sasaki, K.; Sasakibara, Y. Bzïll. Ch.Soc. Jpn., 1991, 64, 11 18.
70. Chatani, N.; Hanafusa, T. J. Org. Chem., 1986,51,4714.
71. Cassar, L.; Ferma, S.; Foa, M. "Advances in Chemistry 132, Homoge11e0u.s Cutu&sis
Il"; Gould, R.F. Ed.; American Chernical Society, Washington, DC, 1974, p 252.
72. Prochazka, M.; Siroky, M. Coll. Crech. Chem. Cornmutt, 1983, 48, 1765.
73. Okano, T.; Iwahara, M.; Jitsuo, K. Splett, 1998, 243.
74. Okano, T.; Kiji, J.; Toyooka, Y. Chem. Lett., 1998,425.
75. Amatore, C.; Saveant, J.M.; Combetlas, C., Robeveille, S-Thiebaunt, A. J.
Elec froanal. Chem., 1985, 184,25.
76. Yoshida, K. J. Am. Chem. Soc., 1979,101,21I6.
77. Smith, P. A. S.; Wirth, J.G.J. J. Org. Chem., 1968, 33, 1145.
78. Zergenyi, J.; Raz, B. Eur. Pat. Appl. 122 355, 1984; Chem. Abstr. 1985, 102, 62275s.
79. Friedrich, K.; Oeckl, S. Chem. Ber., 1970, 103, 395 1.
80. Hermann, K.; Simchen, G. Liebigs Ana Chem., 1981,333.
8 1. Sakai, K.; Ishida, 1. Jpn. Pat. 60 184 095, 1984, Chem. Absir. 1985, 104, 50689t.
82. Umemura, S.; Enomiya, T.;Shiraishi, Y.; Nakamura, T. Jpn. Pat. 79 145 63 3, 1978,
Chem. Abstr. 1980,92, 146462d.
83. Moore, H.W.; Chow, K.; Nguyen, N.V. J. Org. Chem., 1987,52,2530. 84. Chow, K.; Moore, XW.Tetrahedron Lett., 1987, 28, 50 13.
85. Wallenfels, K.; Witzler, F.; Friedrich, K. Tetrahedron, 1967, 23, 1845.
86. Friedrich, K.; Oeckl, S. Chem. Ber., 1973, 106, 3796.
87. Konecny, V.; Cesnekova, E. Czech. Pat. 137 623, 1968, Chern. Ak,1971, 75,
12% 152.
88. Konecny, V.; Drabek, J. Czech. Pat. 134 643, 1968, Chem. Abstr., 1971, 75,55032.
89. Konecny, V.; Drabek, J. Czech. Pat. 134 644, 1968, Ch.Absh-., 1971, 75,55 10z.
90. Suwa Seikosha Co. Jpn. Pat. 58 18 349, 1983, Ch.Abstr., 1983,99, 122045h.
91. Suwa Seikosha Co. Tpn. Pat. 58 32 858,1983, Ch.Abstr., 1983,99, 122048m.
92. Gray, G. W.; Lacey, D.; Toyne, K.J.; McDonnell, D.G. UK Pat. 890 8 1 02, 1989,
Ch.Abstr., 1991,112, 118469g.
93. Schulman, S.G.; Vincent, W.R.; Underberg, W.J.M. J. Phys. Chem., 1981,85, 4088.
94. Corsaro, A.; Buemi, G.; Chiacchio, U.; Pemni, G.; Pistara, V.; Romeo, R
Tetrahedron, 1996, 52, 78 85.
95. Konstantinov, A.; Kingsrnill, C. A.; Ferguson, G.; Bunce, N. 3, J. Arn.Chem. Soc.,
1998,120, 5464.
96. Silverstein, R.M.; Bassler, G.C;Momll, TC.,Spectrometnc Identzfiation of Organic
Compmndq Wiley, 1981, 4th Edition, p 22.
97. Al! calculations were done on a Silicon Graphics 4D-35TG cornputer system with the
NRCVAX system of programs (Gabe, E.J.; Le Page, Y.; Charland, J-P.; Lee F.L.;
White, J. Appl. Ctysf..,1989,22, 384.) for refinement with observed data on F, or with SHELXL-97 (G.M. Sheldrick, 1997, University of Gottengen, Germany) for
refinernent with al1 data on F~.
98. Friedrich, K.; Oeckl, S. Chem. Ber., 1973, 106,236 1.
99. Bunce, N.J.; Bergsma, J.P.; Bergsma, M.D.; DeGraaf, W.; Kumar, Y.; Ravanal, L.R-
J. Org. Ch.. 1980,45,3708.
100. Pseudo-first order kinetics are typical for a reaction with Iow light absorption; when
ail the incident light is absorbed by the substrate the reaction becomes a zero order.
10 1. Bunce, N. J. J. Photchem., 1987, 38, 99.
102 . CRC Handbook of Chemistry and Physics, Ed R.C. Weast, 64th Ed., CRC Press
Inc., Roca Baton, Fla., 1983-1984, p A-63. Entry 526.
103. Soumillion, J.P.; De Wolf, B. J. Chem. Soc. Chem. Commzm , 1981,436.
104. Murov, S.L.;Carmichael, 1.; Hug, G.L. Hmdbook ofPhofochemistry, 2nd Ed.,
MarceI Dekker, Inc. NY, 1993. Tabte 1, pp. 6-54.
105. Under the conditions of the negative ionisation the possible di-hydroxy produas
would fonn dianions with m/t corresponding (FW-2)/2. Carehl examination of the
ES-MS spectra revealed the absence of peaks in the range of = 220 daltons, which
would correspond to polychloro-polycyano-dihydroxybiphenyls.
106. Grimshaw, J.; deSilva, A.P. Chem Soc. Rev., 1981,10,181.
107. Bunce, N.I.; Kumar, Y.; Ravanal, L.; Safe, S. J. Ch.Soc. Perkin Trans. 111, 1978,
880.
108. Omura, K.; Matsuura, T. Tetrahehon, 1971,27,3 10 1. 109. Wiley, J.R; Chen, E.C.M.; Chen, E.S.D.; Richardson, R-;Reed, W.R.;
Wentworth, W.E. J. Electrmi. Chem., 1991,307,169.
110. Shadev, S.; Evans, D.H. J. Am. ChSoc., 1989, Ill, 2667.
Ill. Konstantinov, A.; Bunce, N. J. J. Photochem. Photobzol., A:Chem-, 1996, 94, 27.
112. Hatchard, C.G.; Parker, C.A. Proc. Roy. Soc. A., 1956, 236. 518.
113. Gould, LR.,in Handoook of Orgmic Photochemistry, Vol. 1, Ed By Scaiano, J-C.,
CRC Press, Boca Raton, FL, 2989, p. 47.
114. Murov, S.L.; Carmichael, 1.; Hug, G.L. Hmzd&ookof Photochemistry, 2nd Ed.,
Marcel Dekker, Inc. NY, 1993, p. 299.
115 Data collection on an Enraf Nonius CAD4 diffractometer with graphite
monochromatised Mo-Ka radiation (A 0.71 O67 A).
116 Vogel, 1. Practical Organic Chemisa 2nd Ed., Longmans, Green and Co.,
Toronto, 1951, p. 574.
117 lCramlyn, R.G.W.; Gronjc, J. Phosphorous and Sul*, 1979,6,494.
1 18. Aldrich C hemical Company, Aldrich Catalog-Handhok of Fine Chemicals,
Milwaukee, WI, 1984, p. 344.
119. Merz, W., Ber., 1888,20, 1627.
120. Bucsis, L., Friedrich, K., Chem. Ber., 1976, 109, 2462. CHAPTER III
DEVELOPMENT OF A SIMPLE METHOD FOR DESTRUCTION OF DIOXINS IN
LIQUID WASTE UNDER LABORATORY CONDITIONS AND IDENTIFICATION
OF THE PRODUCTS OF PHOTOLYSIS OF 2,3,7,8-TCDD
3.1. Introduction
The goals of this study were to develop a method for the destruction of dioxins in liquid laboratory waste using inexpensive and easily available equipment, and to identify the major products of photoiysis of 2,3,7,8-tetrachlorodibenzo-p-dioxin.
3.1.1. A Brief Review of the Dioxin Issue
Since early 1970's the issue of dioxins has been one of the hot topics for environmental activists, scientists and the general public. The best known representative of the dioxin family - 2,3,7,8-TCDD - was referred to as 'the most toxic substance on
Earth' and 'chernical AIOS'.' Public concem about dioxins is based on the findings of the severe toxic effects of a number of PCDDRCDF* congeners in animals2 and the occurrence of several accident^^*^*' which led to environmental contamination by
PCDD/PCDF. A large number of research papers on the toxicity of dioxins in animals has been published to date, including several At this time it is unclear whether or not dioxins are carcinogenic to humans," and although several acute and subacute symptoms in human beings were associated with the dioxin exposure by the US
Department of Health and Human Services," the oniy proven effect in humans is a skin rash called chl~racne.'~It is worth noting that 2,3,7,8-TCDD was first identified by German physician Karl Schultz in 1955, while treating workers eom a pesticide piant who had contracted chloracne. l3
Dioxins are formed as unavoidable by-products in the manufacture of chlorinated phenols and some other chlorinated chemicals, or as a result of incomplete combustion of chlorine-containing organic mixtures. Although present in trace amounts, dioxins have the potential to accumulate in both the environment and living organisms due to high lipophilicity, as well as remarkable chernical and thermal stability. l4
A recent study estimates that the annual global emission of total PCDDECDF from major industrial sources is 3000 kg/yr.I5 Environrnent Canada's dioxin inventory16 suggests that about 300 g PCDDPCDF were released to the Canadian atmosphere in
1997. In Canada, the most important potential anthropogenic sources of dioxins include municipal, medical and hazardous waste incinerators, cement kilns, steel plants, and wood combustion. Other sources include the dispersion of commercial products contaminated by PCDD and PCDF, such as certain pesticides and chlorinated solvents. l6
In April 1995, the FederaIProvincial Task Force on Dioxins and Furans (FTFDF) was est ab1ished by the Federal-Provincial Advisory Committee for CEP A (CEPA-FP AC) with a mandate to develop an inventory of sources of releases of PCDDRCDF in Canada.
According to the FTFDF inventory, a 15% reduction in atmospheric releases was observed between 1990 and 1997 in Canada due to up-grades or closures. Due to the adoption and implementation of new pulp and paper regulations, almost 99% reduction in that sector was achieved compared to the base year 1990, and releases to effluents fiom the pulp and paper sector have now been reduced to below the "measurable concentration" level. The total releases to mil from the ash resulting fiom the combustion of salt laden wood in B.C. and the disposa1 of treated wood have not changed between
1990 and 1998.16
The information available at this time indicates that the total release of
PCDDRCDF fiom chernical production in Canada amounts to 2.0 g TEQ/~.'~This
amount is expected to be reduced to 0.4 g TEQ/y by end of 1998, due to changes carried
out by Dow Canada in its Fort Saskatchewan plant. It should be noted that the dioxin-
containing liquid wastes generated by academic, industrial and governmental research
laboratories did not seem to be included in the inventory.
Although the total amount of dioxin-containing hazardous liquid waste generated
in Canada appears to be insignificant, the imperfections of the current regulations may
cause some serious problems in dealing with the dioxin containing waste for a waste
generator.
3.1.2. Lena1 As~ectsof Dioxins in Environment and Dioxin-containina Wastes
PCDDRCDF bearing two or more chlorines have been declared toxic and have been added to Schedule I (List of Toxic Substances) of the Canaciian Environmental
Protection Act (CEPA), based on the assessment published in 1990 by Environment
Canada and Health Canada. In 1992, Environment Canada adopted:
the Pulp and Paper Mill Effluent Chlorinated Dioxins and Furans Regulations, which
prohibited the release of measurable concentrations of 2,3,7,8-TCDD or 2,3,7,8-
TCDF.
the Pulp and Paper Mill Defoamer and Wood Chip Regulations, which prohibit the use
of defoamer containing more than 40 ppb of 2,3,7,8-dibenzofùran or 10 ppb of 2,3,7,8- dibenzo-p-dioxin, and wood chips that have been treated with polychlorinated
phenols.
In f 997 Environment Canada included PCDDRCDF in the list of 'Track 1' substances -
toxic substances that are persistent, bioaccumulative and anthropogenic. Virtual
elimination of Track 1 substances fiom the environment is one of the main objectives of
the Toxic Substances Management ~olicy.'~
The legal consequence of the regulations mentioned above is that Liquid waste
containing "measurable concentration'' of dioxins must be identified and declared as a
hazardous waste. The current legislative document, outlining the identification,
manifestation, transport and storage of hazardous waste in Ontario is Regulation 347; this document does not even mention dioxins.lg The "measurable concentration" is determined by the current detection limits of analytical instruments (most commonly -
GC-MS). The detection limits differ for various types of mass spectrorneters, these limits are also constantly improved with the progress in the analytical field. At the present time neither "measurable concentration'' of dioxins nor the detection limits are defined or regulated by Environment Canada. The waste disposal sites or the storage facilities for dioxin-containing wastes are not identified in current regulations. Commercial destruction of the dioxin-containing hazardous waste is also problematic since the technologies for the PCDDPCDF destruction are under development.
3.1-3. Emerging; Techno1og;ies for the Destruction of Dioxin Waste
Methods of destruction of dioxins and dioxin-containing waste are generally further f?om commercialization than PCB treatmentS4A variety of thermal and non- thermal methods were proposed within the last decade. 3.1.3.1. Thermal Methods
Incineration as the method for destruction of PCDDECDF is controversial. Since municipal waste and hospital incinerators have been identified as sources of dioxins, environmental groups such as Greenpeace strongly oppose any form of incineration. The largest sale atternpt to date was the 1984 incineration of over 2 million gallons of Agent
Orange. The herbicide was burned at sea, aboard the ship 'Vulcanus', in two large tùrnaces operating at 1500 OC with excess of air.20 Despite a combustion eficiency of over 99.9% and the srna11 amount of 2,3,7,8-TCDD involved (< 50 mg), traces of dioxin were still detectable after incineration. Current regulations in the United States require that incinerators achieve a destruction and removal efficiency @RE) of 99.9999 percent for wastes containing PCB and PCDDIPCDF. l3 The Apms incinerator in Coffeyville,
Kansas, is the only stationary incinerator in the United States to have received an EPA permit for dioxin incineration. Since 1993, about 1 million pounds of dioxin-containing waste has been destroyed with > 99.9999 percent DRE.
A number of promising dioxin treatment methods was presented in the 1995 SITE
Technologies Profiles Report by US EPA. (The Superfùnd Innovative Technology
Evaluation (SITE) Program evaluates prospective treatment, monitoring and measurement technologies for use in the cleanup of hazardous waste site^).^'
General Atomics' circulating bed cornbustor (CBC) uses high velocity air to entrain circulating solids and create a highly turbulent combustion zone capable of destroying PCDDFCDF contaminants in liquids, slumes, solids and sludges with
99.9999 DRE. The unit operates at lower temperature (750 - 900 OC) than conventional incinerators, with natural gas or diesel as auxiliary fuel, so the operating costs and potential ernissions of NOx and CO are reduced. Several 36 inch-diameter CBCs have been built and operated in Alaska, California and Germany. The treatment of dioxin- contaminated soils has been performed on a pilot plant scale, although at the Swanson
River project in Alaska, over 100,000 tons of PCB contarninated soi1 was remediated to the limits of PCB/PCDD/PCDF detection, far below allowable lirnitse2'
The gas-phase chemical reduction process, patented by Eco Logic (Rockwood,
Ontario), uses the reaction of hydrogen with organic and chlonnated organic compounds at 850 OC to produce a hydrocarbon reach gas. The post-reactor scmbber removes HCl, heat, water and particulate matter. The mobile unit, mounted on two standard tractor trailers, is equipped with a computerized process control system and in-line mass spectrometer, which is used to continually monitor selected organic compounds.
Although designed for treatment of PCB contaminated aqueous and oily waste strearns, soils and sludge, the process was tested using PCB contaminated wastes with low
PCDD/PCDF concentration by EPA in CO-operation with Environment Canada. A triplicate test resulted in net destruction of trace dioxidfuran compounds as well as a 'six nines' removal of PCB. Several commercial scale systems operate in Australia and
Canada, mostly treating PCB wastes2'
Catalytic hydrodechlonnation substitutes hydrogen in place of chlorine and forms hydrogen chloride as a byproduct in the presence Pd, Ru or Cu ~atal~st.~~The reaction with PCDD/PCDF takes place at 100 - 300 OC, depending on the catalyst. Copper catalyzed hydrogenation has been successfully applied at the laboratory scale for dioxin decomposition in a liquid laboratory wa~te.~~ In situ vitrification (IsV)~' can immobilize inorganics and destroy or remove
organics, including dioxins and PCB in contaminated soils, sediments or sludges. ISV uses electncity to melt soil at temperatures ranging fiom 1,600 to 2000 OC, destroying
organic pollutants by pyrolysis. The pyrolysis products are captured in a hood, which
draws the off-gases into a particulate removal treatment system. A typical setup may
encompass a total melt mass of 1,400 tons. The end result is a vitrified glass monolith, possessing high strength and excellent weathering and leaching properties. ISV treatment of soil contaminated with pesticides, metals, and low levels of dioxins at the Parsons
Chemical site (Michigan) for the SITE dernonstration was successful: the DRE limit was surpassed, and al1 air emissions and vitrified product sarnples had non-detectable levels of dioxins, furans and PCB.~' The ISV technology is currently being employed for treatment of dioxin, chlorophenol, herbicide and other organic-contaminated soil at the
Wasatch Chemical Superfund site in Salt Lake City, Utah.
3.1.3-2. Non-Thermal Methods
Non-thermal methods often have an advantage of lower fuel and energy consumption, and hence, lower operating costs. Most of them are based on the ability of a base to replace chlorine in the aromatic ring. A prospective non-thermal method for dioxin treatment involves dechlonnation with potassium graphite intercalate (C~K).~~
This was show to dechlorinate OCDD at 20 OC to dibenzo-p-dioxin in 45% yield within a minute. Excellent results were shown on a laboratory scale with dechlorination of
2,3,7,8-TCDD at 85 OC by nucleophilic replacement of arornatic chlorine by alkali polyalcohols or polyhydroxy ether~;~'decomposition was reported to be above 99.9%. A commercial process for destroying PCDDRCDF and PCB in solid waste by treatment with solvent and inorganic alkali bases at elevated temperatures is cutrently being developed by Trinity Environmental ~echnolo~ies.~~Currently the technology is able to treat individual chlonnated dioxin isomers and specific PCB congeners. In bench- scale studies this chernical process has reduced PCB concentrations fiom 2,000 pprn to less than 2 ppm in synthetically contaminated material in about 30 minutes using moderate power input.
A combined method, using thermal desorption and base-catdyzed decomposition, was developed by the companies ETG Environmental and SRS.'' The process is initiated in a medium temperature thermal desorber at 450 OC. Sodium bicarbonate is added to contarninated soils, sediments or sludges containing PCB and PCDD/PCDF.
Detoxification occurs by removing chlorine fiom contaminants and replacing it with hydrogen. The result is a clean, inexpensive, permanent remedy where al1 the residuals, including dechlorinted organics are recyclable or reco~erable.~'Levels of PCB and dioxins were reduced below detection limits during a Superfund site demonstration in
Morrisville, North Carolina, in 1993.
3.1.3.3. Photolysis of Dioxins
Clear advantages of photochernical methods for dioxin treatment over the methods discussed previously are minimal requirements in energy, added chemicals, and usually simple equipment. A few technological processes, employing UV light for the destruction of chlonnated organic contaminants, like PCB and dioxins, have been developed in the recent years. These will be descnbed after a brief review on the photochemistry of polychorinated dibenzo-p-dioxins. A number of studies have shown that PCDDPCDF appear to efficiently
photodegrade under both natural sunlight and artificial W radiation. Photolysis and
oxidation by hydroxyl radicais are proposed as the major pathways for the fate of the dioxins in the atrno~~here.~'~~Many laboratory studies have been performed to determine the photostability of the dioxins in organic solutions and absorbed ont0 solid substrates such as fly ash, soil, silica gel and glass. In solution, the photodecomposition of
PCDD/PCDF in the absence of additives has been shown to proceed by first order kinetics?' It is generally believed that the photolysis of PCDDPCDF in solution occurs by reductive dechlorination, giving less chlorinated products, which undergo further photodecomposition. A detailed discussion of the mechanism and products of photolysis of dioxins follows in Section 3.2.4.
Several studies have suggested that the rate of photolysis of PCDD increases with the decreasing degree of chlorination.3031.32 These studies compare the apparent rate constants for loss of PCDD, without taking into account differences in the amount of light absorbed by the various substrates (e-g Muto & ~akizawa~~).Data on the quantum yields of the disappearance of dioxins in aqueous acetonitrile (2:3 v/v) found in the literature do not support this trend, they indicate that there is no relation between the quantum efficiency of the photolysis and the number of chlorines on dioxins. Thus, the
&iss values for 1,2,3,7-tetra-, 1,3,6,8-tetra-, 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxins, and OCDD in aqueous acetonitrile at 3 13 nm were 5.4x104, 2.2~lo5, lSx lo5, and
2.3~10~resp.ectively. 34.35 The same authors have determined that the quantum yields of disappearance of 1,2,3,4,7-pentachloro-, and 1,2,3,4,7,8-hexachlordibenzo-p-dioxinin the similar reaction conditions were 9.8~10" and 1.1 x 1o4 re~~ectivel~.'~Higher quantum yieldç of disappearance were reported for 2,3,7-~r~D~,"2,3,7,8-~~~d~ and OCDD~~ in hexane, 0.20, 4.9~IO-', and 1.3~10" respectively. The decrease in k, for higher- chlonnated dioxins is apparent, however the data are insuscient to conclusively support the above-mentioned trend.
A number of researchers3439.40 came to the conclusion (again, based on relative rate constants) that dioxins bearing chlorine substituents on the lateral positions 2, 3, 7 and 8, are more susceptible to photodegradation than those with the apical chlorines on the positions 1, 4, 6, and 9 of a dibenzo-p-dioxin ring. Indeed, the q&s for 1,2,3,4-,
1,3,6,8-, and 2,3,7,8-TCDD, measured in 1,4-dioxane at 295 nm were in the order of
2.1 x 105, 3.0~10;' and 3.3~IO-*, with the highest value comesponding to the dioxin having only lateral ch lori ne^.^' The regiochemistry of dechlorination is important because it determines whether the photolysis of highly chlonnated congeners is likely to produce significant quantities of laterally substituted, congeners such as 2,3,7,8-TCDD, because these are recognized to be the most toxice2
3.1.3.4, Use of the UV Irradiation for Treatment of Dioxin-Containing Wastes
The ability of sunlight to degrade PCDD in the presence of hydrogen donors was employed in an attempt to destroy 2,3,7,8-TCDD in Seveso, a month after the accident.27
Hand application of a solvent emulsion to a tile roof caused complete disappearance of the TCDD in a full day's sunlight. Next, a 400 Llha spray application of a 40% aqueous olive oil to a highly contaminated field resulted in the removal of 90% of the TCDD within 9 days.
Large scale photolysis was used to decontaminate a former trichlorophenol manufacturing plant in Verona, ~issouri.~~Approximately 4300 gallons of waste were found to contain about 5 kg of 2,3,7,8-TCDD. The dioxin was extracted nom the aqueous
sludge with hexane and continuously degraded with 10 kW mercury arc lamps in the
presence of isopropyl alcohol. The plant-scale operation succeeded in reducing the
TCDD level by over 99.9%.
Several technologies employing W irradiation and applicable for dioxin waste
treatment were described in SITE A photocatalytic degradation process
developed at the State University of New York directly treats solids contaminated with
organochlorine cornpounds. A slurry fiorn soi1 and recycled water, formed in a 1,500 L
mixing tank is fed by gravity onto a series of inclined tables, coated with catalyst and
illuminated by W Iamps. The treated slurry is screened and dewatered as it falls off the
edge of the reactor surface; the solids are disposed of and the water is reused in the
mixing tank. The nature of the catalyst and the additive (10 ppm) used in the process are
not revealed in the SITE report,2' although it States that the organics are destroyed by
combination of direct photolysis, catalyst-induced radical production, and oxidant attack
initiated by the additive.
A two- stage, in situ photolytic and biological detoxication process for shallow
soil contamination is being developed by IT ~or~oration.~'In the first step, the organic
contaminants are degraded by means of a 450 W mercury lamp, with the soil tilled by a power tiller and sprayed with surfiictant. Biodegradation, the second step, further destroys organic contaminants and detoxifes the soil. A 20-hour test irradiation of contaminated sandy soils showed greater than 90% removal of both PCDD and PCB. A recent study demonstrates that photolysis of PCBs in the presence of small amounts of surfactant @rij 58) increases the quantum yield of dechiorination six times in cornparison with the
photolysis in water done."
Waste treatment processes based on photoinduced oxidation reactions of OH
radicals are used for destruction of a wide range of contarninants other than dioxins.
More than 80 industnal scale installations worldwide are marketed by the Canadian based
Company Calgon Corp (formerly ~olarchem)." Calgon's oxidation process, Rayox@, is based on reactions of OH, generated by irradiation of femc perchlorate in the presence of hydrogen peroxide." Pignatello and ~uan~~~recently demonstrated that PCDD/PCDF are effectively removed by the same ~e~'/'z/hvcombination. Oxidative photolysis of z,~,~,~-('~c)TcDDresulted in 64% minerakation of the contaminant to CO2, HCl and
HzO,with the remainder of TCDD converted to hydrophilic forms, not extractable by to~uene.~~
Titanium dioxide is another source of photogenerated hydroxyl radicals in solution. The Matrix photocatalytic system, based on photodestruction of the contaminants in water in the presence of TiOz, was accepted in the SITE prograrn.21The process was demonstrated to convert organic pollutants, including PCDDPCDF and
PCB, to carbon dioxide, halides and water within 30 seconds to 2 minutes of exposure time.
Extensive literature search yielded only two papers, published by the same authors, with regards to the use of photolysis for the treatment of the dioxin laboratory waste. In contrast to the predominantly particle bound occurrence of the PCDDIPCDF in the environment, in the laboratory waste they are already in solution or could easily be converted into the liquid phase by extraction. Ritterbusch et al.46.47 investigated the UV degradation of polychlorinated and polybrominated dibenzo-p-dioxins and dibenzofurans
in laboratory waste. A high pressure mercury lamp (150 W) was used as the W source.
Dioxin levels in hexane, iso-octane and toluene were reduced by 96-87% &er 30 minutes
of irradiation, and after 1 h no PCDDIPCDF could be quantified in any of these solvents.
Liquid and solid wastes generated by a dioxin-analysis laboratory were analyzed and
detectable amounts of dioxins (fiom 8 -700 nglkg) were found in waste samples
designated as non-hazardous, such as disposable pasteur pipettes, broken glassware,
adsorbents, oil from the turbornolecular pump of the GC-MS used for dioxin analysis,
etc. The authors4' suggested separate collection and treatment of al1 dioxin laboratory waste on site, using their setup. They recommended a disposal concept (Figure 42), in which al1 Iiquid wastes are irradiated directly and the photolyzed solutions may be brought for the solvent recycling. Al1 solid wastes like adsorbents, pasteur pipettes and sample vials are extracted by organic solvent (e.g. hexane); the recycled solvents may be re-used. The extracts are fbrther treated in the same manner as the liquid waste. The extracted waste may be recycled (vials) or disposed of with the normal laboratory waste.
laboratory waste 1
Figure 12. A Scheme of the Disposal of the DioEn Laboratory
143 3.2. Development of a Simple Method for Destruction of Dioxins in the Licpid
Laboratow Waste
The owner of Wellington Laboratories (Guelph, Ont.) approached us with the request to design and test an apparatus that would allow for the destruction of dioxins in liquid laboratory wastes by W irradiation. The above company specializes in synthesis and marketing of native and I3c labeled dioxin standards and in dioxin analysis. Dunng its operation, the laboratory generates small volumes of liquid waste with the concentrations of dioxins ranging fiom 104 to 10~~mol L". The waste is currently accumulated and stored by Wellington Laboratones, since there is no mechanism for disposal of such waste in Ontario (Section 3.1.2). The company plans to use the developed method for detoxication of the waste, so it can be disposed of as the regular organic solvent waste. There were also indications that the methodology may be of interest for the other dioxin analysis labs, especially in ~a~an.~'The requirements for the waste destruction technology were:
- removal of dioxins below the detection limits of high resolution GC-MS
- inexpensive, commercially available equipment with minimal proprietary
modifications
- compact size, easy to setup and operate in a laboratory fume hood, with
minimum fire hazard
Wellington Laboratones financed this project; the company also provided some individual PCDD standards, samples of liquid waste, and performed GC-MS analyses of the waste mixtures before and after the irradiations. 3 -2.1. The se tu^ for the Laboratorv Waste Treatment
The irradiation apparatus is shown on Figure 43. The reaction vesse1 is a standard
100 mL three-neck round botîom flask, equipped with a magnetic stirrer and a reflux
condenser. The condenser is sealed with a rubber septum and a balloon and needle in
order to relieve any gaseous pressure. The PenRayd UV lamp is placed in the custom
designed quartz sleeve inserted in the RBF. The third neck of the reaction flask is covered
with the septum for taking samples during the irradiations. The apparatus is mounted
using clamps on a retort stand over a magnetic stir-plate. The setup was positioned in a
fume-cabinet, in which the sliding Plexiglas window was covered by the aluminum foi1
for the W protection dunng photolyses.
The W source, a 254 nm PenRayO lamp, has several advantages over high or
medium pressure mercury lamps. Because of its low power (4.6 W), the light source has
an operating temperature of 60 OC, it does not require the extemal water cooling and the
£ire hazard due to the overheated solvent is minimal. In Our setup the light source was
placed in a quartz sleeve to avoid the deposition of materials on the surface of the lamp
and to minimize the heating of the reaction mixture (the temperature inside the reactor never exceeded the ambient temperature by 3 OC). Although the reaction temperature was low, we encountered a significant solvent loss due to the evaporation dunng overnight photolyses in hexane even though the reflux condenser was attached to the RBF. We found that sealing the condenser with septa, equipped with a balloon for pressure Dimensions of Quartz Sleeve
Needle
Reflux Condenser
Lamp li 7P~~R~~@uv Quartz Sleeve
Septum for Samplin
150 mL 3-Neck Round Bottom Flask Reaction Mixture
Figure 43. Apparatus for Photolysis of Dioxin-containing Liquid Laboratory Waste relief prevented solvent loss even during 48 hour irradiations. The quartz sleeve was easily made form a commercially available 24 mm quartz male ground joint tube; the sleeve's dimensions are shown in Figure 43.
3 -2.2.Irradiations of Dioxin Standards
Preliminary photolysis expenments were conducted using individual dioxin congeners. Decachlorobiphenyl @CB) was also included in the trials to demonstrate the applicability of the method for detoxification of PCB-containing waste. It was necessary to determine whether the energy of the light source was sufficient to bring about the photodecomposition of dioxins, as well as to find the appropriate solvents and optimum irradiation times. Hexane was a solvent of choice since it is a good hydrogen donor; in addition, our GC-MS was optimized for analyses of hexane solutions. Irradiation of
1,2,3,4-TCDD solution in hexane (1.0~10~~mol L-') resulted in 99% conversion of the starting material in 60 minutes. Irradiation of 1,3,6,8-TCDD (hexane solution, 1.34~10~ mol L") resulted in >99% conversion of the starting material in 120 minutes. For OCDD
(5x IO-' mol L") and DCB (3 x 10" mol L-l) >99% conversion was reached within 15 and
12 minutes of the irradiation, respectively. In every case, the compound of interest dropped below the level of detection limit of Satum HI GC-MS (10 pg for TCDD) in less than five hours of photolysis (Figure 44). Thus, 1,2,3,4-TCDD and 1,3,6,8-TCDD fell below the level of detection in two and four and a half hours respectively, while no traces of OCDD or DCB were found after 60 minutes of irradiation. Disappearance of each compound followed first order kinetics (Figure 45). The rate constants were measured and used to calculate the irradiation times required for the removal of a compound to a -O 1,2,3,4,-TCDD -0 1,3,6,8,-TCDD ---- A OCDD -X DCB
Figure 44. Disappearance of 1,2,3,4-TCDD, 1,3,6,8-TCDD, OCDD, and DCB During the Photolysis at
O 2000 4000 6000 8000 1 O000 12000 Time. sec
Figure 45. First Order Plots for disappearance of 1,2,3,4-TCDD, 1,3,6,8-TCDD, OCDD, and DCB During the Photolysis at 254 nm desired concentration. The measured rate constants ranged from 5.5~10-3 81 for DCB (Co
4 1 = 3.55~10"mol L-') to 7x10 s' for 1,3,6,8-TCDD (Ca = 1.22~lo4 mol L-~);the values
were not compared to each other since the Iight absorbance was different for each
solution. The quantum yields were not detennined because the &fiSs measurements were
not the objectives of this project.
Toluene was tested as a possible solvent media for the photolysis of dioxins at the
request of Wellington Laboratories, due to the fact that some of the company's dioxin waste contained significant amounts of this solvent. We knew that toluene has a very
high Iight absorbance at 254 nrn but considered the possibility that the photolysis may
1 49 proceed through sensitization of dioxins by the solvent (fi= 347 kT mol- ). .
A slow photolysis of decachlorobiphenyl @CB) in toluene was observed at 254 nm. The concentration of 1,3,6,8-TCDD (~0=2.3xl0-'mol L-~)was unchanged after 3 hours of irradiation of a toluene solution. An attempt to assist the photolysis by adding triethylarnine as an electron dono?' dunng the irradiation was unsuccessful. Apparently, sensitization of 1,3,6,8-TCDD by toluene did not take place. Triplet energy of 1,3,6,8-
TCDD may be much higher than ET of DCB @oth values are unknown) due to the absence of conjugation in dioxin ring. For example, triplet energy of benzene is 353 kJ mol-' versus 274 kJ mol-' for unsubstituted biphenyL4' ET value for 1,3,6,8-TCDD rnay be comparable to that of chlorinated benzenes4' (335-345 kJ mol-').
Ritterbusch et at6 reported facile disappearance of PCDD in toluene irradiated using a 150W high pressure mercury lamp. This lamp had emission maxima at 254 and
313 nm. Although toluene transmits light at 313 nrn the absorbance of dioxins at this wavelength is extremely low and photolysis is slow. 1,3,6,8-TCDD photodecomposed upon irradiation at 300 nm in a rnerry-go-round reactor, however, much longer irradiation time was required: after 2 hours only 80% conversion was reached.
Continuous irradiation of 2.2~ IO-^ mol L" solution of 1,3,6,8-TCDD at 254 nm in hexane containing small proportions of toluene (1 to 1W v/v) resulted in disappearance of the starting material (Figure 46). For the solvent mixture containing 1% toluene,
99.8% conversion was achieved after 18 h, whiIe 13% of 1,3,6,8-TCDD remained afler
31 h in 10% toluene. It was concluded that the UV treatment of the dioxin mixtures in solutions containing up to 3% toluene (vh) is feasible, but will require longer irradiation times.
àe OC .-r J C O O
O 5 10 15 20 25 30 35 Tirne, hours
Figure 46. Photolysis of 1,3,6,8-TCDD in Hexaneftotuene Matures at 254 nm 3.2.3. Photolvsis of the Dioxin Wastes fiom a Commercial Laboratorv
Two types of samples of liquid waste were provided by Wellington Laboratories.
The iso-octane solution from the standards lab contained relatively high concentrations
(nom 5 pg!mL to 100 pg/mL; see Section 3.4.8, Table 49) of 1,4,7,8-TCDD and
1,2,3,7,8-PeCDD, and unspecified pentachlorobiphenyl, trichlorodibenzofuran, and pentachlorodibenzofuran. The mixture was irradiated at 254 nm for a total period of 27 h.
The disappearance of the contaminants was monitored by GC-MS (Figure 47). After 4 h of photolysis pentachlorodibenzofùran could not be detected; and only trace amounts of tnchlorodibenzofiiran and 1,4,7,8-TCDD were found afier 24 h. Ln 27 h the concentrations of al1 components of the mixture dropped below the level of detection of the Satum III GC-MS, and no traces of less-chlorinated products were found.
120%
O 5 10 15 20 25 30 Time, hours
Figure 47. Conversion of the Components of the Waste Mixture (Standards) in Photoiysis at 254 nm The second waste sample that undement a test treatment was a mixture of ca. 300 environmental samples, mostly extracts of PCDD/PCDF fi-om tly ash, rnixed with I3c- labeled standards. It contained extremely tow concentrations of PCDDPCDF - ca. 10 ng/mL, far below the detection limits of Satum III GC-MS. The mixture was photolyzed for 6 h at 254 nm using the setup previously descnbed. The analyses of the time-course samples were performed by Wellington Laboratones using High Resolution Mass
Spectrometer (HRMS), operating at 10000 resolution (Appendix C). The instrumental detection limits were 0.25 pg for tetra-congeners, 1 pg for penta-, hexa- and hepta- congeners, and 2.0 pg for octa-congeners. Mer 6 hours of the irradiation the TIC abundances dropped by 1 to 3 orders of magnitude for al1 analyzed congeners, except tetrachlorodibenzo-p-dioxins. Photolysis for additional 21 h resulted in the reduction of the abundances below the instrumental detection limit for al1 analytes (Figure 48). Total Tetra-Dioxins Total PentaDioxins Total He~oxins Total Hepta-ûioxins
27 HRS
a Total Pent*Furans O Total HexaFurans Total HeptaFurans
27 HRS
Figure 48. Change in Peak Areas (HRMS) of PCDD (a) and PCDF (b) During the Photolysis of the Waste with Low Concentration of Contaminants at 254 nm 3 -2.4. Identification of Products of PCDD Photolysis
3 -2.4.1. Mechanism of the Photodeg;radation of PCDDRCDF
The experiments described above dealt mostly with the disappearance of the
starting materials under W light. However, for the practical implementation of the
photolpic waste treatment it was necessary to identify the products of the photolysis of
dioxins.
Photolysis of PCDDFCDF in organic solvents is accompanied by the formation
of less chlorinated congeners." On the basis of this observation, reductive dechlorination
via homolysis was suggested as the rnechanism for the photodegradation of the
dioxins.38,40,52 Homolysis afbrds radical intermediates which yield dehalogenation
products in hydrogen donor solvents (Scheme 3).
ArCl + hv -A~CI* A~CI*- Ai + CL- Ar' + R-H -ArH + R'
Scheme 3. General Mechanism of Reductive Dechlorination by Homolysis
Reductive dechlorination is established to occur upon photolysis of a number of
chlorinated aromatic compounds. Homolytic dechlorination of lower chlorinated
benzenes proceeds with high quantum yields f?om the triplet excîted tat te.'^ Photolytic
e=ciencies of 1-chloronaphthalene, 4-chlorobiphenyl, 9,lO-dichloroanthracene and 1-
chloropyrene are very low, because homolysis is thermodynamically unfavorab~e.~~
Earlier reports30"54*55 have shown that reductive dechlorination is an important transformation pathway in the photolysis of dioxins. The following observations supported the idea that photodecomposition of dioxins occurs via homolysis.
Dechlorination products have been the only produas detected in these studies; and the
reaction efficiency was higher in solvents with readily abstractable hydrogens, e-g.. the
quantum yield of disappearance of OCDD was 20 times higher in hexane than in aqueous
acet~nitrile.~~The addition of sodium b~roh~dride,'~a good hydrogen atom donor,
greatly enhanced photolysis of a number of chlorinated dioxins (please see notd7). A
similar enhancement effect was observed when 2-propanol was added to a methanolic
solution of 2,3,7,~-TcDD.~*
On the basis of extensive GC-MS analyses of the products of photolysis of
2,3,7,8-TCDD in hexane, ~tormer'~suggested that photolysis of dioxins proceeds via a
successive loss of chlorines by homolysis of C-Cl bond resulting in non-chlorinated
dibenzo-p-dioxin. The latter cumpound forms 2-hydroxydiphenyl ether and then 2,2'-
dihydroxybiphenyl by C-O bond scission (Figure 49). Stormer observed the formation of
a dark oil that separated fiom hexane upon concentration of a photolyzed sample, and
concluded that 2,2'-dihydroxybiphenyl undergoes polymerization.
Figure 49. Photodegradation of 2,3,7,û-TCDD by Reductive ~echlonnation~~
155 3.2.4.2. YieIds of Dechlorination Products in Photolvsis of Dioxins
Some experimental data found in the literature suggest that reductive
dechlorination is a minor pathway in the photolysis of dioxins. Dulin et muld not
detect dechlorination products when 2,3,7,8-TCDD was photolyzed in aqueous
acetonitrile. Crosby et aL30 detected only traces of dichloro- and trichlorodibenzo-p-
dioxins after exposing Agent Orange containing 2,3,7,8-TCDD to sunlight for several
hours. leatlwong et al." estimated that only 10% of 2,3,7,8-TCDD is converted to
2,3,7-TrCDD upon irradiation at 3 10 nm in iso-octane.
It was suggested that the reductive dechlorination is important for the dioxins bearing more that four chlonnes, while the less chlorinated congeners undergo some other transfor~nations.~~Experimental data did not support that idea. Wagenaar et uL6' found that less than 5% of OCDD was degraded by reductive dechlonnation in hexane after 136 minutes of irradiation. In our earlier study on the photolysis of OCDD in the presence of triethylamineSowe detennined that the yield of less chlorinated dioxins from photolysis of OCDD in iso-octane in the absence of amine does not exceed 20% even at the early stages of the reaction. The yield significantly improved in the presence of 0.23 mol L-' Et3N (up to 67%), but the less chlorinated products were formed via an electron transfer from the amine to OCDD. In this case the change in the reaction mechanism was demonstrated by the absence of deuterium incorporation in the products when OCDD was photolyzed in the presence of Et3N in cyclohexane-dlî (irradiation of OCDD in cyclohexane-di? resulted in greater that 99.9% incorporation of deuterium into the primary dechorination products in the absence of triethylarnine).sO We observed formation and disappearance of less chlorinated dioxins during the photolysis of OCDD (Figure 50) and 1,3,6,8-TCDD /1,2,3,4-TCDD (not shown) in hexane at 254 nrn. In these experïments the quantitation of the less chiorïnated products was not done and the peak areas in Figure 50 refer to the sum of peak areas of al1 congeners with the same number of chlorines.
We measured yields of less chlorinated dioxins in the photolysis of l.Zxl~-~mol
L-' 2,3,7,8-TCDD in hexane at 300 nrn. The calibration curves of dibenzo-p-dioxin, 2-
MCDD, 2-3-DiCDD, 2,3,7-TrCDD, and 2,3,7,8-TCDD were generated using GC-MS.
The products of reductive dechlorination were quantitated at 36, 84 and 98% conversion of 2,3,7,8-TCDD (Table 42). The maximum yield of less chlorinated dioxins was less
-mono a di +tri *tetra -. penta -t. hexa -t- hepta -octa
O 20 40 60 80 100 120 Time (mln u tes)
Figure 50. Formation and Disappearance of Less Chlorinated Dioxins During the Photolysis of OCDD in Hexane at 2% nm than 2 percent at 36% conversion, the yield dropped at the reaction progressed. No any other products were observed by GC-MS.
Table 42 Determination of the Yields of Dechlorination Products in Photol~sisof 2,3,7,8-TCDD Irradiation the, min O 4 IO 20
Conversion, % O Yield of Less Cldorinateci - Dio.- %
Although low yields of less chlorinated dioxins were noticed by many researchers, no other products have been identified in the literature. To date, several explanations for the low matenal balance have been suggested. ~amantov~~proposed that photodissociation is a heterolytic cleavage yielding the carbon-carbene resonance structure (Figure SI), which may undergo dimerkation with the parent molecule.
However, there is neither experirnental evidence supporting his theory nor, to my knowledge, a precedent for such mechanism in the photochemistry of haloaromatic compounds. cc____)h -Cy-Jy-J-; \ \ \ \ + \ \ C CI CI-
Figure 51. Proposed Formation of Carbene Intermediates in the Photolysis of PCDD~~
Recently Minghui et claimed that lower chlonnated benzenes are the major produas in the photolysis of PCDD, however, 1found some major discrepancies in their paper. First, although these authors suggest that C-O cleavage is the major reaction pathway for the PCDD photodegradation, they state that no phenolic products were detected in the reaction mixhires. Second, according to Minghui et al., di-, tri- and tetrachlorobenzenes were identified as the products of photolysis of 1,2,3,7,8-PeCDD in
carbon tetrachioride, however this solvent is a poor H atom source - it is hydrogen fiee,
so the origin of these 'products' is rather mystetious.
In the study of the sunlight photolysis of 14c labeled 1,2,3,4,7-PeCDD and
1,2,3,4,6,7,8-HpCDD in methanol the authod4 attempted to trace the product formation
by monitoring 14cactivity using HPLC coupled with LSC (Liquid Scintillation Counter).
Analysis of the time-course dichloromethane extracts of the aqueous solutions of these
two dioxins revealed the appearance of transient bands of 14c activity, attributed to
trichloro- and tetrachlorodibenzo-p-dioxins. The bands represented only 15% of 14c
activity at the maximum, they appeared during the first sampling time and disappeared to
baseline with fùrther exposure. Analysis of several extracted water samples for I4c
revealed that the major portion of non-extractable activity was present in the water. The
authors came to the conclusion that the major photoproducts were too polar to be
extracted by dichloromethane and hence a degradation route other than reductive
dechlorination is involved in the photolysis of PCDD?
3 -2.4.3. Material Balance of Photolvsis of 2.3.7.8-TCDD in Hexane
The dificulties in solving the problem of low material balance in photolysis of
dioxins are associated with their low solubility in cornrnon organic solvents. The choice
of analytical techniques is limited to GC, GC-MSor HPLC. We approached this problem
by tracing the products of photolysis of 2,3,7,8-TCDD spiked with 1,6-[~~]-2,3,7,8-
TCDD. We used scintillation counting to quanti@ the amount of tritiated products in
hexane solution, irradiated at 300 nm.. A sample drawn with a gas syringe from a sealed tube containing 2,3,7,8-~~~~/1,6-[~HJ-2,3,7,8-~~~~solution, and irradiated for 3 h (99.9% conversion by GC-MS), exhibited virtually the sarne total radioactivity as before irradiation (3 16Sf 110 DPI versus 33 8OSO DPI per 10 pL aliquot). This observation signifieil that the products of photolysis are still present in solution, rather than adsorbed on the glass surface of the reaction tube, precipitated, or present in the gas phase. We could not observe any significant peaks after injecting the irradiated solution of 'cold'
2,3,7,8-TCDD on HPLC. However, scintillation counting of small fiactions collected by passing 20 pL, of irradiated 1,6-[3Hj-2,3,7,8-~~~~through a Cis column enabled us to construct HPLC chromatograms of the products (Figure 52).
3 -2-4.4.Identification of Polar Products in Photolvsis of 2.3.7.8-TCDD in hexane
Short retention times of the product peaks on reversed phase HPLC column indicated that the products of photolysis of 2,3,7,8-TCDD are polar substances. The conditions for the product separation on HPLC were optimized using the mixture from photolysis of 1,6-['H]-~,~,~,~-TcDD.
With pure acetonitrile as the mobile phase one broad peak with retention time of
4.5 min was obsewed. In pure methanol two peaks were resolved: the smaller peak had the retention time of 3.5 min and the larger peak eluted at 5.5 min. Addition of water
(5% V/V) to methanol did not improve the separation, since the retention times shortened and the peak at 4 min exhibited tailing. We were able to observe the products by conventional HPLC equipped with W absorbance detector (h = 227 nrn), when the product mixture corn the photolysis of 'cold' 2,3,7,8-TCDD was concentrated 500 times.
The chromatogram acquired using 100% MeOH as the mobile phase at a flow rate of 1 mL/rnin had a shape similar to that of C in Figure 52, except the 4.00 6.00 8.00 Time, min
Figure 52. HPLC Chromatograms of 3~-containingProducts of Photolysis of 1,6-[.'~-2,3,7,8-~~~~ in Hexane at 300 am The mobile phase was: A - 100°/o acetoniide, B - 95% MeOH and %H20, and C - 100% MeOH. The flow rate was 1 mL/min in aii nuis. retention times of the two major peaks were longer by 0.2 min and six new smaller peaks were resolved. Fractions 3 to 4 minutes and 4 to 6 minutes were collected fiom five injections on the analytical colurnn, the solutions were concentrated and subjected to analysis by ES-MS. Three major peaks with m/r 271, 303 and 321 in the ratio of 1:3:4 were observed in the fiaction 3 to 4 minutes, and m/z 321 was the only peak fi-om the fiaction 4 to 6 minutes. Mass spectral data were acquired in negative ionization mode and the odd m/z values correspond to M-Y with the protons lost nom hydroxyl groups. The substance with m/r 321 (Figure 53) demonstrates a characteristic four-chlorine pattern and has a molecular formula of Cl2&ChO2,and dz303 and 271 both display a three- chlorine pattern and correspond to molecular formulas of C12H7C1303 and C12H7C130 respectively.
Comparison of total radioactivity of 1,6-[3J3J-2,3,7,8-~~~~to the activities of the polar products allowed us to estimate their yield. Datapoints from chromatogram C,
Figure 52, were converted to ASCII text and integrated using PeakSimple II sohare.
The peak areas were 1380 DPM and 4480 DPM for the peaks at 3.5 and 5.5 min respectively. The total activity count of 5860 DPM accounts for ca 87% of the total 'H, since the activity of 1,6-[3~-2,3,7,8-~~~~was 3380k60 DPM per 10 pL, and 20 pL of the photolyzed mixture was injected to obtain chromatogram C. The ratio of the peak areas is 1 :3.3; it appears that the substance with the retention time of 5.5 min, with the molecular formula established by ES-MS as Cl2&Cl4O2, was the major product in the photolysis of 2,3,7,8-TCDD. The molecular formula suggests that this substance (Mz
321) may either be 2-hydroxy-3,3',4,4'-tetrachlorodiphenyl ether 45 or 2,2'-dihydroxy-
4,4r,5,5'-tetrachlorobiphenyl 46. One piece of evidence favors the structure 45: Figure 53. ES-MS Chromatogram of Products of Photolysis of 2,3,7,8-TCDD in ES-MS, under the conditions of negative ionization the compound 46 would lose both protons and become a dianion; such species would give a signal with m/z of 160, corresponding to (M-2H)/2. We did not observe any fragments with Mz 160 in the ES-
MS spectra of the photproducts. For either structure it is significant that the newly formed rnolecule retained al1 chlorines.
cl'
Ne attempted to isolate the products Rom irradiation of a large amount (25 mg of 2,3,7,8-TCDD. The dioxin was photolysed in hexane at 300 nm for 3 hours, until 97% conversion of the starting matenal was reached. The mass of the brown material obtained after the solvent removal was 24.8 mg. TLC analysis with a mixture of hexane and chloroforrn (1 :1) showed three spots with Rf of 0.2, 0.24 and 0.5, and a dark spot at the origin, which could be mobilized only by methanol. Separation by flash chromatography yielded three fractions: 0.5 mg (Rf0.2), 2.7 mg (Rr 0.24) and 1.9 mg (Rr 0.5). The substance remaining on the column bed was washed off by MeOH, its weight was 19 mg.
This material was separated into two fractions (5.8 and 7.6 mg) by semi-preparative
HPLC. ES-MS analysis revealed that each fraction was a complex mixture of products
(Appendix D). The compounds with m/z 205 and m/z 322 were present in each fraction.
In the early fractions from flash chromatography a large number of the compounds with high m/z, displaying five-, six- seven- and eight- chlorine patterns were found. The starting material was >99.9% pure by HRMS. Some of the higher chlorinated products are formed by chlorination of the primary photoproducts with the atomic chlorine formed by homolysis of 2,3,7,8-TCDD eg. a substance with m/r 357 had a five-chlorine pattern and the molecular formula Ci2HsC1s02. The other products appeared to mise from dimerization of the primary hydroq produas, eg the substance with m/t 642 displayed an eight chlorine pattern.
A wide array of hydroxylated products was also found after the irradiation of
2,3,7,8-TCDD and 1,2,3,4-TCDD bonded to C is solid support in aqueous 2-propanol.s8
According to the authors' claim, traces of poly- to non-chlorinated phenols, catechols, and hydroxydiphenyl ethers were detected using GC-MSD. It was noted that the produa composition was strongly dependent on the solvent composition and the length of irradiation.j8 We did not observe any monoaromatic hydroxylated compounds in Our experiments. It is worth mentioning that Our numerous attempts to observe other than the reductive dechlonnation products by GC-MS were unsuccessful, probabl y due to the adsorption of hydroxylated aromatic produas ont0 the silylated surface of the injector sleeve.
3 -2.4.5. Possible Mechanism of Photolvsis of 2.3.7.8-TCDD
Observation of the hydroxylated products by ES-MS and the finding that the formation of such products accounts for at least 87% of the disappearance of 2,3,7,8-
TCDD during W irradiation suggests that the breakage of the C-O bond rather than the
C-Cl homolysis is the major pathway in photolysis of dioxins. A few observations found in the literature support the idea that C-O bond cleavage is a feasible alternative to the
reductive dechlorination mechanism. Diphenyl ether (47) rearranges to O- and p-
phenylphenols upon irradiati~n.~'Photolysis of unsubstituted dibenzo-p-dioxin yields
2,2'-dihydroxybiphenyl (48). 66.67 It was proposed that the primary photochernical step is
a ring opening to give biradical 49, which on reduction by hydrogen abstraction from
solvent gives 2-phenoxyphenol (50). Subsequent phototransformation gives 48 as the
major product, with a minor amount of 4-hydroxydibenzofùran 51.66.67 It should be noted
that in the ES-MS chromatogram of the products of photolysis of 2,3,7,8-TCDD we
observed a trace of a substance with m/z 285 with the molecular formula of C12H5C1302r
matching a hydroxy-trichlorodibenzofiiran
48 HQ major
Guan and Wan 68,69 suggested alternative mechanism for photolysis of dibenzo-p- dioxin. During photolysis these researchers observed formation of a transient by W-Vis spectroscopy and suggested 2,2'-diphenoquinone intemediate 52, rather than JO, as the precursor for 2,2'-dihydroxybip henyl 48. The authors presented the following arguments in favor of their mecahnism. During the photolysis of dibenzo-p-dioxin the yield of 48 54 rl 55
Figure 54. A Mechanism of Photolysis of Dibeazo-p-Dioxin grew steadily with essentially no change in the yield of 50 (less than 1%). When 50 was photolysed in the control experiment, the major product was catechol, while 48 could not be detected in the product mixture. According to the mechanism suggested by Guan and
Wan (Figure 54), reduction of biradical 49 by solvent is a very minor pathway; 49 recombines by ipso attack to give a spiro ketone 53, with the latter rearranging to
2,2'-biphenylquinone 52 by homolysis of aryl-oxygen bond. The intermediate 52 is reduced by solvent to give 48; in aprotic solvents 52 undergoes ring closure via secondary photolysis, giving rise to biradical 54, resulting in a second major product,
1-hydroxydibenzofùran 55 (the structure established by a comparison of 'H NMR to that of authenic ~arn~le)~'
The mechanism of photolysis of 2,3,7,8-TCDD may differ f?om the one shown in
Figure 54. According to Guan and Wan, photolysis of 2,3,7,8-tetramethyldibenzo-p- dioxin 56, a structural analog of 2,3,7,8-TCDD, in THF resulted in 4,4',5,5'-tetramethyl-
167 2,2'-dihydroxybiphenyl (57) as the only product;.69a formation of pink-colored transient
was noted dunng the reaction. We never observed color changes during the photolyses of
2,3,7,8-TCDD even when reasonabl y concentrated solutions were irradiated (Section
3.4.13). The major contra-argument to the above mechanism is the ES-MS evidence for
the formation of a chlorinated diphenyl ether rather than a chlorinated dihydroxybiphenyl
as the major product in photolysis of 2,3,7,8-TCDD.
Accumulated evidence allows us to derive some conclusions on the nature of
processes occurring during the photolysis of 2,3,7,8-TCDD and, probably, other
chlorinated dioxins. It is most likely that in the photolysis of 2,3,7,8-TCDD in hydrogen
donor solvents C-O bond scission is the pnmary step. The product, similar to 45
undergoes fùrther phototransformations, including dechlorination, dimerization and re-
chlonnation by atornic chlorine. Homolysis may compete with C-O rupture in the
photolysis of 2,3,7,8-TCDD, but the less chlorinated dioxins rnay also undergo the ring
opening reaction. As a result, a highly cornplex mixture of polar products with various degrees- of chlorination is fonned; the product composition rapidly changes as the
reaction progresses. 3.2.5. Recevtor Bindine Assavs of the Products of Photol~sisof 2.3.7.8-TCDD
Mere demonstration of the disappearance of dioxins under UV light and the
knowledge of the types of products formed are not sufficient for the practical application
of the developed method for the destruction of dioxins in liquid waste. It is necessary to
detemine whether the products formed in photolysis lose the toxic effects associated with dioxins.
This task was accomplished using a biological assay, in which one of the most important aspects of the dioxin to~icit~~~~~'- the ability to bind to Ah receptor was tested. 2,3,7,8-TCDD was chosen as a mode1 compound since this substance is known for its strong binding afinity to the Ah receptor.' In the experiment, products from various stages of photolysis of 2,3,7,8-TCDD in hexane were incubated with 1,6-1~HJ-2,3,7,8-
TCDD and rat hepatic cytosol containing the Ah receptor. The cornpetitive binding experiment was performed by J. Petrulis. The experiment was designed so that the unp hotolyzed control was in excess over 1,6-[~~]-2,3,7,8-TCDD, and thus outcompeted it for occupation of the Ah receptor. In the photolyzed samples the concentration of the
2,3,7,8-TCDD dioxin decreased, and, given that the photoproduas lack the Ah receptor affinity, the decline in the observed cornpetition for Ah receptor binding sites should follow the conversion of 2,3,7,8-TCDD with irradiation.
The graph representing the relationship between percent specific binding of 1,6-
[3~-2,3,7y8-~~~~and the conversion of 2,3,7,8-TCDD is shown in Figure 55. The specific binding of 1 y6-[3KJ-2,3,7,8-~~~~and conversion of 2,3,7,8-TCDD, both -t%Specific Binding - - O - - % Conversion
1.00E-09 1.00E-10 [2.3,7.STCDD] mol C'
Figure 55. The Relationship between Percent Specific Binding of 1,6-[313J-~,3,7,8-~C~~and the Conversion of 2,3,7,8-TCDD in Photolysis
expressed in percentages are plotted against the concentration of remaining 2,3,7,8-
TCDD.
The binding afinity of the photolyzed mixture declined with [2,3,7,8-TCDD] .
The binding decline was slower in the early stages of the reaction, but at high conversion the photolyzed mixture did not express any binding affinity. The results of this experiment are significant since we demonstrated that the product of photolysis of
2,3,7,8-TCDD do not exhibit toxicity associated with the Ah receptor binding. It is also important to realize that longer irradiation times are required since the binding loss lags behind the loss of 2,3,7,8-TCDD in p hotolysis. Endocrine disrupters and modulators have been the subject of many recent reports in scientific and public media. Several phenolic compounds such as nonyl- and octyl- phenols were identified as the estrogen endocrine disrupters. Chlorinated compounds, such as 4-hydroxy-2',3',4',5'-tetrachlorobiphenyls were shown to exhibit binding to the estrogen receptor. The products of photolysis of 2,3,7,8-TCDD were chlorinated aromatic cornpounds bearing at least one hydroxy group and could exhibit the estrogen receptor binding ability. We were able to test this possibility using the estrogen receptor binding assay developed in our laboratory by B.J. Cox.
An assay in which the products of photolysis of 2,3,7,8-TCDD were competed against tritiated estradiol for the estrogen receptor in a protein was perfonned with B.J.
Cox to determine whether the photoproducts from 2,3,7,8-TCDD possess afinity for binding to the estrogen receptor. Cornpetitive displacernent of tritiated estradiol with increasing concentration of the photoproduas would indicate that the products of photolysis are able to bind with the estrogen receptor, and therefore may be estrogenically active. The samples taken at 0, 13, 43, 88 and 99% conversion of 2,3,7,8-
TCDD were tested. The proteins, associated with specific and non-specific binding were separated by HPLC using gel-filtration column; the activity in the collected eactions was measured using scintillation counter and plotted against the elution time. The peak area of the chrornatographic peaks (Figure 56) was measured using PeakSimple II software.
The results of the experiment are presented in Table 43. Table 43. Measured Radioactivity of the Sarnples in the Estradiol Assay on the Products of Photolysis of 2,3,7,&TCDD Ratio of Specific Product Concentmtion, mol Specioc Bindin& DPM % Conversion Non-S~ecific and non-specsc L" Binding*, DPM bindine O O 3928 8882 0.44 13.2 3.58~10" 3855 7807 0.49 43.3 1.17~lo4 3950 8291 0.47 81.8 2.2 1 x lo4 3382 6887 0.49 99.9 2.7 1 x 104 3212 7492 0.43 *-includes fiee radio-ligand
Figure 56. Change in Radioactivity Associated with Estrogen Binding Afiinity after Photolysis of 2,3,7,&TCDD Although both the specific and non-specific binding decreased, there was no significant change in their ratio. The specific binding decreased by 18% at the highest concentration of the photoproducts, similar reduction, by 15% was observed for the non- specific binding. This reduction may be or may not be considered as an indication of a weak binding to the estrogen receptor due to significant margin of error associated with the integration method. From data in Table 43 it is certain that the products of photolysis of 2,3,7,8-TCDD do not exhibit *ong estrogen receptor binding affinity.
3 -3. Conclusions for Cha~ter
A safe, efficient and inexpensive method for on-site detoxification of small volumes of PCDDPCDFIPCB-containing liquid laboratory waste using W irradiation from a low-pressure mercury lamp was developed. The method was successfully tested on individual PCDDRCB congeners and on actual waste from the dioxin laboratory.
New data on the mechanism and products of photolysis of 2,3,7,8-TCDD under conditions of homolysis were obtained. We demonstrated that the majority of the products of photolysis of chlorinated dioxins &se not from reductive dechlorination, but frorn the scission of the C-O bond in the dioxin ring in experiments with 2,6,-l3HI-
2,3,7,8-TCDD. By means of the scintillation counting of the tritiated reaction mixture, separated on HPLC it was established that the polar photoproducts account for at least
87% of the material balance in the photolysis of 2,3,7,8-TCDD in hexane at 300 nm. The major products were tentatively identified by electrospray mass spectrometry. We were unable to isolate the individual products due to the cornplexity of the resulting photolysis mixtures. In future considerations the most practical approach to structural product identification would be an LC-MS comparison of the product peaks with the authentic samples.
Receptor binding assays on the products of photolysis of 2,3,7,8-TCDD, conducted in association with B. Cox and J. Petrulis, revealed that the products do not exhibit strong estrogen receptor binding afinity, and they do not bind to the Ah receptor. Since the toxicological effects of dioxins are mostly associated with Ah receptor binding affinity, we can conclude that the products of the W treatment of PCDD/PCDF laboratory waste would not exhibit the toxicity by this pathway and could be disposed of as regular organic solvent waste. 3 -4. Ex~erimentaiSection
3 -4-1.Materials
HPLC grade hexane and iso-octane, ragent grade toluene were fiom Fisher
Scientific. Al1 solvents were used as received. OCDD (98% purity) was previously synthesized in our laboratory; 1,2,3,4-TCDD and decachlorobiphenyl (99% purity) were purchased fiom Accu-Standard; 2,3,7,8-TCDD, 2,3,7-TrCDD, 2,3-DiCDD, 2-MCDD, and dibenzo-p-dioxin (al1 +99% punty) were supplied by Wellington Laboratories; 1,6-
[3~-~~~~(specific activity 37 CVmmol, > 98% purity) was purchased fkom Chemsyn.
Cytoscint ES scintillation fluid was fiom ICN Pharmaceuticals.
3 -4.2. Instrumentation
Qualitative and quantitative gas chromatograp hy - mass spectral analyses were performed using Satum 3 GC-MS (the description of the instrument is in Section 2.4.4,
Chapter II). Two sets of the acquisitions methods were used:
Method for analysis of tetra-chlorinated dioxins (Program 1):
Injecter: 90 OC, 0.5 min isothermal; 200 Oc/min to 250 OC; 1.1 min isothennal Colzimn: 50 OC, 0.1 min isothermal; 50 O~/minto 150 OC, 20 OWrnin to 250 OC; 8 min isothennal Auxiliury Line: 270 OC
Ion Trap: EI mode, 2 segments:
Segment 1: Filament Delay 2.5 min Mass range: 100 to 380 m/z Mass defect: -66 muA00u Segment Length: 7.10 min
Segment 2: Filament Delay O min Mass range: 100 to 380 m/z Mass defect: -66 mu/100u Segment Length: 7.9 min Method for analysis of OCDD and DCB (Program 2):
Injector: 90 OC, 0.5 min isothermal; 200 *clmin to 250 OC; 1.1 min isothermal CoIumn: 50 OC, 0.1 min isothermal; 50 O~/minto 150 OC, 20 O~/minto 250 OC; 27.90 min isothermal AuxiIimy Line: 270 OC
Ion Trap: EI mode, 3 segments:
Segment 1: Filament Delay 2.5 min Mass range: 100 to 380 m/z Mass defect : -66 mu/ 100u Segment Length: 7.10 min
Segmertt 2: Filament Delay O min Mass range: 100 to 480 dz Mass defect: -66 md100u Segment Length: 12.90 min
Segment 3: Filament Delay O min Mass range: 100 to 520 m/z Mass defect: -57 md100u Segment Length: 15 min
Method 1 was modified for the analysis of the samples in toluene: the starting injector temperature was raised to 110 OC, and the filament delay in the segment 1 was increased to 4.00 min.
3 -4.3. Calibration Curves
Calibration curves for dibenzo-pdioxin, 2-MCDD, 2,3-DiCDD, 2,3,7-TrCDD,
1,2,3,4-TCDD, 1,3,6,8-TCDD, 2,3,7,8-TCDD, OCDD, and DCB were constmaed using automated intemal standard quantitation. The interna1 standard was hexachlorobenzene
(5.0~10~mol L-'). The appropriate arnount of the analytes (1.5 - 2 mg) was weighed using either Mettler AT 250 or Mettler AE 260 analytical microbalances (reliability 0.01 and 0.1 mg respectively), and dissolved in hexane in a 25.00 mL volumetnc flask using an ultrasonic bath. Solutions of different concentrations were prepared by the method of successive dilutions. At least five points were taken to construct each curve. The volume of the injections was 1 pL. For each analyte the ratio of the area of the sample to the area of the internal standard was plotted versus the amount of the sample injected. The concentration in the unknown sample using this method was calculated by Saturn software using the following formula:
Ani~i~~t-~~,= (AreaqiJAreais) x (Amozmt&F) x IS-,
where IS is an internal standard, RF is a response factor, determined fiom a calibration curve and ISfaa, is the coefficient, adjusting for the deviations in the injection volumes. The calibration curves are given in Table 44.
3.4.4. Photolysis of 1.2.3.4-TCDD, 1.3.6.8-TCDD, OCDD. and DCB in Hexane
In each experiment 25.0 rnL of a hexane solution of an individual dioxin @CB) of known concentration was irradiated at 254 nm in the apparatus described in Section
3.2.1, with 0.5 rnL aliquots being drawn at set intervals. The aliquots were mixed with
0.5 mL of a hexane solution of the internal standard (HCB, 1.0~IO-' mol L-') and subjected to GC-MS analysis. The results of the analyses, including concentrations and percent conversions, are presented in Table 45.
3 -4.5. Photolvsis of 1.3.6.8-TCDD in Toluene at 254 and 300 nm
A 2.3x10-~ mol L-' solution of 1,3,6,8-TCDD in toluene was prepared by removing solvent from 10 mL of a 2.3 x lo5 mol L-' solution of 1,3,6,8-TCDD in hexane on a rotary evaporator and redissolving dned dioxin in 100 mL of toluene. Twenty Table 44. The Cdibration Curves of Dibenzo-pdiorin, 2-MCDD, 2&DiCDD, 2,3,7-TrCDD, 1,23,4- TCDD, 1,3,6,8-TCDD, 2,3,7,8-TCDD, OCDD, and DCB
Concentration, Concentration, Concentration, Area/AreaIS mol L" WArears mol L" A=a/AreaIs mol L"
Slope =IO9602 mol-' L , = 0.994 Slope = 123505 mol-' L, 2 = 0.998 Slope = 108964 mol-' L, >r = 0.997
Concentration, Area/AreaIs Concentration, Area/AraIS Conceniration, ArealArais mol L" mol L-' mol L-'
Slope = 72848 mol-' L, 2 = 0.997 Slope = 42039, r' = 0.995 Slope = 25850 1, 2 = 0.988
1,3,6,8-TCDD OCDD DCB Concentration, AreafAreaIS Concentration, ArealAreau Concentration, Area/Areais moi L" mol L" mol L-'
Slope = 250750mol-' L, 2 = 0.995 Slope = 86055, 2 = 0.985 Slope = 168685, 2 = 0.996 * - The concentration of the internai standard was 5x10" mol L-' Table 45. Photolysis of 1,2,3,4-TCDD, 1,3,6&TCDD, OCDD, and DCB in Herane
The, sec Concentration, mol L" Conversion, % WC&
- Tirne, sec Concentration, mol L-' Conversion, %
10800 8.00~IO-' 99.9 7.33 * - 30 rnL solution was photolyzed OCDD
Time, sec Concentration, mol L-' Conversion, % In (cc&
4-23x
1.28~10-~
8.70~IO-'
2.60~LO-~
Not detected
------Time, sec Concentration, mol L-' Conversion, % Mc&
3.55~104
8.56~104
1.15~lo6
3.20~10-~
1.40~IO-7
1.20~10-~
4.00~10-'
Not detected ** - the volume of the photolyzed solution was 20 mL, 200 pL samples were taken and
mixed with 200 p.L interna1 standard solution
milliliters of the toluene solution were irradiated at 254 nrn in the reaction apparatus
(Section 3.2.1) for 3.5 hours. After 2 hours of irradiation a drop of triethylamine was added to the reaction mixture. At set intervals 0.5 mL aliquots were taken, mixed with 0.5 mL of a hexane solution of the internal standard (HCB, 1.0~1 o5 mol L-') and subjected to GC-MS analysis.
Two 8 mm Pyrex0 tubes containing 2.0 mL of 2.3~10~mol L-'solution of
1,3,6,8-TCDD in toluene (a drop of Et3N was added to one of the tubes) were placed in the meny-go-round photoreactor and irradiated at 300 nm. At 30, 60 and 120 minutes 0.2 rnL aliquots were drawn, mixed with 0.2 mL of the internal standard solution and analyzed by GC-MS.
The results of the analyses are presented in Table 46.
Table 46. Photolysis of 1,3,6,8-TCDD in Toluene at 254 and 300 nm
Time, min O 10 20 45 60 120 180
Concentration, mol L-' 2.15~10- 1.99~10' 2.18~10' 2.26~10' 2.02~10- 2.25~10- 2.09~10-
No Et3N Et3N added The, min Concentration, mol L-' Conversion, % Concentration, mol L-' Conversion, %
3 A.6. Photolysis of DCB in Toluene at 254 nm
The experimental setup was described in Section 3.2.1. The initial concentration of DCB in toluene was 5.4x10-~mol L-'. The volume of irradiated solution was 10 mL. The aliquots (0.5 mL) were taken at 0, 35, 65 and 90 minutes, rnixed with the equal volume of the intemal standard solution and subjected to GC-MS analysis (Table 47)
Table 47. Photolysis of DCB in ToIuene at 254 nm - .- Time, min Concentration, mol L-' Conversion, 5%
3 -4.7. Photolvsis of 1.3-6.8-TCDD at 254 nm in Hexane Containing: 1. 3 and 10 Percent
Toluene
Solutions containing 1, 3, and 10% toluene were prepared by adding to 0.25, 0.75 and 2.5 mL solutions of 2.3~lo5 mol L-~1,3,6,8-TCDD in toluene a solution of 1,3,6,8-
TCDD in hexane (2.3~10"mol L'~)to total volumes of 25.0 mL in volumetrk flasks. The volumes of photolytes in each case were 10.0 mL; the solutions were irradiated at 254 nm, using the experimental setup described in Section 3.2.1.The aliquots (0.5 rnL) were taken at set time intervals, mixed with the equal volume of the interna1 standard solution and subjected to GC-MS analysis (Table 48).
Table 48. Photolysis of 1,3,6,8-TCDD at 254 am in Hexane Containing 1,3 and 10 Percent Toluene 1% ~oluene 3% Toluene 10% Toluene Time Concentration Conversion Time Concentration Conversion Tirne Concentration Conversion min mot L-' % min mol L" YO min mol L-' YO O 2.25 x IO-' 0.0 O 2.19~IO-' 0.0 O 2.19~IO-' 0.0 3 -4.8.Photolvsis of the Dioxin Waste fiom Wellington Laboratones (Standards)
Awaste mixture containing known concentrations of a pentachlorobiphenyl, a
trichlorodibenzofuran, a pentachlorodibenzofuran, 1,4,7,8-TCDD, and 1,2,3,7,8-PeCDD
(Table 49) in iso-octane was provided by Wellington Laboratories. A five-point
calibration file was prepared by successive dilutions of the analyte and quantitation of the dilution series on Satum III GC-MS with HCB as the interna1 standard (5x10~mol L") using Program 2 (Section 3.4.2). Thirty milliliters of the waste solution were irradiated at
254 nm using the experimental setup, descnbed in Section 3.2.1; 0.5 mL aliquots were drawn after 1.5,4, 24, and 27 hours of the irradiation. An equal volume of 1 x 10" mol L-l solution of HCB in hexane was added to the aliquots pnor to the analysis by GC-MS. The results of the analyses are presented in Table 49.
Table 49. Photolysis of the Diosin Waste from Wellington Laboratories (Standards)
Compound CI5 PCB TrCDF TCDD PeCDF PeCDD Quantitaion 326 272 322 340 356 Mass Time, hours Concentration, mol L"
Time, hours Conversion, % 3.4.9. Photolvsis of the Dioxin Waste fiom Wellinaon Laboratories Extracts1
Thirty milliliters of a complex mixture of l3cC-pc~I3,"c-PCDF and native
PCDD/PCDF with the concentrations of the individual congeners not exceeding 10
ng/mL in iso-octane were obtained from Wellington Laboratones. Two milliliters of the
solution has been retained for the analysis by HRMS, while the remaining volume was
photolyzed for 27 h at 254 nm using previously described setup. One milliliter samples
for the analyses were taken after 1.5, 3.5, 6, and 27 hours of irradiation. The mass
spectrorneter was Micromass VG70SE, coupled with Hewlett Packard 5890 Series II gas
chromatograph, equipped with 60 m capillary column (J&W, 60m x 0.25mm x 0.25 pm).
Peak areas of the congener groups were integrated using selective ion channels (Table 50).
Table 50. PhotoIysis of the Dioxin Waste from Wellington Laboratories (Extracts)
Tirne, hours O PCDD** Peak Area
Total Tetra-Dioxins 9.2~lo7 9.8~107 1.0~lo7 9.4~107 1.5~106
Total Penîa-Dio.- 4.4~10' 8.8~107 6.7~lo7 2.6~10' 1.1EM6
Total Hexa-Dioxins 1.Ox 108 9.9~107 3.3~107 5.2~106 O
Total Hepta-Dioxins 7.2~107 6.2~107 6.0~106 O O OCDD 1.7~10' 1.2~lo7 2.4~los 1.7~10' 2.0~105 PCDF Peak Area
Total Tetra-Furans 1.6~10' 9.6~10' 4.0~10' 1.5x108 O
Total Penta-Furans 7-6x log 6.3~log 3.4~10% 1.6~10' O
Total Hed--Furans 5.6~log 4.3~10' 1.9x108 6.8~107 7.5~lo4
Total Hepta-Furans 2.5~10' 4.5~lo7 1.7~106 O 1.2~los
OCDF 8.8~10' 1.2~106 1.2~los 1.8~10' 2% 10' * - A signai 40-noise ratio was Iess than 3 for 27 hour sample 3.4.10. Determination of the Yields of Dechlorination Products in Photolvsis of 2.3.7.8-
TCDD
Two milliliters of I.ZXIO-~ mol L-' solution of 2,3,7,8-TCDD in hexane in 8 mm
Pyrex@ tube, sealed with a rubber septum, were photolyzed at 300 nm in a merry-go-
round photoreactor. The aliquots (200 pL) were taken der 4, 10 and 20 minutes of the irradiation, mixed with 200 pL of 5x10~mol L-' HCB in hexane, and subjected to
anaIysis by GC-MS. The material balance of dechlorination, yield of Less-chlorinated products and percent conversion were calculated based on the quantitation data (Table
Table 51. Determination of the Yields of Dechlorination Products in Photolysis of 2,3,7,&TCDD Irradiation time, min O** 4 10 20
Dio.uin congener: Concentration, mol L-'
Total Concentration, mol L-' 1.19x10-~ 7.67~lo4 1.95~104 2.88~10" Total Concentration of the 8.86~10" - -- dechlorination D~O~UC~S,7 mol- L-' * -Two other possible dichioro-products, e.g. 2,6-DCDD and 2,7-DCDD were accounted for by the integration of the comesponding peaks using the quantitation fite for 2,3-DCDD. ** - Trace amounts of less-chlorinated dioxins (total of 0.75%) were present in the original sample.
3 -4.11. Photol~sisof 2.3.7.8-TCDD Spiked with 1.6-[-'Hl-2.3.7. 8-TCDD
Ail the experimental work including handling of the tritiated TCDD was performed in the areas of the laboratory designated for the expenments with radioactive rnaterials, using the appropriate means for personal safety. The activity of 3~-containing materials was measured on Beckman LS 7000 Scintillation counter. Wipe tests were
perfomed afier each experiment.
The stock solution of tritiated 2,3,7,8-TCDD was prepared by dissolving 10 pL of
1.0~10~mol L-* of I,~-[~HJ-~,~,~,~-TcDDin DMSO with specific activity of
37Ci mmol-' in 1.0 mL of water and extracting the water containing dioxin with 2.0 mL of hexane. The resulting solution of 1,6-[3~-2,3,7,8-~~~~in hexane exhibited the activity of 2.2 1M.1 1x 1o4 DPM for 10 pL aliquots (3.7~10" mol L-'1. A stock solution of
2,3,7,8-TCDD (9.32~lo5 mol L-', 0.9 mL) was placed into two 8 mm Pyrex@ tubes.
Hexane (0.10 mL) was added to one tube and 0.1 mL of the stock solution of I,~-[~H]-
2,3,7,8-TCDD in hexane to the other. The tubes were capped with rubber septa, placed in the mex~y-go-roundphotoreactor and irradiated at 300 nm. The disappearance of 'cold'
2,3,7,8-TCDD was monitored by GC-MS. For this purpose, aliquots of 20.0 JLLvolume were taken fiorn the tube, not containing 1,6-[3q-2,3,7,8-~~~~, before and dunng the irradiation (20 pL of 5x10~mol L-' HCB in hexane were added to each aliquot prior to the injections). After a 3-hour penod the conversion of 'cold' 2,3,78-TCDD had reached
99.9% (GCMS). Three 10 pL aliquots were taken with a gas-tight 25 pL syringe through the rubber septum fiom the tube containing 1,6-[3~-2,3,7,8-~~~~before, and then after irradiation; each sample was mixed with 2.0 mL of the scintillation cocktail and placed in the scintillation counter.
3-4.12. Separation of the Products of the Photolvsis of 1.6-r3m-2.3.7. 8-TCDD on HPLC
The HPLC setup was as follows: Perkin Elmer Model 250 isocratic LC pump,
Rheodyne 7010 Injector equipped with 20 pL sample loop, Waters p-Bondapack Ci*
3.9~300mm colurnn, Gilson Model 202/204 fraction collecter. For most runs the flow rate was 2 mL min-'. In a typical run 50 fi of the product mixture fiom photolysis of 1,6-
[3HJ-2,3,7,8-~~~~(Section 3.4.1 1) was injected; the fiactions of a set volume were collected directly in the scintillation vials. Chromatograms were obtained by measuring the activity in each fraction on the scintillation counter and plotting the DPM values against the elution time. The separation conditions were optimized by varying the mobile phase and the ilow rate (Table 52).
3 -4.13. Large Scale Photolvsis of 2.3.7.8-TCDD
Twenty five rnilligrams of 2,3,7,8-TCDD of > 99.9% purity, provided by
Wellington Laboratones, were dissolved in 400 mL of hexane with aid of the ultrasonic bath, thus making ca 2x10~mol L" solution. Four Pyrex@ tubes of 60 mL capacity and
24 mm 0.d. were filled with the prepared solution (CU. 50 mL in each tube), cored with rubber septa and placed in the merry-go-round photoreactor. The solutions were irradiated at 300 nm, the progress of the reaction was followed by GC-MS, with 100 pL aliquots taken every 30 minutes and diluted 100 times with hexane before the analysis.
After 3 h of irradiation, 97% conversion was reached and the remaining solution of
2,3,7,8-TCDD was photolysed for the same period of time. Combined photolytes were stripped of the solvent on a rotary evaporator yielding 24.8 mg of the products. This material was coated onto 2 cm3of silica gel (60-200 mesh) and placed ont0 a bed of pre- equilibrated flash chromatography cohmn. The column was eluted with 500 rnL of the Table 52 Separation of the Producîs of the Photolysis of 1,6-[3m-2,3,7,8-~~~~on HPLC
100% MeOH* 90% MeOH: 10% HzO 100% CH3CN Elution the, min Activity, DPM Elution time, min Activity, DPM Elution time, ruin Activity, DPM
*- Composition of the mobile phase.
188 mobile phase, comprised of 50% hexane and 50% chloroform. Collected fiactions (18
fiactions of 25 mL) were subjected to analysis by TLC and combined, if similar. The
column was then washed with 300 mL MeOH resulting in 19 mg of dark brown material.
The above material was redissolved in MeûH (2.0 rnL) and subjected to separation by
HPLC using Alltech Econosphere Clel OU semi-preparative colurnn (250 mm x 10 mm)
with MeOW1û??H20as the mobile phase at the flow rate of 6 mUrnin. Collected fiactions
were stripped of the solvent on a rotary evaporator and subjected to analysis by ES-MS.
3 -4.14. Preparation of Samples of Photolvsis of 2.3.7.8-TCDD for Toxicological Assavs
A solution of 2,3,7,8-TCDD in hexane (2.40 mL, 7.8~10" mol LS)was placed in
8 mm Pyrex0 tube and irradiated at 300 nm in the merry-go-round photoreactor.
Aliquots (200 pL,) were taken afier O, 0.5, 1, 2, 4, 8, 12, 20, 35, and 60 minutes and
placed in 1 mL sample vials. Twenty five microlitres were drawn fkom each aliquot and
mixed with 25 pL of 5x 104 HCB in hexane. The mixtures, prepared fiom 0, 2, 8, 12, 20,
35 and 60 minute samples were analysed by GC-MS with the injections done in triplicate.
The measured concentrations of 2,3,7,8-TCDD were averaged, and the first-order rate
constant was determined fiom the plot of h(C&) a t. The concentrations of 2,3,7,8-
TCDD in solution (Table 53) were calculated using the following equation:
The solvent in the sample vials (hexane, 175 pL) was carefully removed under a slow
Stream of nitrogen and the contents were redissolved in 50 p.L of DMSO (Fisher,
Spectranalyzed Grade). Table 53. Measured and Calculated Concentrations of 2,3,7&TCDD During Photolysis at 300 nm Measured Concentrations
Time, sec Concentration, mol L-' Conversion, % (cdc')
Caicuiated Concentrations* Percent Time, sec Concentration, mol L-' Conversion, % Remai ning
* - The first order constant for disappearance of 2,3,7,8-TCDD was 2.364~10~~with 2 = 0.99 1. 3 -4.15. HAP Assav with the Products of Photolvsis of 2-3.7.8-TCDD
The procedure of Gasiewicz and ~eal~'was followed without major
modifications. Each sample was prepared in quadniplicate. Stock solutions of 2,3,7,8-
TCDD in DMSO (Section 3-4-14}were diluted 1000 times before use. The concentration
of 1,6-[3~-2,3,7,8-~~~~was 1x10' mol L-'. Aliquots of cytosol with 10 pL of [3H+
TCDD and 10 pL of each photolyzed solution were incubated for 45 minutes at 23 OC.
This was followed by analysis of the binding of I~W-TCDDto the Ah receptor with the
HAP assay." Aliquots of cytosol incubated with ~H+TCDDalone gave a value for
"total binding", while samples incubated with E~H]-TCDDand a 200-fold excess of
TCDF were used to determine "non-specific" binding. The quadruplicate values of the
activities measured by scintillation counting, were averaged. The percent specific binding
of [3H+~~~~to the Ah receptor (Table 54) was then calculated using the following
equation:
% Specific binding = [sample @PM - non-specific (WPhflJ/[total@PM) - non-specific (DPW] x 100
Table 54. HAP Assay with the Products of Photolysis of 2,3,7,8-TCDD Concentration of 2,3.7,8-TCDD, mol L-' Activity, DPM* Specific Binding, % Conversion, %
60 1845.5 88.9 99.98 * - The total bindïng activity was 1923 DPM, the non-specific binding activity was 170 DPM 3 -4.16. Estradiol Assay with the Products of Photolvsis of 2-3.7.8-TCDD
A gel filtration chromatographie method for determining relative estrogenic
binding afinities, as developed by Cox and unc ce,^' was used without alterations. The
concentration of 2,4,6,7-[3~-estradiol(fiom Amersham, specific activity of 87 Ci/mmol)
in the assay was 1.5~10" mol L". Five samples of the photolyzed 2,3,7,8-TCDD (Section
3.4.14) were used in the assay (10 pL of the solutions in DMSO were taken). The final
concentrations of the produas of photolysis of 2,3,7,8-TCDD are shown in Table 55. Rat
liver cytosol was obtained from six female Long Evans rats. The HPLC setup was as
described in Section 3.4.12 with the exception that a Phenomenex Biosap S400 size
exclusion column (306 mm x 7.8 mm) was used. For each concentration point 0.5 mL fractions were collected at the flow rate of 0.5 amin for 12 minutes, starting the collection 6.5 minutes afier the injection. The rneasured activites for each point are presented in Table 55. Table 55. Estradio1 Assay with the Products of Photolysis of 2,3,7,8-TCDD Activiîy, DPM IProdu~l, - O 3.58~10" 1.17~lo4 2.21~lo4 2.7 1x 104 mol L-' Elution Total 0% 13-2% 43.3% 8 1-8% 99.9% time, min Binding Conversion Conversion Conversion Conversion Conversion 6.5 42 29 34 55 43 51 Biblioeraphy for Cha~terIII
1. Tokar, B. ZMagazine, 1996,5, 50.
2. Safe, S. Crit. Rev. Toxicol., 1990, 21, 51.
3. Whiteside, T. The Pendulm and the Toxic Cloud, YaIe University Press Ltd., London,
1979, pp 3 1-66.
4. Bunce, N. J. Environmental Chistry, Wuen Publishing Ltd., Winnipeg, 1994, p 3 15.
5. Deegan, J., Jr. Environ. Sci. Technol. 1987, 21, 328.
6. Safê, S.; Bandiera, S.; Sawyer, T.; Robertson, L.; Safe, L.; Parkinson, A-; Thomas, P.
E.; Ryan, D. E.; Reik, L. M.; Levin, W.; Denomme, M. A.; Fujita, T. Dzviron.
Healfh Perspect., 1985, 60, 47-56.
7. Sde, S. Environ. Carcinog. & EbotoxicoI. Rev., 1991, C9,26 1-302.
8. Safe, S. Crif. Rev. Toxicol., 1994, 24, 87-149.
9. Rowlands, J. C.; Gustafsson, I-A. Crif. Rev. ToxicoL, 1997,27,109-134.
10. Johnson, ES. Emiron Healfh Perspecf., 1993, 99,383.
11. U.S. Department of Health and Human Services In: Toxicological profile for
chlorinated dibenzo-p-dioxins, Dr& for public comment, 1997, 17.
12. Crow, K. D. Ann. Occup. Hyg., 1978,2I, 297-298.
13. Wentz, C.A. Hazardms Wmte Management, 2nd Ed., McGraw-Hill Inc., Toronto,
1995, p 253.
14. Tysklind, M. Mziltitative Chernical Characterizaiion and Modelling of
Poiychloriinaled Dioxins and Dibenrofrans , Softjarden Offset AB, Umea, 1993, p 6.
15. Bnuzy L.P. ; Hites, R. A. Environ. Sci. Technol., 1996, 30, 1797. 16. Environment Canada and the FederaVProvincial Task Force on Dioxins and Furans,
Dioxins md Furans and Hexachorobenzeene Invetzfory, Dr& 4, April 1998.
17. TEQ or Toxic Equivalent is the additive value based on the internationally adopted
Toxic Equivalent Factors (TEF) masure system developed to express toxic potency
of PCDD/PCDF mixtures.
18. Environment Canada, Scienn)ïc JwtliJcananon,PCDD and PCDF, Candidate
Subsfancesfor Management under Track I of the ISMP,Ma& 1997.
19. General-Waste Management R.R.O. 1990, Reg. 347. Revised Regulations of
Ontario, 1990, Amended to O. Reg. 5 12/95
20. Esposito, MT.; Tieman, T.O.;Dryden, F.E. Dioxins, US EPA Report EPA-600/2-
80-197, 1980., p. 257.
2 1. Superfùnd Innovative Technology Evaluation Program. Technology Profiles. 8th
Edition. U.S. Environmental Protection Agency. October 1995. pp 59-409.
22. Dreschler, W.D. ChernoJphere, 1986, 15, 1529.
23. Anon Ew. ChNews, 1986, JO, 20
24. Lissel, H., Kottrnann, J., Chemuqhere, 1989, 19, 1499.
25. Tundo, P., Facchetti, S., Tumiatti, W., Fortunari, U.G., Chemosphere, 1985, 14, 403.
26. Superfûnd Innovative Technology Evaluation Program. Technology Profiles. 9th
Edition. U.S. Environmental Protection Agency. December 1996. pp 320-321.
27. Mill, T. In Dioxzns Nz the Environnzent, Eds: Kamrin, M.A.; Rodgers, P.W.,
Hemisphere Publishing Corporation, 1985, p 173.
28. Pennise, D.M.; Kamens, R.M. Environ. Sci Technul. 1996,30,2832. 29. Choudhry, G.C.; Webster, G.RB. Chemo~here,1985, 14, 893.
30. Crosby, DG.; Wong, AS.; Plimmer, J.R.; Woolson, E-A Science, 1971,173, 748.
3 1. Crosby, DG.; Moilanen, K.W.; Wong, AS. Environ. Health Perspeci., 1973,5, 259.
32. Koester, C.f .; Hites, R Environ. Sci. Techtol. 1992, 26, 502.
33. Muto, H-; Takizawa, Y. Chem. Leff., 1991,233.
34. Choudhry, G.G.,and Webster, G.R.B., 1 Agric. Food Chem., 1989,37,254.
35. Choudhry, G.C.; Webster, G.R.B. Chemosphere, 1986,15, 1935.
36. Choudhry, G.G.;and Webster, G.R Chemosphere, 1985, 14,9.
37. Dulin, D.; Drossman, H.; Mill, T. Environ. Sci. Technol., 1986,20, 72.
38. Hung, L.S., Ingram, L.L., Jr., Bull. EEnviron. Contam. Toxicol., 1989, 44, 380.
39. Nestrick, T.J.; Lamparski, L.L.; Townsend, D.I. Anal. Ch.,1980, 52, 1865.
40. Dobbs, A.J., Grant, C. Nature, 1979,278, 163.
41. Koshioka, M.; Ishizaka, M.; Yamada, T.; Kanazawa, J.; and Murai, T. 1Pesticide
Sci., 1990, 15, 39.
42. Chu, W.; Jafiert, C. T.; Diehl, C. A; Marley, K.; Larson, R. A. Eizviroil. Sci. Tech.,
1998, 32, 1989.
43. Bolton, J.R.; Cater, S.R. In Apufic and Surface Phofochemistry, CRC Press,
Toronto, 1994, pp. 467-490.
44. ~oiton,J.R.; Amiri, A.; Cater, S.R. In Alternative Ftrels and Environment, Ed.,
Sterret, F.S., Lewis Publishers, Boca Raton, FL, 1995, pp. 187-192.
45. Pignatello, J. J.; Huang, L.Q.Wat. Res., 1993, 2 7, 173 1.
46. Ritterbusch, J.; Lorenz, W.; Bahadir, M.; Chemoqdiere, 1994, 29,457. -
47. Ritterbusch, J.; Lorenz, W.; Bahadir, M.; Chemosphere, 1994, 29, 1829.
48. Persona1 communication with Mr. Brock Chittam, the owner of Wellington
Laboratories.
49. Murov, S.L.; Carinichael, L; Hug, G.L. Handbook of Photochemistry, 2nd Ed.,
Marcel Dekker, Inc. NY, 1993. Table 1, pp. 4-53 .
50. Konstantinov, II;Bunce, N.J. J. Photochem Photobzol., A: Chern., 1996,9427.
5 1. Buser, HR1 Chromatogr-, 1976,129,3 03.
52. Crosby, D.G.; Wong, AS.; Plirnrner, J.R.; Woolson, E.k Science, 1971, 173, 748.
53. Bunce, N.J. In Hand7~ookof Organic Photochemistry, Ed. Horspool, W.M., CRC
Press, Boca Raton, FL, 1995, Chapter 86.
54. Crosby, D.G.; Wong, AS. Science, 1977, 195, 1337.
55. Nash, R.G.; Beall, M.L., Jr. J. Agric. Food Chem., 1980, 28, 614.
56. Epling, G.A; Qui, Q.; Kumar, A, Chemo.phere, 1989, 18,329.
57. A different reaction mechanism such as a radical-chah reaction may have been
involved in this case.
58. Hosoya, K.; Kirnata, K. Fukunishi, K.; Tanaka, N. Chemo@ere, 1995, 31., 3687.
59. Stormer, K. A. In Demxication of Hazardous Wmte, Ed. Exner, J.H., Ann Arbor
Science, Ann Arbor, MI., 1982, p 280.
60. Kleatlwong S.; Nguyen, L.V.; Herbert, V.R.; Hackell, M.; Miller, G. Environ. Sci-
Technol., 1990, 24, 1575
6 1. Wagenaar, W.J.; Boelhouwers, E.J.; de Kok, H.A.M.; Groen, C.P.; Govers, H.A.J.;
Olie, K.; de Gerlache, J.; de Rooij, C.G. Chemosphere, 1995, 31, 2983. --
62. Mamantov, A. Chemosphere, 1985, 14,897.
63. Minghui, 2.;Zhicheng, B.; Xiaobai, X.; Keou, W. Chemosphere, 1996,32,603.
64. Fnesen, K.J.; Muir, D.C.G.; Webster, G.R.B.Environ. Sci. Technol., 1990,24, 1739.
65. Ogat, Y. Tagai, K. Ishino, 1. Tetrahedon, 1970, 26,2703.
66. Plimmer, J.R.; Klingebiel, U.I.; Crosby, D.G.; Wong, AS. In Chlorodoxins, 0rig.l-n
andFate, Ed.: Blair, E.K, - ACS, Washington, D.C.,1973, p. 49.
67. Massé, R.; Pelletier, B. Chemoqhere, 1987, 16, 7.
68. Guan, B.; Wan, P. J. Chem. Soc.,Chem Commun., 1993,409.
69. Guan, B.; Wan, P. J. Photochem Photobiol., A: Chem., 1994,80, 199.
70. Gasiewicz, T.A.; Neal, R.A. Anal. Biochem,, 1982,124, 1.
7 1. Cox, B.J. ; Bunce, N.J. Anal. Biochem., 1998, acceptedfor publication. CHAPTER IV
GLOBAL CONCLUSIONS
Most of the objectives stated at the begiming of this thesis were accomplished.
In the study of successive photocyanation of highly chlorinated aromatic compounds we demonstrated that the reaction is common for various classes of aromatic compounds. A number of similarities were found for different compounds under study.
The products in photolysis of polychlorinated aromatics in the presence of sodium cyanide were polycyanated hydroxychlorocompounds with various degrees of chlorine replacement; with few exceptions complex product mixtures were formed. The produas can only be isolated in the form of highly polar phenoxide sodium salts; this constitutes the other major hindrance for the synthetic development of the reaction. Products fiom some substrates were identified and characterized.
Significant information regarding the niechanism of successive photochemical cyanation was gathered. Quantum yields of disappearance were measured for a number of compounds. The quantum efficiency of photocyanation of chlorinated benzenes was found to increase with the number of chlorine substituents on a substrate. The reaction under study also appeared to be an excellent example of autocatalysis in photocyanation since each successive cyanation proceeds with higher quantum yield.
Photophysical aspects of the cyanation were studied in detail. The Sp~2Ar* mechanism was consistent with the observations on the photocyanation of polychlorinated aromatics. Sensitization and quenching experiments established the triplet excited state to be reactive for al1 tested compounds, consistent with the suggested mechanism. At its present state the synthetic potential of the reaction is low. The future work on successive photocyanations may include study of the reaction in anhydrous media. In the water-fiee system complete substitution of chlonnes by cyano groups may be feasible since the hydroxide ions capable of the early termination of photocyanation is absent.
This will increase the reaction selectivity. Moreover, it would be easier to handle, separate and characterize the formed polycyanoarornatics. A number of other developments of the research are possible; for example,. a detailed study of products and mechanisms for the photoreactions of polyfluoro- and polybromoaromatics with cyanide ion.
We developed a method for on-site destruction of dioxins in small quantities of liquid laboratory waste using UV light. The equipment used for destruction is inexpensive and easy to setup; the light source, a low pressure mercury PenRayO lamp is free of the fire hazard associated with the high-wattage UV lamps. The method was demonstrated to be efficient on actual waste samples.
The problem of missing material in photolysis of dioxins was resolved by using tritiated 2,3,7,8-TCDD as a tracer compound. Products of photolysis of 2,3,7,8-TCDD were identified by ES-MS as hydroxylated aromatic compounds. With the aid of scintillation counting we estimated that the yield of such products may be as high as 80%.
This evidence, along with observed small yields of reductive dechlorination products presented a solid proof that C-O rather than C-Cl homolysis is the major reaction path for dioxins.
Biological assays on the products of photolysis of 2,3,7,8-TCDD served as an excellent example of successful interdisciplinary cooperation. The Ah receptor assay demonstrated that the produas formed in photolysis of 2,3,7,8-TCDD lose receptor binding ability and hence this information is very important for the practical applications of the UV treatment methods.
The products of photolysis of 2,3,7,8-TCDD could exhibit estrogen receptor binding ability. We were able to test this possibility by the estrogen receptor binding assay. The products did not exhibit strong binding affinity. We could not detekne whether or not some weak binders were present. Overall, the results of binding assays indicate that the TCDD photoproducts do not exhibit toxicity associated with the Ah and estrogen receptors. It is not known whether or not any other types of toxicity may be associated with these photoproducts, although the following information is worth mentioning.
The structure of one of the major products (45) tentatively identified in photolysis of 2,3,7,8-TCDD is similar to that of ri clos an' (2,4,4'-trichloro-2'-hy droxydiphenyl ether), an antibacterial ingredient of Colgate-Palmolive's "Total@" toothpaste2. This piece of information does not suggest that the photolysis of dioxins can be used in production of a toothpaste component but rather indicates that the primary products of photolysis are non-toxic.
The future research on PCDWPCDF photolysis may include the following topics:
- structural identification of major photoproducts of 2,3,7,8-TCDD photolysis
(rnost feasible by cornparison to authentic samples)
- study of the secondary photoreactions (using major primary photoproducts as
mode1 substrates) - elucidation of the reaction mechanism (both photochernical and
photophysical) for photolysis of 2,3,7,8-TCDD in hydrogen-donor solvents
- development of a larger-scaie method for dioxin-containing laboratory waste
detoxification (for example, using a 15W medium pressure herbicidal lamp as
the W source) BibIiomaphv for Chapter IV
1. Bhargava, & N.; Leonard, P. A. Am. J. Infect. Control, 1996, 24, .209.
2. Jackson, E. M. Cosm. Dermat., 1998,1423. APPENDIX A
1H NMR Spectrum of Compound 42 APPENDIX B
HSQC Spectrum of Compound 42 APPENDIX C
HRMS Chromatograrns of PCDDFCDF Congeners from UV-treated Dioxin-containing
Laboratory Waste
80 i i , ,.8. , ? 60-l \ 6.0 HRS 11.3~6 4 zo 4 , i \ ::ig F o-om O 1 L . ;,. L - 24:bO ' ' ' 25:bO' 26:OO 27:bO 28:bO 29:bO Tirne 100 % ,LSE5 Ii E 12, 27 HRS 60 40 6.1 E4 20 2 F9-2-3.1E4
O--, O--, --.-.-..- , L~~O.OE0 24:ûû 25:bO 26:ûO 27:bO 28:bO 29:OO Time C-1. TCDD File:3O0CT_PH #1-246 Acqi30UCT-1998 11:0825 GC EIt Voimgc SIR 70SE Sample#3 Texts2UL-PHûTOLYSIS-OHR 7- -""""""-- ' 1WT A
1.5 HRS
3.5 HRS
6.0 HRS file:30ocT-PH # 1-141 Acq:30-OCT-1998 11:08:25 GC EI-t Voitage SiR 70SE SampItrW Tcxt:2UL-PHOTOLYSISOHR Exp:PCBPCDD-M Io0 24 80 z OCDD
1.5 HRS
, , O- ... , , r_O.OEO 44:12 4424 ' 44:36 ' 44:48 '45:bO ' ' ' 45:12' ' 45:24 45h6 45:48 Time Io0 56 F~-~~~ 80 : 1.9E4 27 HRS ' .. 60: 40: Es:::
2o j,-Lm- ~-/J--.A 14.8E3 , , , O- , , . r 0-OEO 44:k . ' '&24 ' 44~36' ' 44148 45: 45:12 45:24 '45:56 ' '45:48 Tirnc
C-3. OCDD File:30ûCT-PH #I422 Acq:30-OCl-1998 l l:08:25 GC EI+ Voltage SIR 70SE Sample#3 Text:2Ut-PHOTOLYSIS-OHR Exp:PCBPCDD-M 100 %
80 _! 1 I 1 1t 1-8E7 i \ 1 1' !- 60 i TCDF ., OHRS k 1.3~7 a 1 40 1 4, 1 , . k9.0~6r 20 i .\ 14.5~6 \ .:: , \ - (7 0 p. J ,, L J IL ' I ' ,:o.oEO 23:Oo 24:OO 2500 26:b 27:OO 28:h 29:b Time LOO % 1 ,I -5M
. . ,!.. . 13.0~6 \,'b :-, .. 4, . A--. .u ,,,.0 F 0.0~0 26:ûû 27:OO 28:h' 29:bO Time LOO X
L 12.3E6 F 6.0 HRS L 1 -7E6 l.LE6
!\ 15.6E5 AtO.OE0 23 :O0 24: 29:oOI - Time 100 1 -2.OE4
C-4. TCDF File:30Off-PH #1-246 Acq:30-Off-1998 1l:08:25 GC EI t VoI~geSIR 70SE Sampie#3 Tcxt:2UL-PHOTOLYSISOHR Exp:PCBPCDD-M 100 98
O HRS
1.5 HRS 11.7E7 L t i.rn
3.5 HRS file:3OOcT_PH 81-141 Acq:30-OCT-1998 11:08:2S GC EI+ VoIcage SIR 70SE SampW3 Text:2UL-PHOTOLYSISoHR Exp:PCBPCDD-M 100 96 8.8E6 1 80 4 7.lE6 OCDF ., . 40 4 3.5E6 1 t 20; k 1.8~6 1 O- 1 O.OU) m:48Time LOO 96 ,l.lE5 2 r 80 4 L8.9~4
O - F 0.OEO - '44:iz 4&24 44:36 ' 45:b ' '4ki2 45:24 4536 45:48 Timc 100 46 1.7E4 80 ; i\ r
O-, , . F O.OE0 44:I2 ' &:i4' ' 44:36 ' ' U:48 ' '45:6 ' '45:ii 45:24 45:36 ' 45:48 Time 100 % - -2.lE4
C-6. OCDF AE'PENDIX D
ES-MS Spectra of Products of Photolysis of 2,3,7,8-TCDD in Hexane at 300 nm
D- 1. Products before Attempted Separation
D-3. Flash Chromatography, Fractions 5-8 D-4. Flash Chromatography, Fractions 1 1-1 2 Scan ES- 4,5407
23,
b 11-i Scan ES- 3.0287