Utah State University DigitalCommons@USU

Reports Utah Water Research Laboratory

January 1987

Engineering Treatment of Hazardous Wastewaters Utilizing Dye- sensitized Photooxidation

Betty-Ann Naeger

R. Ryan Dupont

William M. Moore

Follow this and additional works at: https://digitalcommons.usu.edu/water_rep

Part of the Civil and Environmental Engineering Commons, and the Water Resource Management Commons

Recommended Citation Naeger, Betty-Ann; Dupont, R. Ryan; and Moore, William M., "Engineering Treatment of Hazardous Wastewaters Utilizing Dye-sensitized Photooxidation" (1987). Reports. Paper 190. https://digitalcommons.usu.edu/water_rep/190

This Report is brought to you for free and open access by the Utah Water Research Laboratory at DigitalCommons@USU. It has been accepted for inclusion in Reports by an authorized administrator of DigitalCommons@USU. For more information, please contact [email protected]. Engineering Treatment of Hazardous Wastewaters Utilizing Dye-sensitized Photooxidation

Betty-Ann Naeger R. Ryan Dupont William M. Moore

Utah Water Research Laboratory Utah State University Logan. Utah 84322-8200 WA TER QUALITY SERIES November 1987 UWRL/Q-87/01 ENGINEERING TREATMENT OF HAZARDOUS WASTEWATERS UTILIZING

DYE-SENSITIZED PHOTOOXIDATION

Betty-Ann Naeger, R. Ryan Dupont, and William M. Moore

WATER QUALITY SERIES UWRL/Q-87/01

" Utah Water Research Laboratory Utah State University Lo~an, UT 84322-8200

November 1987 ABSTRACT

Studies were conducted to determine the applicability of photooxidat ion for the degradat ion of select ed hazardous and refractory organic compounds. These photochemical oxidation react ions occur through the trans fer of energy from elect ron­ ically excited sensitizer molecules which attain excited states by absorbing visible light energy.

Optimum conditions for photooxidation were established based on sensitizer concentration and reaction pH for four polynuclear aromatic pollutants.

The rate of photooxidation was found to be independent of the initial substrate concentration for methylene blue-sensitized reactions, and dependent on substrate concentration for solutions without a sensltlzins dye. Photolysis of substrat e mixt ures established acridine and anthracene as photochemically active substrates.

Photochemical reaction data suggest predictable trends in substrate reactivity based on pKa values of both sensitizer and substrate, initial substrate concentration and light absorbance characterist i'cs.

The photoproducts formed during the photolysis of acridine were found to be more toxic than the parent compound. These reaction products appear to be stable and warrant further study.

iii TABLE OF CONTENTS

Page

INTRODUCTION

OBJECTIVES 3

LITERATURE REVIEW 7

Shale Oil 7 Retort Wastewater 9 Polynuclear Aromatic Hydrocarbons 12 Treatment Options 13

Conventional processes 13 Dye-sensitized photooxidation 14

Actinometry 19

. MATERIALS AND METHODS 23

Photolysis Bench 23 Reactors 25 Photolysis 25

Optimum reaction conditions 25 Effect of substrate mixtures 27 Effect of initial substrate concentrations 27

Actinometry 27 Extraction Procedures 29 Constituent Analysis 29 Toxicity Testing 29 Statistical Analysis 30

EXAMPLE CALCULATIONS 3 J

Substrate Reaction Rate Constants 31 Actinometry 31

Actinometer reaction rate constants 31 Light energy 31 Average incident light intensity 33 Actinometer correction factor 33 Quantum yield 33

RESULTS 35

Optimum Reaction Conditions 35

Acridine 35 Anthracene 44

v TABLE OF CONTENTS

Page

Quinoline 47 Naphthalene 52

Initial Substrates Concentration 52 Substrate Mixtures 56 Substrate Half-Lives 56 Actinometry 60 Toxicity Testing 62

DISCUSSION 65

Optimum Reaction Conditions 65

Anthracene photodimerization 65 Effect of sensitizing dye 68 Effect of pH 70

Initial Substrate Concentration 71

Anthracene reaction 71 Acridine reactions 73

Effect of Mixing Substrates 73 Analysis of Toxicity 75 Prediction of Reaction Rates 76

ENGINEERING SIGNIFICANCE 79

Design of a Sensitized Photooxidation Lagoon 79

Determination of flowrate 79 Design considerations 79 Actinometry 80 Determination of rate constants 80 Design volume 81 Design area 81

Effluent Concentrations 82 Chemical Costs 83

SUMMARY AND CONCLUSIONS 85

RECOMMENDATIONS FOR FURTHER STUDY 87

REFERENCES 89

APPENDIX A: SUPPORTING DATA--REACTION RATE CONSTANTS 95

APPENDIX B: SUPPORTING DATA--REACTION QUANTUM YIELDS • 10 I vi LIST OF FIGURES

Figure Page

l. Schematic representation of three surface oil shale retorting processes (U.S. Dar 1973) . 8

2. Schematic representation of an in-situ oil shale r~torting process, a) vertical cross section; b) plan-view (iowacki 1981) 10

3. Molecular oxygen states and their electronic configuration (Foote 1968) 16

4. Radiation as a function of wavelength emitted from the DuroTest Vitalite lamps 24

5. Percent of total light intensity occurring at each wavelength for natural sunlight and artificial light 24

6. Absorbance spectrum for a 3 mg/l methylene blue solution 28

7. First order kinetic decay constants for the photolysis of 5 mg/l acridine solutions as a function of pH; comparison of methylene blue concentrations of 10, 5, 2, I, and 0 mg/l . 38

8. First order kinetic decay constants for the photolysis of 5 mg/l acridine solutions as a function of methylene blue concentration; comparison of pH 3, 5, 7, 9, and I I . 39

9. Reaction quantum yields for the photolysis of 5 mg/l acridine solutions as a function of pH; comparison of methylene blue concentrations of 10, 5, 2, and I mg/l 39

10. Reaction quantum yields for the photolysis of 5 mg/l acridine solutions as a function of methylene blue concentration; comparison of pH 3,5, 7,9, and II 41

11. First order kinetic decay constants for the photolysis of I mg/l anthracene solutions as a function of pH; comparison of methylene blue concentrations of 5, 2, and 0 mg/l 45

12. First order kinetic decay constants for the photolysis of J mg/l anthracene solutions as a function of methylene blue concentration; comparison of pH 3,5,7,9, and II 46

13. Reaction quantum yields for the photolysis of 1 mg/l anthracene solutions as a function of pH; comparison of methylene blue concentrations of 5 and 2 mg/l 49

14. First order kinetic decay constants for the photolysis of 5 mg/l quinoline solutions as a function of pH; comparison of methylene blue concentrations of 5 and 0 mg/l 50

vii LIST OF FIGURES (Continued)

Figure Page

15. Reaction quantum yields for the photolysis of 5 mg/l quinoline solutions as a function of pH in a 5 mg/l methylene blue solution (95 percent confidence limits shown for each data point) 51

16. First order kinetic decay constants as a function of initial anthracene concentration; comparison of methylene blue concentrations of 2 and 0 mg/l 54

17. First order kinetic decay constants as a function of initial acridine concentration in a 2 mg/l methylene blue solution (95 percent confidence limits shown for each data point) . 55

18. Reaction quantum yields as a function of init ial anthracene concentration ~n a 2 mg/l methylene blue solution 57

19. Reaction quantum yields as a function of initial acridine concentration in a 2 mg/l methylene blue solution 58

20. Methylene blue absorbance spectrum compared with the light source emission spectrum 61

21. Jablonski's diagram 66

22. Absorbance spectrum for a ) mg/l anthracene solution 67

23. Photodimerization of anthracene 68

24. Absorbance spectrum for 5 mg/l acridine solutions,S mg/l quinoline solutions and ) mg/l anthracene solutions 74

viii LIST OF TABLES

Table Page

I. Characteristics of selected polycyclic aromatic hydrocarbons meeting criteria for this study 4

2. Physical characteristics of selected polycyclic aromatic hydrocarbons 5

3. Water quality of oil shale retort wastewater (Nowacki 1981) I I

4. Buffer solutions utilized for investigation of the effect of pH on the sensitized photolysis rates 26

5. Experimental design configuration 26

6. Photolysis data from the experiment involving anthracene with 5 mg/l methylene blue and buffered to pH 9; Run 2 32

7. Photolysis data representing experimental conditions and results for the example photolysis run 32

8. Percent light absorbed in solutions of varying depth and methylene blue concentrations 34

9. First order kinetic decay constants (I/hour) for the photolysis of 5 mg/l aqueous solutions of acridine 36

10. Reaction quantum yields (moles/einstein) for the photolysis of 5 mg/l aqueous solutions of acridine 40

11. Comparison of treatment efficiencies (%) for the photooxidation of 5 mg/l solutions of acridine during a 48-hour photolysis period 42

12. Shaded representation of statistical equality among the observed reaction rate constants for the photolysis of acridine 43

13. Shaded representation of statistical equality among the observed reaction quantum yields for the sensitized photolysis of acridine . 43

14. First order kinetic decay constants (l/hour) for the photo- oxidation of 1 mg/l solutions of anthracene 44

15. Comparison of treatment efficiencies (%) for the photooxidation of ) mg/l solutions of anthracene with dye for a 24 hour irradia- tion period and without dye for a 50 minute irradiation period. 48

\6. Reaction quantum yields (moles/einstein) for the photooxidation of I mg/l aqueous solutions of anthracene 48

ix LIST OF TABLES (Continued)

Table Page

17. First order kinetic decay constants (l/hour) for the photolysis of 5 mg/l aqueous solutions of quinoline 50

IS. Removal efficiencies (%) for the photolysis of 5 mg/l solutions of quinoline during a 72 hour photolysis period 51

19. Percent of naphthalene lost due to volatilization under photolysis reaction conditions with light eliminated; mean values from duplicate analyses 52

20. First order kinetic rate constants (I/hour) for the photolysis of 1 mg/l aqueous solutions of naphthalene 53

21. First order kinetic decay constants (l/hour) from the examination of the effect of initial anthracene concentration on the rate of photolysis 53

22. A comparison of the first order decay reaction rate constants (I/hour) for substrate mixture (acridine, 5 mg/l; anthracene, 1 mg/l; and quinoline, 5 mg/l) with the photolysis rates of each compound treated individually 59

23. A comparison of quantum yields (moles/einstein) for substrate mixtures (anthracene, 1 mg/l; acridine and quinoline, 5 mg/l) with the quantum yields of each compound photolyzed individually. 59

24. Calculated half-lives for the photolysis of anthracene under different treatment conditions 59

25. Calculated half-lives for the photolysis of acridine under different treatment conditions 60

26. Calculated half-lives for the photolysis of quinoline under different treatment conditions 60

27. Results of the toxicity analysis utilizing the Microtox acute toxicity test 63

2S. Optimum reaction conditions for the photodegradation of 5 mg/l solutions of acridine and quinoline and 1 mg/l solutions of anthracene 66

29. Change in the concentration of anthracene with time when photolyzed without a filter and with a 420 nm filter 6S

30. Results from tryptophan-methylene blue actinometer performed under natural light conditions on a cloudy day at 420 N latitude SI

][ LIST OF TABLES (Continued)

Table Page

31. Results from triplicate analyses of 5 mg/l acridine solutions; reaction rate constants from sensitized and direct photooxidations 96

32. Results from triplicate analyses of I mg/l anthracene solutions; reaction rate constants from sensitized and direct photooxidations 98

33. Results from triplicate analyses of 5 mg/l quinoline solution; reaction rate constants from sensitized and direct photooxidations 99

34. Observed reaction rate constants from the photolysis of a mixture containing anthracene, acridine and quinoline 99

35. Observed reaction rate constants from experiments investigating the effect of initial anthracene concentration on the rate of photolysis 100

36. Observed reaction rate constants from experiments investigating the effect of initial acridine concentration on the rate of photolysis 100

37. Summary table from triplicate analyses of the photolysis of 5 mg/l acridine solutions; average incident light intensity, reaction rate constants from actinometry, reaction rate constants from photolysis, absorbance correction factor, and reaction quantum yields 102

38. Summary table from triplicate analyses of the photolysis of 5 mg/l acridine solutions; average incident light intensity, reaction rate constants from actinometry, reaction rate constants from photolysis, absorbance correction factor, and reaction quantum yields 103

39. Summary table from triplicate analyses of the photolysis of 5 mg/l acridine solutions; average incident light intensity, reaction rate constants from actinometry, reaction rate conscants from photolysis, absorbance correction factor, and react ion quantum yie ids 104

40. Summary table from triplicate analyses of the photolysis of I mg/l anthracene solutions; average incident light intensity, reaction rate constants from actinometry, reaction rate constants from photolysis, absorbance correction factor, and reaction quantum yields 10'j

xi LIST OF TABLES (Continued)

Table Page

41. Summary table from triplicate analyses of the photolysis of I mg/l anthracene solutions; average incident light intensity, reaction rate constants from actinometry, reaction rate constants from photolysis, absorbance correction factor, and reaction quantum yields 106

42. Summary table from triplicate analyses of the photolysis of 5 mg/l quinoline solutions; average incident light intensity, reaction rate constants from actinometry, reaction rate constants from photolysis, absorbance correction factor, and reaction quantum yields 107

43. Summary table from triplicate analyses examining the effect of combining substrates on the yields and reaction rates; mixtures were photolyzed in 2 mg/l methylene blue solutions and buffered to a pH of 7 108

44. Summary table from the triplicate analyses examining the effect of initial anthracene concentration on the reaction rates and yields; average incident light intensity, reaction rate constants from actinometry, reaction rate constants from photolysis, absorbance correction factor, and reaction quantum yields in 2 mg/l methylene blue solutions 109

45. Summary table from the triplicate analyses examining the effect of initial acridine concentration on the reaction rates and yields; average incident light intensity, reaction rate constants from actinometry, reaction rate constants from photolysis, absorbance correction factor, and reaction quantum yields in 2 mg/l methylene blue solutions 110

xii INTRODUCTION

The unpredictability of the inter­ estimated by Fox et al. (1980) to be national energy market and the potential from 78 to 278 gallons per barrel oil, of global energy shortages in the next d·epending on the retort technology few decades have convinced most national utilized. In addition to the high energy planners of the long-term need to industrial demand for water, the Green develop major new energy sources. One River Format ion is located in a water means of supplementing domest ic oil and deficient area, much of the limited gas deposits is with synthetic fuels groundwater and surface water is moder­ derived frOOl fossil fuel sources. The ately to highly saline (Israelsen et al. largest untapped fossil fuel resource in 1980), and water rights are difficult to the United States is the oil bearing obtain. Present plans call for the shale in the Green River Formation si t ing of many s hal e oil produc t ion located in Colorado, Utah, and Wyoming. facilities along major western rivers This area contains one of the largest which are to be used as a water source deposits of hydrocarbons in the world and wastewater sink. and is projected to be a potential source of more than two trillion barrels The rate of oil shale production of shale oil (Yen 1978). may be limited by water availability (Maase 1980, Maase and Adams 1983) When energy resources are ex­ and it may be imperative that the large tracted, processed, converted, and volumes of wastewater produced during utilized, the related impacts to human production be treated for reuse within health and the environment often require the plant. The wastewater may be a new and increasingly more efficient valuable resource for the arid regions pollution control methods. New syn­ if effect ive and economical treatment thetic fuel processes, including oil methods are found. The shale oil extraction from shale, are expected to industry, to be economically competitive have unique air, water and solid waste with the other fuel industries, requires control requirements. low cost water pollution control tech­ nologies, capable of supplementing The Department of Energy has conventional treatment methods, in order established a research, development and to substantially reduce or eliminate demonstration (RD&D) program for en­ contaminants in oil shale processing couraging the development of this wastewaters. country's oil shale resource (Hartstein and Harney 1981). The aim of the oil Conventional biological and chemi­ Shale RD&D Program is to stimulate the cal pollution control technologies commercial development of shale oil by will sufficiently handle the biological eliminating technical and environmental and inorganic pollutants produced. barriers to its produc tion. Among the However, numerous researchers (Fox key environmental needs that were et a1. 1980, Mercer 1980, and Klieve outl ined, development of effic ient oil et al. 1981) have shown biological shale wastewater treatment systems was treatment by activated sludge to be strongly stressed. ineffective in sufficiently reducing total organic carbon for reuse or Development of oil shale resources discharge. Treatment processes, i n­ will require large quantities of water, eluding activated carbon, solvent extraction, ion exchange, reverse sensitizing molecules, or triplet osmosis, air stripping and coagulation, sensitizers, return to ground state by have major technical and/or economic transferring energy to molecular oxygen limitations. Photooxidation of hazard­ or organic substrates present in the ous organic wastewater may be a cost­ system. If dissolved oxygen is present effective process capable of degrading in even trace amounts, a triplet trace organic compounds since sunlight sensitizer will selectively transfer its needed to drive the decomposition energy to oxygen due to its low energy reaction is available at no cost. requirement in forming singlet oxygen (Foote 1958). Singlet oxygen is a Photooxidation reactions can be highly reactive species capable of divided into two types: direct and quickly oxidizing substrate molecules. indirect. Direct photolysis occurs when the chemical itself absorbs light energy While few research studies have and undergoes reaction from its excited investigated photochemical treatment state. Organic compounds absorb 1 ight methods sl?ecifically applicable to oil energy primarily in the UV region. shale retort wastewaters, several Since little of the energy of UV wave­ researchers have observed successful length reaches the earth's surface, photodegradation of refractory organic direct photolysis is not a common compounds similar to those found in decomposit ion pathway for most organic retort water. Spikes and Straight pollutants. (1967) reported that many organic compounds including alcohols, nitrogen In the sensitized, or indirect, heterocycles, organic acids, olefins, photolysis process, another chemical benzenoids, phenols ,and aromatic species, called a sensitizing molecule, compounds, are susceptible to absorbs the 1 ight energy emitted in photodecomposition. Watts (1983) along the visible region where there is an with Sargent and Sanks (1974, 1976) have abundant supply reaching the earth's shown dye-sensit ized photooxidat ion to surface. The sensitizing molecule be an effective treatment method for becomes electronically excited to degrading biorefractory pesticide sand a new unstab Ie energy level. Excited aromatic compounds, respectively.

2 OBJECTIVES

The objective of this research­ Table 1, based on the criteria for was to invest igate the use of dye­ selection. Pertinent physical charac­ sensitized photooxidation as a potential teristics of the compounds are shown in treatment method for removal of organic Table 2. compounds from retort wastewaters rend er i ng these waters sui tab Ie fo r Specific objectives included: subsequent biological treatment, reuse and/or discharge. 1) Analysis of dye-sensitized photooxidation as a treatment method A literature search provided to completely degrade, decrease the information regarding the most abundant, toxicity of, or increase the biodegrad­ recalcitrant and/or toxic compounds on abili ty of selected hazardous compounds which to concentrate research efforts. identified in oil shale production Four chemicals were chosen for invest i­ wastewaters. The effect of pH, dye gation based on the following criteria: concentration, and initial substrate concentration on photodecomposition 1. existence in retort wastewater, reaction rates was used as criteria in determining optimum photooxidation 2. recalcitrance to conventional conditions. biological treatment, 2) Evaluation of the tOX1C1ty of 3. proven toxic, mutagenic or parent compounds and the photodegrada­ carcinogenic activity, and/or tion produc ts to determine whether dye-sensitized photooxidation treatment 4. included on the USEPA Consent reduce~ toxicity prior to biological Decree, Priority Pollutant List treatment or discharge. (U . S • EPA I 976 ). 3) Development of a chemical Two polynuc lear aromatic hyd rocarbons, actinometer, closely simulating actual (PNAs). anthracene and naphthalene, and experimental conditions, suitable for their nitrogen heterocyclic analogs, use in monitoring light intensity quinol ine and acridine. were selected. during sensitized photodecomposition These four compounds are classified in exper iment s •

3 Table 1. Characteristics of selected polycyclic aromatic hydrocarbons meeting criteria for this study.

Compound Anthracene Naphthalene Quinoline Acridine Criteria

Retort water x x x x

Pol ynuc 1 ear aromatic hydrocarbon x x

Nitrogen heterocycle x x

EPA pr iority pollutanta x x

Suspected carc inogen/ mutagena ,b X X X X

Envi ronmental levels: Toxicity based permissible concentrat ion Based on heal th effects (mg/l) C 2.0 0.69 0.14 0.8

Based on ecological effects (mg/l) C 0.05 0.5 0.25

a-U.S. EPA (1976) b-Pelroy and Petersen (1979) C-Kingsbury et al. (1979)

4 Table 2. Physical characteristics of selected polycyclic aromatic hydrocarbons.

Physical Anthracene Naphthalene Quinoline Acridine Charac teri st ics

Struc ture :::::-.. ~ ~ ~ ~ (() ~ "'.4- ceo CO ~ N"" 0::0N

Fonnula C14HlO ClOH8 C9H7N C13H9N

Molecul ar weight 178 128 129 179

Melt ing point, °c 216 88 -15 111

Boiling point, °c 340 218 238 345

Aqueous solubility, mg/l .07 30

5 LITERATURE REVIEW

Shale Oil Six h u nd red b i lli 0 n bar r e 1 s 0 f 0 it equivalent are contained within high Organic rich sedimentary rock, grade deposits, capable of producing 25 referred to as oil shale, was originally to 100 gallons of shale oil per ton of deposited over 50 million years ago as shale. The U. S. Department of the sediment in inland lakes. As a result Interior has estimated that 80 billion of a complex geologic history, these barrels of this reserve are recoverable sediments have been converted into by present mining technology. This marl stones , rich in solid hydrocarbons. total is approximately twice the present The organic fraction of oil shale domestic crude oil reserve in the United was not subjected to the same geologic States exclusive of Alaska, and repre­ conditions that yielded petroleum, tar sents 75 percent of the oil the United sands, and coal (Yen 1978). In shale States has produced since the Civil War oil, chemical bonds between individual (Pfeffer 1974). organic molecules transformed the original organic matter into solid Three basic methods of processing hydrocarbons known as kerogen, a high shale oil have been tested at pilot mol ec ul ar we ight po lyme r cons is t ing scale facilities: surface retorting, primarily of heterocyclic molecules with in-situ, and modified in-situ retorting. smaller quantities of aliphatic and Sur face retort ing processes (Figure 1) aromatic compounds (Fox 1980). During involve mining the oil shale with oil shale retorting, kerogen is thermal­ conventional techniques, crushing and ly broken down to release shale oil. retorting the material in surface operations, and disposing of the spent The process of extracting oil from shale residue. Spent shale solids pose s hal e is not new. Th e ear 1 yIn d ian s a major disposal problem. Redente et re ferred to the "rock that burns," and a1. (1981) estimate that although the private retorting of eastern U.S. shale mass of the shale oil is reduced 12 to took place in numerous small plants in 15 percent during retorting, its volume the mid-1800s (Maase 1980). By the will increase by as much as 30 percent. early 1900s nearly 200 claims had been staked in the western United States for The second method of extracting oil oil shale mining (Nowacki 1981). involves retorting the oil shale in place (in-situ) and is not as technical­ Oil shale occurs throughout the ly advanced as sur face re to rt oper­ world on every continent and in at ations. In-situ retorting offers least 30 of the United St ates. The a number of advantages over surface world t s richest reserves occur in the retort processes. It eliminates the 25,000 square mile Eocene Age Green necessity of disposing of large quan­ River Formation of Colorado, eastern tities of spent shale; and it may be Utah and southern Wyoming (U.S. DOl used where the shale is deeply buried, 1973). Deposits of known resources are where it occurs in relatively thin estimated to contain an equivalent of layers of rich and lean shale, or where 2.0 trillion barrels of crude oil in the conventional m1n1ng methods cannot be Green River Formation alone (Yen 1978). applied (Yen 1978).

7 Air r ~- -- -1- ---i--

Spent shole

8urning Product cooling zone Retorlin9 RetOrlin9 zone

Gos-combustion zone

Heal-exchonge zone Oil outlet

Disk feeder Piston

Retorted+ shale Oil GAS-COMBUSTION RETORT UNION OIL RETORT Recycle gas is mixed with air and burned within the Shale is introduced near bottom of retort retort_ Gases flow upward and shale moves down­ and forced upward. Air enters at the top ward. and flows downward.

Flue gas to atmosphere Row shale y I scr{bber I

TOSCO RETORT Ceramic bolls transfer heat to shale. No combustion takes place in retort.

Figure 1. Schematic representation of three surface oil shale retorting processes (U • S. 001 1973).

8 One application of the in-situ Gas condensate is the waste stream approach, known as true in-s itu retort­ generated from air pollution devices ing, is shown in Figure 2. This process used to treat process off-gas streams. consists of drilling a predetermined Gas condensate is characterized by high pattern of wells in a rectangular levels of ammonia, C02, alkalinity shale sect ion. If naturally occurring and dissolved organics (Hicks et al. permeability is low, permeability 1980), wh ile inorganic salt content is is increased by fracturing the shale negligible. with explosives (Nowacki 1981) or by excessive hydraulic injections Retort water is formed when water (Pfeffer 1974). The overburden is left vapor condenses in cool, rubblized intact. The fractured shale section is shale ahead of the fl arne front during ignited at one end, air is injected pyrolysis, and is collected with the through ignition wells to support recovered oil. Depending on the pro­ combustion, and shale is retort~ with a cess, up to 22 barrels of retort water horizontally moving combustion front, will be generated per barrel of oil converting the organic matter to oil produced. The volume may be higher if (Dinneen 1974). The oil, water and gas groundwater int rus ion occurs (Fox et al. produced is recovered from other wells 1980). Retort water leaches large in the section. In a true in-situ quantities of inorganic salts and process all the retorting is done dissolved organics from the shale oil. underground. The organic content may reach 4 percent while inorganic species at concen­ A modified in-situ process entails t rat ions of as much as 5 percent are mining some of the oil shale to create a typical (Fox et a1. 1980). Oreanic permeable combustion zone for in-situ concentrations can range up to 4,000 retorting with the mined shale retorted mg/l as TOC (Fruchter and Wilkerson above ground. 1980). Concentrations of chemical constituents in retort water are com­ Retort Wastewater pared in Table 3 for direct heating and indirect heating surface retorting, true The main wastewater streams gener­ in-situ, and modified in-situ with above ated from oil shale processing in­ ground retorting. The wide range of clude mine drainage, gas condensate, concentrations reflect the differences and retort water. No mining/retort ing in retorting processes and the degree of can begin until the water table is groundwater infi! tration. While mine lowered below the level of the mining/ drainage may potentially be the largest retorting operation. The quality and wastewater stream produced, retort water volume of mine drainage water is ex­ may be the most complex and difficult to tremely site specific and depends upon treat (Klieve et al. 1981). the type of mining, rate of mine expan­ sion, and the permeabil ity and porosity With the breakdown of the kerogen of the material in which the mine is in the oil shale, polynuclear aromatic located. Generally, mine drainage (PNA) compounds, poly~yclic aromatic water contains high levels of total arnines, various heterocycles and other dissolved solids, boron, and fluorides carcinogenic compounds are distributed (Nowacki 1980). The developers in between the shale retort water and the the Piceance Creek basin of the Green spent shale ( Jones eta 1. 1982). The River Formation, employing modified principal inorganic components of retort in-situ retorting techniques, antici­ water are ammonium, sodium and bicarbon­ pate 2.4 to 6.8 barrels of water per ate ions (Fox et ale 1980). These barrel of oil produced to be extracted materials can interfere with other as mine drainage water (Hicks et al. treatment processes, via toxicity to 1980). biological systems and scaling in gas

9 .. ..

(a)

AIR IN GAS OUT COMBUSTION FRONT OIL PRODUCTION WELL AND x x PUMP BLOWERS x x x VACUUM RETORT X ZONE X

x x X x

/ -x x X " , / "- *- --- -x- --- * --- -X- --- w OFF-GAS "X - BLAST HOLES TREATMENT I-==:::;:J FACILITY

(bl DE-MISTER SEPARTORS

Figure 2, Schematic representation of an in-situ oil shale retorting process, a) vertical crOss section; b) plan-view (Nowacki 1981),

10 Table 3. Water quality of oil shale retort wastewater (Nowacki 1981).

Above Ground Above Ground True Modified Indirect Di rec t In-situ In-situ Condu.c t iv i ty, mhos/em 120 ,000 16,000 49,600 82,250 pH, units 8.4 8.8 8.7 8.1 Oil & grease, mg/l 964 392 1,432 Phenol, mg/l 8.7 8.2 8.0 Chemical oxygen demand, mg/l 86,400 7,700 28,000 15,536 Total organic carbon, mg/1 19,290 15,000 12,300 Inorganic carbon, mg/l 1,600 Total dissolved solids, mg/l 160,000 1,856 13,700 1,002 Volatile dissolved solids, mg/l 159,700 1,810 Total suspended solids, mg/l 70 1 50 Chlorides, mg/l 70,000 2,300 5,000 53 Sulfate, mg/l 5,500 29 1,600 376 Fluorides, mg/l 5 3.1 31 5.1 Cyanide, mg/l > 0.1 11.8 0.19 Total nit.rogen, mg/l 33,600 5,100 30,300 Ammonia, mg/l 31,700 3,388 18,200 38,000 Nitrat.es, mg/l 580 138 33 Nitrites, mg/l 0.1 0.3 Total alkalinity as CaC03, mg/l 35,200 12,800 92,343 Arsenic, gil < 5 0.19 5.9 0.52 Mercury, gil < 0.1 < 0.1 0.1 < 0.002 Total sul fides, as S, mg/l ( 0.5 1,354 14 915 Trace elements, mg/1 Aluminum 2.4 0.36 0.11 Andmony <0.9 0.009 Barium <0.3 0.012 Boron 1.2 0.18 5.2 CadmilJlll (0.9 Calcium 18 2.5 0.6 Chromium 0.3 0.02 0.01l Copper 9 0.07 0.008 Iron 9 1.1 77 . Lead 3 0.12 Magnesium 30 2.5 3.2 Manganese 1.2 0.03 0.042 Nickel' 0.2 0.04 1.1 Potassium 3 0.36 8 Sodium 30 2.5 210 Tin (0.3 0.14 8.9 Zinc <1.2 0.26

II strippers, and consequently, require to man and other organisms from organic removal (Hicks et al. 1980). chemicals released into the env irornnent from both natural and man-made sources. The general plan developed by the Polynuclear aromatic compounds represent oil shale industry is for zero or near­ one of the largest groups of carcinogens zero discharge of wastewater from re­ causing concern (Lee et al. 1981). torting facilities. Because of the high levels of organics, inorganics, oi Is, PNA compounds consist of two or ammonia and alkalinity in retort waters, more fused benzene rings in linear, pretreatment for the removal of some or angular or cluster arrangements. all 0 f these const ituents is required Substitution of carbon in the benzene prior to full-scale treatment and/or ring with nitrogen, sulfur, oxygen or water reuse. Ai?i>roximately 50 percent other elements creates heterocyclic of the orGanic carbon in retort water is aromatic compounds. amenable to bioxidation; the remaining organic compounds are recalcitrant. Physical and chemical characteris­ These compounds are predominant ly non­ tics of PNA compounds vary with the polar solutes such as polycyclic aroma­ molecular weight of the compound. The tic hydrocarbons, and nitrogen and oxy­ lowest molecular weight member, naphtha­ gen heterocycles (Jones et a1. 1982). lene (ClOH8, MW 128), consists of two Due to the carcinogenic or mutagenic fused benzene rings. One of the largest characteristics of these refractory com­ hydrocarbons commonly studied is pounds they may pose a potential danger coronene (C24H12' MW 300) (Neff 1979). to human health and the environment. Within this range exists a large number of PNA compounds which differ in ring Retort water is generally used for number, structure and chemical activity. spent shale disposal and dust control and it is anticipated that oil-water The majority of the PNA's are separation and steam stripping of formed· during the incomplete combustion ammonia, along with the removal of trace of organic matter at high temperatures. organics, will be required prior to Several mechanisms have been proposed reuse in full-scale facilities (Nowacki for the formation of PNA's by pyrolysis 1980). and pyrosynthesis (Schmeltz and Hoffmann 1976, Crittenden and Long 1976). In Raphaelian and Harrison (1981) pyrolysis, complex organic molecules, characterized the organic fraction of exposed to elevated temperatures, are retort water from an in-situ process. partially broken, yielding lower molec­ Resul ts indicate that nitrogen hetero­ ular weight, unstable free radicals. cycles are among the most abundant or­ Pyrosynthesis of PNA's then proceeds ganic species present. Within the nitro­ by rapid combination of these free gen heterocyclic class, the quinolines radical fragments containing one or more in the basic fract ion of retort water carbons, to form larger reI at ively represent the largest group (Santodonato stable aromatic hydrocarbons. and Howard 1981), and many nitrogen heterocycles including quinolines have In general, all organic compounds exhibited carcinogenic activity (Coomes containing carbon and hydrogen may serve 1 9 79 ) . Ra 0 eta 1. (1 9 7 9, 1 9 81) h av e as precursors to PNA's. Where oxygen, shown in-situ product water and in-situ nitrogen or sulfur is present in the groundwater from a simul ated modified starting material, oxygen-, nitrogen-, in-situ retort to be mutagenic. or sulfur-heterocyclic analogs can be expected (Lee et al. 1981). Polynuclear Aromatic Hydrocarbons There has been growing concern The yield and molecular weight regarding the possible harmful effects dis t rib uti 0 n 0 f the P NA's form e d

12 during combustion depend on the pyro­ absorption and phototoxicity. Among the lysis temperature and duration of heterocyclic compounds tested, acridine the exposure (Lee et al. 1976, Lee and absorbed UV-A light more effectively Hites 1976) and the chemical composition than the other heterocyclic compounds of the starting organic matter. and was the only heterocycle found to be phototoxic. During pyrolysis of oil shale, in which kerogen acts as the starting Although the presence of nitrogen material, a large number of PNA com­ heterocycles in the environment has been pounds, po lye yc 1 ic aroma t ic pr imary known for many years, the biological amines, various heterocycles and other effects of these compounds have not been carcinogenic compounds are formed and as intensively investigated as have the distributed between the shale oil, the polynuclear aromatic hydrocarbons. retort water and the spent shale. The Several decades ago researchers demon­ PNA's and their heterocycles are of strated that several naturally occurring critical environmental and health nitrogen heterocycles were carcinogenic concerns because of: 1) chronic heal th to mice (Santodonato and Howard 1981), effects (carcinogenicity), 2) microbial leading to the present concern over recalcitrance, 3) high bioaccumulation their potential human health signi fi­ potential, and 4) low removal effi­ cance. In recent studies in bacteria, ciencies in traditional wastewater acridine has been demonstrated to cause treatment processes (Fox et al. 1980). photodynamic damage to both DNA and cell membranes (Wagner et al. 1980, Snipes et PNA compounds have received addi­ a1. 1979). tional attention recently due to the existence of many PNAs in the EPA The basic fraction of many environ­ Priority Pollutant Consent Decree mental samples usually contains a number (U.S. EPA 1976). Tabek et a1. (1981) of nitrogen heterocycles. Pelroy and examined the biodegradability of Petersen (1979) demonstrated that the PNA compounds along with other organic basic fraction of shale oil was more compounds listed as priority pollutants. mutagenic than any of the other frac­ The PNAs demonstrated varied rates of tions (acidic, neutral, tar, PNA, crude biodegradation with different acclima­ oil) tested. Quinoline and acridine are tion periods depending on the test considered bases possessing these compound and the dose of substrate. mutagenic characteristics. Concentra­ Tricyclic aromatic hydrocarbons were tions of 0.5 mg/l to 1.0 mg/l of found more susceptible to biodegradation quinoline in water may cause tainting of than the tetracyclic and higher poly-· fish flesh (Kingsbury et a1. 1979>­ cyclic hydrocarbons. Acridine has been shown to be toxic to fish at 5.0 mg/l levels (Kingsbury et Kochevar et al. (1982) have demon­ al. 1979). Permissible environmental strated that some PNA compounds commonly concentrations for the test compounds, found in coal tar exhibit phototoxic based on ecological and health effects, potent ial . The degree of erythema are contained in Table 1. produced in guinea pig skin after exposure to the compound and UV-A Treatment Options radiation (320-400 nm) was taken as the measure of phototoxic i ty. Pyr ene, Conventional processes anthracene and fluoranthene were strong­ ly phototoxic. Acridine was phototoxic, The oil shale industry is still but to a lesser degree. These compounds struggling to fully develop and at absorbed available 1 ight more e ffec tive­ present no active oil shale mines are ly than other PNA compounds and showed a operating on a commercial scale. If and correlation between relative light when commercialization begins, a

13 variety of retorting processes may be atmosphere is well documented (Al t­ used, each creating unique environmental schuller and Bufalini 1971, Leighton problems. Since full scale designs have L961), but comparatively little is known not been completed, environmental about photodegradat ion of pollutants in concerns can be considered and incor­ water where other competing processes porated into process design from the such as biodegradation and hydrolysis early stages of development. occur (Zepp and Cline 1977). In a water system, biodegradation is often the The need for treatment opt ions for major pathway for elimination of the removal of the refractory, and often organic compounds. When the system times hazardous, organic compounds in includes refractory species however, oil shale retort water has been identi­ biodegradation is too slow to be effec­ fied, and various physical and chemical tive and photodegradation may become treatment options have been investi­ the most important pathway for these gated. Many of the options investi­ compounds. gated, such as wet air oxidation, ion exchange, reverse osmosis, or excess Light, a form of electromagnetic permanganate treatment, are highly radiation, has a dual nature, exhibiting capital and/or operationally cost properties of both waves and particles. intensive. Others, such as solvent The quantum theory introduces a rela­ extraction, coagulation and steam tionship between the wave and quanta stripping, do not provide adequate nature of light (Smith 1971). Light is removal of the residual organic fraction a form of energy and the individual par­ in the retort water. ticles, or quanta, possess energy that depends on the wavelength of light, with Biological treatment is widely used more energy supplied by light quanta as as a cost effect ive means of removing the wavelength of light decreases biodegradable organic materials from (Plimmer 1971). In keeping with the municipal and industrial wastewaters. theory of duality of light, photo­ Malaney et al. (1967), Hicks et al. chemists use a particular wavelength in (1980), and Harrison et al. (1975) have calculating relevant extinction coeffi­ shown that many refractory and/or car­ cients and count quanta in determining cinogenic compounds found in industrial the efficiency of a photochemical waste streams cannot be effectively reaction (McLaren and Shugar 1964). removed by means of conventional bio­ logical treatment. The first 1 aw of photochemistry, known as the Grotthus-Draper law, states Successful treatment of oil shale that only 1 ight which is absorbed by a wastewater will require the application system can induce a chemical reaction. of numerous physical/chemical and As a molecule absorbs light, it receives biological unit processes. A particular energy in the form of discrete units process may be extremely effective for called photons or quanta. The second the removal of specific classes of law of photochemistry, the Stark­ compounds, however, when appl ied to Einstein law, involving the particle a mixture such as retort water, it may nature of light, states that one quantum be able to remove only a fraction of light is absorbed per molecule of of the pollutants. Complete treatment absorbing and reacting substance that will require a combinat ion of methods disappears. Since changes in single used in series or parallel applications. molecules of reactants are not ordi­ narily observed, a more tangible quan­ Dye-sensitized photooxidation tity, the einstein, is used in dis­ cussing photochemical react ions j an Fundamentals. The importance of einstein is one mole of quanta (McLaren photochemical transformation in the and Shugar 1964).

J4 Absorption of light can cause gap between ground state and the first oxidation reactions which would other­ excited state of that molecule must be wise not occur. Chemical reactions are less than the absorbed energy of the not restricted to the molecule absorbing triplet sensitizer, which is on the energy. as excitation energy may be order of 40 kilocalories/mole {kcal/ transferred in a number of ways to the mole} for common sensitizing organic species that ultimately undergoes dyes. Most organic molecules require chemical change. An almost unlimited more energy to be excited than is number of photooxidation reactions are ava i 1 a b 1 e from a t rip 1 e t sen sit i z e r . possible and these can be broken down Oxygen however. is unusual because its into two types: direct and indirect. ground state is a triplet and the energy Direct photolysis occurs when the gap between ground state and the first chemical molecule itself absorbs 1 ight excited state. which is a singlet. 1S and undergoes reaction from its excited unusually low (22 kcal/mole>. Thus if state. Indirec t photolysis occurs when dissolved oxygen is pre~ent in even another chemical species. called a trace amounts. a triplet sensitizer will sensitizer molecule. absorbs light selectively transfer its energy to energy and then transfers this energy oxygen (Foote 1968. Sargent and Sanks from its excited state to another 1976) . chemical which undergoes reaction (Mabey et a1. 1982). Singlet oxygen is a strong oxi­ dizing agent with an average half-life Absorption of light energy, pri­ of 2 U seconds (Kearns 1971). capable of marily in the ul traviolet (UV) region. quickly oxidizing photosensitive sub­ is a common characteristic of most strate molecules. Foote (1968) and organic pollutants. Transformations of Pitts et al. (1963) suggest that the two organic pollutants by direct photolysis singlet oxygen species most likely depend then upon absorption of energy in involved in photooxidation reactions are the UV spectrum. This direct photolysis singlet deltaoxy el\. (It. g). and singlet reaction is generally an inefficient sigma oxygen. ( 1E ). Kearns et al. environmental process since light (1967a) have suggesfed that high energy transmittance in the atmosphere in the sensitizers with triplet state energy of UV and visible region decreases with the formation. Et • ire~ter than 37 kcal. decreasing wavelength. Essentially no th~ energy of E • should produce Ii g h tis t r a nsm itt edt 0 the ear t h ' s IE oxygen as th~ primary product surface at wavelengths less than 295 of genergy transfer. whereas sensitizers :anometers (Zepp and Cline 1977). with Et less than 37 kcal should produce only It.g oxygen which has an energy of In the sensitized photolysis formation of 22 kcal. Figure 3 depicts process. however. sensitizing molecules the electronic states and configurations absorb light in the visible region where of molecular oxygen. there is a wealth of energy reaching the earth's surface. The electronically The 1t.g oxygen state is long lived excited sensitizing molecule or triplet and survives at least 108 collisions sensit izer is usually unstab Ie at this with methano~ in the vapor phase, new level of energy and by a deactiva­ whereas the IE state survives no more tion process, it reverts to a lower, than ten colPisions under the same less energetic. stable state. Triplet conditions (Foote 1968). sensit izers return to ground state by transferring energy to molecular oxygen Porter and Wi lkinson (1961) found or organic substrates which are non­ that. as the triplet state energy of the ab'sorbing in this same visible wave­ sensitizer approaches the triplet state length region. For energy transfer to energy of the acceptor, the energy another molecule to occur. the energy transfer probability 1S significantly

15 III

St ate Occupancy of highest orbitals Energy Lifetime in solution

+ 1 1: g t + 37 kc al 10-9 to 10-12 sec 1 1I g , 22 kcal 10-3 to 10-6 sec

31:; -4- ground state 0'1

Figure 3. Molecular oxygen states and their electronic configuration (Foote 1968). reduced. When oxygen is t he acceptor, oxygen since both species may be formed Kearns et al. (l967b) predicted the in varying amounts and are capable of ratio of lEg to 16g oxygen to decrease reacting with organic matter (Acher and to zero as Et of the sensitizer ap­ Rosenthal 1977): proaches 37 kcal. This study concluded that the two sing let oxyg en spec ies S + hv -+ S* (1) exhibit different reactivities, and that observed variations in reaction product (4) distribution as a function of sensitizer used results from a variation in the 16g 02 +OM -+ oxidation products (5) relative amounts of IE g and 16g oxygen generated. Which of the two reaction types predominates and fhe efficiency of The transfer of energy from the each path depends on the oxygen and excited triplet sensitizer to the substrate concentrations, the rates reactants in the system may proceed by of reaction of sensitizing triplet with two routes, one in which the interaction substrate and with oxygen, the rate of is with the substrate and one in which triplet decay, reaction pH, and other the primary interaction of the sensi­ physicochemical factors (Sargent and tizer triplet is with oxygen (Foote Sanks 1974). Foote (1968) found that 1968). The former involves energy even traces of oxygen will likely transfer from the excited triplet inhibit Type I reactions. sensitizer (S*) to the organic matter (OM) in the system to generate reactive, Engineering applications. Much short lived radical intermediates which scientific research has been conducted dissociate into oxidation products: on dye-sensitized photooxidation, however engineering application of the S + hv -+ S* (1) photolysis process to water and waste­ water treatment has received little S* + OM -+ transient species + S (2) attention. Spikes and Straight (1967) reported that many organic compounds transient species -+ including alcohols, nitrogen hetero­ oxidation products (3) cycles, organic acids, olefins, benze­ noids, phenols and aromatic compounds This reaction mechanism is termed Type I undergo sensitized photooxidation in and may include hydrogen abstraction, natural systems. According to Foote electron abstraction, alkyl radical (1968) many organic compounds are generation and hyd roxide radical gener­ capable of absorbing light to initiate ation as intermediate mechanisms (Heelis these photochemical reactions. Some et al. 1981), A requirement for energy will sensitize only very specific transfer is that the acceptor have a reactions, while others will sensitize triplet energy lower than that of the the photooxidation of many substrates. senslt1Zer; triplet energies of sensi­ Investigations by Sargent and Sanks tizing dyes are very low (on the order (1974 and 1976) indicate that common of 40 kcal) .compared to most organic dyes such as methylene blue, toluidine compounds so this condition is seldom blue, rose bengal and neutral red are satisfied. effective non-specific sensitizers.

The second process, termed Type II Gurnani et al. (1966) observed the involves energy transfer to molecular sensitized photodegradation of the amino oxygen and results in excitation of acid, tryptophan, in an effort to molecular oxygen to singlet oxygen, understand the mechanism of its oxi­ where l6g 02 in Equations 4 and 5 dation. Seven chromatographically represent either species of singlet separable produc ts were formed when an

17 aqueous solution of tryptophan was Engineering design parameters for exposed to light at pH 9 in the presence the treatment of refractory organics of methylene blue and oxygen. In the were determined by Sargent and Sanks absence of methylene blue and/or oxygen, (974) using cresol and methylene blue Some of these same products were de­ as substrate and sensitizer, respec­ tected, though in much smaller quant i­ tively. Design parameters examined ties. Exposure of tryptophan to UV included optimum pH, substrate concen­ light did not result in the formation of tration, dye concentration, effect of any of the products obtained with dye bleaching, and kinetic rate con­ visible light. stants for both anaerobic and aerobic react ions. They found that the photo­ The products formed during sensi­ lysis of cresol occurs as much as 2.8 tized photolysis of cholest-4-en-3B-ol times faster in aerobic systems than have been studied in considerable in anaerobic systems, again suggest ing detail to gain insight into the singlet the importance of singlet oxygen to oxygen mechanism. Only two products effective photolytic reactions. A first are formed during this photooxida­ order rate equation described, with tion; epoxy ketone and enone (Kearns reasonab Ie accuracy, the rate at which et a1. 1967a and 1967b), however, the cresol was photooxidized. The reaction product distribution ratio varied mechanism is probably not first order, depending on the nature of the sen­ however, but a complex function of sitizer. This result is linked to the substrate concentration, dye concent ra­ relative amounts of IE ~ and It,g oxygen tion, and oxygen concentration, as well generated. as temperature, pH, and light intensity.

One of the earlier engineering The pH dependence of singlet oxygen applications of photochemical pro­ produc t ion in aqueous solut ions was cesses was performed by Bulla and studied by Bonneau et al. (1975). Edgerley (1968) who examined design Thiazine dyes were utilized as photo­ and operat ing parameters in the UV sensitizers to investigate the yield of photolys is of three chlorinated hydro­ singlet oxygen as a function of pH. carbons: aldrin, dieldrin, and end rin. Both laser flash photolysis and con­ Kinetic rate equations were used to tinuous irradiation were used. The describe the effect of design para­ production of singlet oxygen by the meters. A first order rate law was thiazine dye, methylene blue, was very applicable to retention time, lignt sensitive to changes in pH in the range intensity, and depth of reactive of 5 to 9. Based on observed photo­ solution. Temperature had no significant oxidation rates for methylene blue in effect on the rates measured. Bioassay aerated solutions, the production of tests indicated that aldrin degradation singlet oxygen was approximately 5 times products were less toxic than aldrin more efficient in basic than in acidic itself. medium.

First order reaction rate equatio~s Cresol and phenol were tested for were applied to studies by Young et al. sensitivity to dye-sensitized photo­ (1971) when examining the effects of oxidation by Sargent and Sanks (1976). solvent polarity in dye-sensitized Dyes used as sensitizers included photooxidation. Three solvent systems rhodamine 6G, sodium phthalocyanine, stud ied were methanol, n-butyl alcohol, 2,7 -dichlorofluorescein, neutral red, and tert-butyl alcohol-:- The react ion methyl ene blue, rose bengal, malachite rate;-;aried due to' the change in the green, hematoporphyrin-D and acridine lifetime of singlet oxygen 1n the orange. Tungsten lamps were used as the different systems, related to the light source. Results indicated that viscosity of the solvent. singlet oxygen was the mechanism of

18 oxidation and that methylene blue and solution at pH 9. Rose bengal, methyl­ rose bengal were the most effect ive ene blue and eosin dyes were better sensitizers studied. The authors singlet oxygen producers by a factor of concluded that the reaction of singlet at least 9 than thiopyronine, proflavine oxygen with refractory aromatic hetero­ and acridine yellow. Singlet oxygen cyclic compounds gives products trac­ produc t ion was not observed with 9- table to biological oxidation, but the aminocaridine, acridine orange, quin­ reaction products were not tested for acrine and ethidium bromide. biodegradability. Wat ts (1983) developed rat ional Acher and Rosenthal (1977) in­ engineering criteria for the design of vestigating the appl icabil i ty of dye­ dye-sensitized photooxidation lagoons sensitized photolysis for the treatment for selected pesticides. Engineering of domestic sewage found that the application parameters investigated chemical oxygen demand (COD) could inc lud ed opt imum pH, maximum lagoon be reduced up to 67 percent and total depth, mixing effects, aeration re­ disinfection was achieved upon photo­ quirements and optimum methylene blue lysis with methylene blue. Experiments concent rat ions. Rate constant s were were conducted in natural and artificial also determined for varying treatment light, and methylene blue was determined conditions including natural sun­ to be the optimum sensitizer at a light photolysis and artificial light concentration of 12 mg/l. reactions.

Natural water was used ~n experi­ Actinometry ments by Zepp et a1. (1977) to examine the steady state concentration of Estimation of the rate constants singlet oxygen. The highest rates for photolysis of a specific chemical in occurred in highly colored water, rich naturally occurring waters, based upon in fulvic acid. Since most organic rich laboratory data, requires information natural waters contain large concentra­ regarding the absorption spectrum of the tions of fulvic acid, the authors chemical, the quantum yield, or effi­ concluded that photochemical generation ciency, of the process and the intensity of singlet oxygen is a widespread of the light as a function of wavelength phenomenon in the aquatic environment. (Dulin and Mill 1982). Chemical actinometry is the usual method for Rates of direct photolysis of two determining the quantum yield of pesticides, carbaryl and trifluralin in photoreactions as well as the light pure naturally occurring river water intensity of the given photosystem. were investigated and modeled by Zepp and Cline (1977) under natural sunlight A chemical actinometer system conditions. Hal f-lives were calcul ated consists of a photochemical reaction as a function of season, latitude, time possessing a known and accurately of day, depth of water bodies and ozone reproducible quantum yield. Ideally, layer thickness. Experimental verifi­ the quantum yield of the system should cation of the computed half-lives was be independent of light intensity presented. over the wavelength range of interest. By establishing reaction quantum yield Using the appearance of stable and light intensity, the kinetic data nitroxide radicals from the photo­ derived from photolysis under one lysis of amine 2,2,6,6-tetramethyl-4- set of light conditions can be applied piperidone, as an indication of reaction to a wide range of I igh t condi t ions rates, Houba-Herinet al. (1982) studied utilizing the photOChemical reaction the production of singlet oxygen with equation indicated below (Dulin and Mill irradiation of several dyes in aqueous 1982) :

19 actinometer for photochemical research (6) (Calvert and Pitts 1966). It is ex­ tremely sensitive over a wide range of wavelengths and is simple to use. When sulfuric acid solutions of K3Fe(C204)3 where: are irradiated in the range of 250 to 577 nm, simultaneous reduction of ::: reaction quantum yield iron to the ferrous state and oxi­ for the chemical, molest dation of oxalate occur. The quantum einstein yields of Fe 2+ format ion have been accurately determined, and are inde­ initial rate constant pendent of wavelength from 250 to 577 for photoreaction of the nm. The reaction is independent of chemical, mole/sec reactant and product concentrations, incident light intensity and reaction ::: initial rate constant for temperature. photoreaction of the actinometer, mole/sec Although not as commonly used as the ferrioxalate system, other chemical a ::: n,\£>.. summation of the product actinometer systems appear to be useful of the light intensity and for environmentai photochemical studies. extinction coefficient of Dulin and Mill (1982) developed two each wavel engt h of the act inometers for iutens ity measurements actinometer, day-l from 300 to 370 am, suitable for natural sunlight eXl,)eriments. Both systems, p­ n,\£,\ c ::: summation of the product nit r 0 ani sol e ,- p y rid i n e and p - nit r 0 - of 1 ight intensity and acetoph.:!llone-pyridine, have quantum extinction coefficient of yields invariant of wavelength. Neither each wavelenjth of the are volatile from water nor readily chemical day - sorbed, and each is easily analyzed by high pet'formance liquid chromatography. tPa ::: quantum yield for the actinometer, mole/einstein An actinometer for the visible Several criteria for chemical and wavelength range, 400 to 580 nm, was spectral properties that should be given presented by Brauer et al. (1982). The high priority when selecting actino­ photooxidation of heterocoerdianthrone meters include: 1) a quantum yield that in toluene to an endoperoxide with is temperature and wavelength inde­ visible light can be evaluated using UV pendent, 2) a quantum yield that is and visible spectroscopy. independent of the concentration of the actinometer, 3) photolysis products that do not interfere with photolysis, An actinometer system should, either chemically or spectrally, 4) the however, closely simulate the experi-' ability to be readily analyzed by mental conditions in the photochemical conventional methods,S) stability 1n research. In this study, the usable water, 6) stability in the dark, and 7) wavelength region was outside the range moderate solubility and non-volatility for the effective utilization of the 1n water (Dulin and Mill 1982). ferrioxalate system. An alternate system was required which would monitor An actinometry procedure described light intensity for the wavelength by Hatchard and Parker (1956) utilizing region 570 to 700 nm where the sensi­ potassium ferrioxalate is regarded as tizer, methylene blue, absorbs most the classic liquid phase chemical strongly.

20 Bellin and Yankus (1968) found that photons absorbed is called the primary the amino acid, L-tryptophan, was quantum yield for the process. A susceptible to sensitized photodegrada­ corollary of the Stark-Einstein law is tion by methylene blue. To maintain a that the sum of the primary quantum reaction quantum yield of 1.0, Bonneau yields of all the processes that de­ et al. (1975) discovered that an aqueous activates the excited molecule equals solution of tryptophan and methylene unity (Mousseron-Canet and Mani 1972). blue must be buffered to a pH greater This does not apply to the overall than 8. The utilization of a methylene reaction quantum yield, (jlrxn,which blue-tryptophan actinometer system would may be much larger than unity ard for closely parallel the actual photolysis certain chain reactions it may be as conditions, uti! izing the same reaction high as 104 to 106 (Turro 1978). It geometry and light emitting apparatus is essential, therefore, to identify the and consequently was adopted for use in type of quantum yield in discussion. this study. The quantum yield, (jlrxn, is defined by Calvert and Pitts (1966) a8:

The quantum yield. ~, is a means of (jl Moles of molecules reacting quantifying the efficiency of a photo­ rxn = Moles of photons absorbed chemical reaction and is one of the most useful and fundamental quantities in the (7) study of photochemical mechanisms. Its magnitude provides important information By selecting a chemical actinometer regarding the nature of photooxidat ion with a photoreaction which is sensitized reaction. Small quantum yields (~«l) by the same dye used in the photolysis indicate that processes such as deacti­ experiment, Equation 6 simplifies to: vation, quenching, fluorescence or other processes that lead to no net chemical change are competing with the photo­ (8) oxidation reaction (Calvert and Pitts 1966). Large quantum yields (~»l) indicate that mechanisms operating in a and was used for the calculation of the chain reaction are occurring. overall reaction quantum yields in this study. The second law of photochemistry, the Stark-Einstein law, states that one Spikes and Straight (1967) reported molecul e is act ivated for each photon that quantum yield is not a function of absorbed in a system. After a molecule light intensity so the quantum yield of absorbs a photon, it is unstable and reaction calculated from laboratory will undergo a variety of competing photolysis data using Equation 8 may primary processes, such as chemical then be used to predict the rate of reaction, light emission, energy trans­ disappearance of the pollutant, kC, in fer, or phys ical deactivation, to return the environment under natural sunlight to a stable state. The fraction of conditions.

21 MATERIALS AND METHODS

Photolysis Bench The pho tobench cons i ;ted 0 flO pairs of 8 foot lamps placed across a 3 Photooxidation experiments were foot wide bench. The lamps were matched performed using artificial constant so that intensity was constant across light provided by Duro Test Vi tal ite'" the length and width of the bench. The lamps, which are 91 percent corrected to bench was of sufficient size to allow the quality of natural sunlight. the examination of photolysis reactions Radiation impinging upon test solutions in 25 reactors at one time. was routinely measured using a chemical actinometer and a Lambda-185 Radiometer. Barltrop and Coyle (1978) suggested A radiometer integrates the light output that at least 1015 photons/second OVer the entire visible spectrum and an light energy be available for effect ive overall intensity output is recorded. photolytic reactions to occur. The The Lambda-185 Radiometer recorded an chemical actinometer system cot.sisting average intensity of 48 W/m2 throughout of a tryptophanmethylene blue mixture the experiment. Of greater importance was run repeatedly at varying distances to sensitized photolysis experiments is from the lamps. By successive appl ica­ the incident light intensity in the tion of Equation 9, wavelength region of maximum absorbance of the sensitizing dye. The chemical actinometer monitored the energy output lICVN light energy (9) for the maximum absorbing wavelengths, = a t 570 nm to 700 nm for methylene blue.

Radiation as a funct ion of wave­ where length emitted by the Duro Test Vita­ lite~ lamps was measured for the visible C = change in tryptophan concen- light spectrum by an Optronic 742 tration (mole/ liter) spectroradiometer (Figure 4). The spec trorad iome ter was a1 so used to V = volume of sample (liters) measure radiation as a function of wavelength of natural sunlight in July a = quantum yield of actinometer during the solar noon period in Logan, (mole/einstein) Utah (latitude, 42 0 N). A comparison of the percent of total light intensity t = reaction time (seconds) emitted at each wavelength in natural light and under the experimental lights N = 6.023 x 10 23 (photons/ is shown in Figure 5. The sensitizing einstein) dye, methylene blue, exhibits its maximum absorbance at 665 nm. The The light energy output in photons/ percent of the total light intensity second was determined for each trial emitted at 665 nm was the same for both distance. The base of the photobench the artificial light source and natural was finally adjusted so that the test sunlight making experimental and ex­ solutions were 14 cm from the lamps pected natural sunlight induced reaction thereby receiving the suggested light results comparable. energy input.

23 --- 200

_ ....

E c: 150 N "E CJ)" I- I- 100 ~ E 50

400 450 500 550 600 650 700 750 800 WAVELENGTH (nm)

Figure 4. Radiation as a function of wavelength emitted from the DuroTest Vitalite lamps.

2.50 ...... Nolurol lighl -% ot 10101 Intensity occurring 0 I eoch A..

--- Artificiol lighl- % of 10101 inlenslly occurring 01 eoch A.. 2.0

• ~ ••!

~' •••* •• "'.''''. >­ I- 1.50 .. / .... \:...... CJ) Z W I­ Z 1.00

". 0.50

'" . :" ...... ~.:

O.O~----r------~---~----~------r----'------'--~---'- 400 450 500 550 600 650 700 750 800 WAVELENGTH. nanometers

Figure 5. Percent of total ligh t i ntens ity occurring a t each wave lengt h for natural sunlight and artificial light.

24 Reactors Those samples buIier€(l to pit 3, 5, and 7 were prepared using a citric acid-ohos­ Glass containers approximately 13 phar:e solution, while samples anaiyzed cm in diameter were used as photolysis at pH 9 and II were buffered with a car­ reaction vessels. The average depth of bonate-bicarbonate solution. Buffer so­ the test solutions was approximately 2 lut ion ingredients for each exper imen­ cm. The reactors were prepared by tal pH condition are shown in Table 4. washing with a 10 percent sulfuric acid solut ion .followed by three rinses wi th Methylene blue was selected as the dist illed deionized water. A Milli-Q'" sensitizing dye for the experiment. The deionizing water system provided labor­ effect of dye concentration on the rate atory pure water for all washings and of photooxidation was examined using dilutions. Glass plates (7" x 7" x dilute dye solutions at concentrations 1/4") were cut to fit over the reactors of 10 mg/l or less. Blanks accounting to minimize evaporation. for adsorption loss on containers and compound loss due to volatilization were Photolysis analyzed for each compound by examining degradation rates in solutions contain­ Two polynuclear aromatic hydro­ ing no dye (direct photolysis) as well carbons, anthracene and naphthalene, as in solutions with methylene blue and their nitrogen heterocyclic analogs, maintained in the dark. acridine and quinoline, were selected as representative pollutants in retorting For each compound analyzed, an wastewater; and were used as photo­ initial run was performed using a test oxidation substrates. Optimum condi­ solution containing 5 mg/l methylene tions for photodegradation were deter­ blue. Table 5 represents an example of mined by varying the pH and sensitizing experimental set-up. Twenty-five 200 ml dye concentration. samples, comprising an experimental run, were divided into groups of five, with Due to the low aqueous solubility each group representing one sampling of the four compounds used in this time interval. Within each group of st udy, t he use of a co-solvent was five, the solutions were buffered to pH necessary. Acetonitrile was used as a 1 3, 5, 7, 9, and 11. Samples were taken percent co-solvent due to its inert at five time intervals to allow deter­ photochemical characteristics (Mabey e t mination of the rate law and rate al. 1982). Reaction solutions were constants governing the reactions. prepared from a stock solut ion of the The time interval s were changed as chemical in HPLC Grade Acetonitril e at necessary to suit the reactivity of the a concentration 100 times that used substrate. during photolysis. An aliquot of this solution was diluted with deionized Following the complet ion of tripli­ water to yield a 1 percent acetonitrile cate analyses of the 5 mg/l sensitizer in water solution, with final compound runs, a lower concentration of methylene concentrations of 1.0 mg/l for anthra­ blue, 2 mg/l, was investigated. The same cene and naphthalene and 5.0 mg/l for experimental design was followed at this acridine and quinoline. lower concentration. The third experi­ mental methylene blue concentration was Optimum reaction conditions selected based upon the results of these first two runs. If thf' 2 mg/l methylene The effect of pH on sensitized blue solut ions were J.) ore success ful in photooxidation rates was investigated degrading the compo'md, then a lower over a pH range from 3 to 11 in 2 concentration of dye, I mb/l was inves­ pH-unit intervals. Samples were buffered tigated. If, however, the 5 mg/l concen­ to prevent pH drift during experiments. tration of dye was more successful in

25 I III I,

Table 4. Buffer solutions utilized for investigation of the effect of pH on the sensitized photolysis rates.

Stock Buffer Solutions pH D

3 844 ml 156 ml 5 486 ml 514 ml 7 130 ml 72 ml 9 40 ml 460 ml 11 500 ml 20 ml A11so1utions were-dTluteato-2-rnersro1:lowing the addit ion of the two stock solutions.

Solution A: 0.1 M Citric acid B: 0.2 M dibasic sodium phosphate C: 0.2 M sodium carbonate D: 0.2 M sodium bicarbonate

,.", 0\

Table 5. Experimental design configuration*.

Reactors Prepared pH3 pHS pH7 pH3 pH5 pH7 pH3 pH5 pH7 pH3 pHS pH7 pH3 pHS pH7 pH9 pH22 pH9 pHll pH9 pHll pH9 pHll pH9 pHll Methylene blue concentration 5 mg/l S mg/l 5 mg/l S mg/l 5 mg/l

Test compound acridine Acridine Acridine Acridine acridine

Sample time (hrs) 0 4 12 24 48 *This procedure was repeated in triplicate for methylene blue concentrations or-2 mg/l, and either 10 mg/l or 1 mg/l. photolys is, then a higher concentrat ion function of initial substrate concentra­ of dye, 10 mg/l was used. tion was analyzed using concentrations of 0.5, 1.0, 2.0, 5 .0, and 10. a mg/l, Due to the bleaching of the dye while the in it ial concentrat ions of during lengthy photolysis runs, the acridine used were 1.0, 2.0, 5.0, 10.0, concentration of methyl ene blue was and 15 . 0 mg/l. monitored every 4 hours and adjusted as necessary. In this study phase, methylene blue concentration and pH conditions used At the end of a desired time were those determined to be optimal in interval, the five reactors used initial experiments. Anthracene was for that sampling time were removed examined using 2 mg/l methylene blue from the photobench. To account for buffered to pH 7 and with no methylene evaporation, the solutions were poured blue present at pH 9. Acridine was into 250 ml glass graduated cyl inders subjected to photolysis conditions of and the volumes were adjusted to 200 mI. 2 mg/l of methylene blue in a pH 5 A 100 ml sample was then wi thdrawn and solution. prepared for ext ract ion and compound analysis. Actinometry

Effect of substrate mixtures Artificial light sources may fluctuate during operation and dete­ To investigate the possibility of riorate wi th age making analys is of PNA compounds acting as sensitizers in light output essential. A tryptophan­ photol ys is reac t ions (Foote 1968), methylene blue actinometer was utilized sample mixtures combining anthracene, to monitor light intensity at the acridine, and quinoline, buffered to pH maximum absorbing wavelengths of methyl­ 7, were examined both with dye (2 mg/l ene blue (570 nm to 700 nm) (Figure 6). methylene blue) and without dye. Due to Methylene blue absorbs visible light the volatil ity of naphthalene and the energy, sensitizes the photodegradation long photolys is time neces sary when of tryptophan, and produces a tryptophan us ing a mixture of compounds, naphtha­ decomposition rate directly proportional lene was omitted. to the light intensity output of the lamps. The disappearance of tryptophan Samples were prepared using a 1 was measured on a Beckman DU UV Spectro­ percent acetonitrile co-solvent and the photometer at 280 nm. same chemical concentrations as 1n previous photolysis experiments: The effective path length of light anthracene (1.0 mg/l) , acridine and through the actinometer was adjusted quinoline (5.0 mg/l). by changing the depth of the actinometer solution, as well as the concentration of methylene blue in the reactor. This Effect of initial substrate adjustment was made to ensure that all concentrations of the impinging light was absorbed through the entire solution depth so The effect of initial substrate corrections for light transmitted concentration on the rate of photolysis and/or for the light reflected back from reactions was investigated using anthra­ the bottom of the reactor were not cene and acridine as the photosensitive necessary. substrates. The change in the rate of photodegradation was determined by After initial tests, concentrations analyzing the reaction rates of five of 3 mg/l methylene blue and 20 mg/l initial substrate concentrations. The tryptophan were chosen for act inometer photooxidation of anthracene as a conditions. A volume of 450 ml was

27 r

1.00

0.80

Q) g 0.60 o .I:i... o III .I:i <.( 0.40

0.20

O.OO~----~~~---.r-----~------'------'------'-~~~------r 400 450 500 550 600 650 700 750 800

WAVELENGTH t nm Figure 6. Absorbance spectrum for a 3 mg/l methylene blue solution.

selected to provide a solution depth of experimental conditions used in sample 5 cm in the reaction beakers. photolysis reactions. The reactor was placed under the lamps and the sensi­ An actinometer was prepared prior tized photooxidation reaction was to the start of each experimental allowed to proceed for 15 minutes. The analysis. A 20 mg/l solution of trypto­ reaction was then terminated since phan was buffered to pH 9 using a bicar­ photo products formed began to interfere bonate-carbonate buffer (Table 4). The with absorbance readings. The reactor solution (450 ml) was poured into a re­ was removed from the 1 amps and the action beaker identical to the vessels absorbance of tryptophan was measured. used for the photolysis experiments. The number of moles of tryptophan photo Methylene blue as added to a concentra­ degraded in 15 minutes is direct ly tion of 3 mg/l. Before placing the proport ional to the intensity of the ac t inometer under the photolamps. the lamps at the maximum absorbing initial absorbance of tryptophan was wavelengths of methylene blue. Lamp measured on the UV spectrophotometer. intensity was calculated using Equation The absorbance was converted to initial 10. concentrat ion from a concent rat ion/ absorbance curve previously developed for varying concentrations of tryptophan (10) in buffered 3 mg/l methylene blue solutions. where A glass plate was placed over the rea c tor t 0 s i m u 1 ate the s am e 10 - incident light intensity, W/m2

28 c speed of light, 3 x 10 10 em/ extract, filled, and labeled wi th the sec appropriate sample number. All samples were stored in the dark at -70 G C until median wavelength, 6500 x 10-8 analyzed via gas chromatography. em Constituent Analysis A surface area of solutions exposed to light, cm2 Disappearance of substrate was monitored using a He'"lett-Packard 5880A Extractions Procedure gas chromatograph ~ith a Model 7672A automatic liquid sampler. A 6 ft x 2 mm Extractions were performed using a (ID) glass column packed with 3 percent 100 ml aliquot of sample from the 200 ml SP-2250 on 100/120 Supelcoport was used photolyzed sample. Federal Register, for compound separation. Chromato­ Method 610 (Federal Register 1979) was graphic conditions for acridine, anthra­ followed for the extraction procedure cene, and quinoline were as follows: wi th all volumes decreased by a factor injector port temperature 225 G C , oven of 10 to accommodate a smaller sampl e temperature 200°C (isothermal), detector slze. temperature 250°C, and nitrogen carrier gas flow rate 30 ml/min. A flame The 250 ml separatory funnels used ionization detector was used for com­ in the extraction procedures were acid pound identification. washed with a 10 percent sulfuric acid solution prior to each use. The 100 ml Naphthalene was analyzed using the sample was poured from the sampl ing same isothermal chromatographic condi­ beaker into a separatory funnel, "the tions except the oven temperature was beaker was rinsed with 6 ml of HPLC lowered to 1aO°C to separate it from thl: grade methylene chloride, and the solvent peak. rinsings were added to the contents of the funnel. The separatory funnel was External standard calibration shaken for 2 minutes and the solutions procedures were used for each compound. allowed to separate for 10 mi nutes. Standards were prepared in the same After solvent separation was compl ete, manner as the photolysis samples, using the methylene chloride layer was removed a 1 percent acetonitrile co-solvent. from the bottom of the funnel and Standards were extracted using methylene collected in a 25 ml glass graduated chloride as per the extraction procedure cylinder. To prevent evaporation previously desc ribed. during the procedure, the cylinders were covered with aluminum foil. Toxicity Testing

After the extract was removed, Sunlight has been shown to increase another 6 ml of methylene chloride the toxicity of water soluble components was added to the contents of the funnel of refined petroleum products to fish and the extraction procedure was re­ (Scheier and Gominger 1976) and algae peated twice. The volume of methylene (Larson et a1. 1979). To determine if chloride extract collected in the the photoproducts from the oxidation of graduated cyl inder was recorded to the each compound were actually more toxic nearest 0.5 mI. A concentration factor than the parent compound, each test was calculated for each sample based on mixture containing dye and PNA compound the volume of extract collected, e.g., was analyzed for toxicity. Toxicity 15 ml of extract from a 100 ml sample experiments were performed by a quali­ has been concentrated 6.67 times. Five fied laboratory technician utilizing a ml glass sample vials, with teflon Beckman Microtox'" Toxicity Analysis lined sc rew caps, were rinsed wi th the system. In the Microtox'" system

29 lyophilized marine luminescent bacteria, cons tants and quantum yields obtained. a strain of Photobacterim phosphoreum, The data were divided into two sets. are re-hydrated and used as the test The first set was used to examine the organisms for a rapid (5 to 30 minutes) rate constants and quantum yield values determination of acute toxicity. Light from tripl icate analyses at a constant output of phosphorescent bacteria methylene blue concentration to deter­ exposed to a range of toxicant concen­ mine the rate constant and yield depend­ trations is used to develop an LC50 ency on pH. The second set was used to based on light output rather than mor­ analyze the rate constant and yield of tality as the index of toxic response. each pH level, to determine the depend­ ency of the photolysis rate and quantum The pr1mary advantages of the yield on the concentration of methylene Microtox.... system as an acute tox1C1ty blue. 'The resul t of an ANOVA was a test are its relatively short duration, calculated F value which was applied to as compared to 48 hours for Daphnia or the homogeneity of variance analysis. fathead minnow standard EPA-approved The F value was compared with the tests, and the fact that the test tabul ated critical value for F. If the organisms are cultured by one lab calculated F from the ANOVA was greater (Beckman) which, in theory, should than or equal to the critical value, reduce response variability. then the variance of the samples was significantly different from each other Separate experimental runs we re indicating that a change in pH and/or performed for the toxicity analysis. dye condition did affect the rate of Two organic compounds, acridine in a photodegradation:-- solution of 2 mg/l methylene blue and pH 5, and anthracene without dye at pH 9 The second statistical treatment of with 2 mg/l methylene blue at a pH of 7, the data sets was performed using were used as photooxidation substrates Duncan's New Multiple Range Test, a for toxicity experiments. For each technique for identifying which rate toxicity experiment, the optimum photo­ constants and quantum yields were oxidation condition for individual significantly different. This analysis compounds was used. Samples were taken was applied if a difference in the prior to analysis, at one-half the run sample values had been identified time, and after photolysis. using ANOVA analysis. Duncan's test identifies the statistical equality and Statistical Analysis difference between a set of mean values. It operates on the basis that the Statistical analysis of the photo­ probability of equality decreases as the lysis data was performed using analysis number of means under testing increases of variance (ANOVA) and Duncan IS (Steel and Torrie 1980). Optimum pH and New Multiple Range Test. The analysis dye concentration conditions for the of variance test was performed to photodegradation of individual compounds determine if a difference in treatment were selected based on the resul ts of effectiveness existed. between the rate Duncan's test.

30 EXAMPLE CALCULATIONS

The photolysis experiment involving photooxidation of tryptophan as a first anthracene in a solution containing 5 order reaction, subsequently, only the mg/l methylene blue and buffered to pH 9 initial and final tryptophan concentra­ was selected as an example to demon­ tions were recorded and used to deter­ strate the calculation methods and mine the rate of reaction. The actinom­ procedures used throughout this study. eter was run in conjunction with The photolysis data from this experiment photolysis experiments to determine the is shown in Table 6. Table 7 contains amount of tryptophan degraded, which is pert inent information ut il ized in the the basis for calculations involving analysis of both the actinometer and light intensity, light energy, and anthracene photolysis data. quantum yields.

Substrate Reaction Rate Constants Actinometer reaction rate constants

The first order reaction rate The first order reaction rate cons tant based on the d i sappe arance constant based on the disappearance of anthracene was determined to be of 2.5 mg/l tryptophan during 15 minutes 0.0973 hour-I. The first order rate of irradiation was 0.4741 hour-I. constant, k C , was converted to an Convers ion of this rate, ka , to an initial rate constant, KC, for calcu­ initial rate constant, Ka, was done lation of the overall quantum yield using Equation 13. using Equation 11. Ka = ka (sec-I) x [Trp] (mole/liter) KC = kC(sec-1 ) x [An] (mole/liter) x Volume (liters) (13) x Volume (liters) (11) Substituting data from Table 7 into Substituting values from Table 7 into Equat ic n 1 3: Equation 11: Ka = 0.4741 I hour KC = 0.0973 1 hour hour x 3600 sec hour x 3600 sec x 9.79 x 10-5 m~le x 0.45 liter x 5.62 x 10-6 m~le x 0.20 liter llter llter = 5.80 x 10-9 ~ (14) sec = 3.04 x 10-11 mole (2) sec Light energy

Actinometry The light energy emitted from experimental lamps was calculated During the development of the using Equation 9. tryptophan-methylene blue actinom­ eter, sufficient photolysis time Light energy (9 ) studies were performed to establish the =

31 Table 6. Photolysis data from the experiment involving anthracene with 5 mg/l methylene blue and buffered to pH 9; Run 2.

Anthracene Time Concentration In C/Co (hours) (mg/l)

o 1.37 0.00 4 0.92 -0.40 8 0.65 -0.75 12 0.42 -1.18 24 0.00

Table 7. Photolysis data representing experimental conditions and results for the example photolysis run.

ACTINOMETER

Substrate photolyzed Tryptophan Methylene blue concentrat ion 3 mg/l Reaction pH 9 Initial substrate concentration 22.37 mg/l (9.79 x 10-5 mole/liter) Final substrate concentration 19.87 mg/l (8.58 x 10-5 mole/liter) Amount of Tryptophan degraded 2.5 mg/l (1.21 x 10-5 mole/liter) First order rate equation y = -0.474lx + 0.00 Solut ion volwne 450 ml Solution depth 5 cm

PHOTOLYSIS EXPERIMENT

Substrate photolyzed Anthracene Methylene blue c oncentr at ion 5 mg/l Reaction pH 9 Initia~ substrate concentrat ion 1.37 mg/l (5.62 x 10-6 mole liter) First order rate equation y = -0.0973x + 0.001 Solut ion volume 200 ml Solution depth 2 cm

32 where the terms have previously been actinometer was the same sensitizer used described. in the photolysis experiments. However, the actinometer and photolysis samples Substitut ing the act inometer data differed in volume and methylene blue from this run: concentration which affected the amount of light absorbed by each solution. mole To correct for the difference in ab­ Light energy == 1.21 x 10-5 liter sorbance, the percent of absorbed light, based on depth of solution and dye x 0.45 liter x 6.023 concentration, was calculated using the Beer-Lambert Law. Table 8 present s the 1023 photons x 1 einstein absorbance and solution depth data x einstein mole ut il ized in calculat ing the percent of light absorbed by the photolyzed samples 1 _ 15 photons and actinometer. x 900 sec - 3.64 x 10 sec A correction factor. Ca. was (15) developed to normal ize the absorb­ ance of the photolysis samples to the Average incident light intensity actinometer. The anthracene example containing 5 mg/l methylene blue in a Incident light intensity was solut ion 2 cm deep absorbed 90 percent calculated using Equation 10. of incident light intensity. The actinometer contained 3 mg/l methylene blue in a solution 5 em deep which 10 == (~~~N) ( ~A ) (10) absorbed 98 percent of the incident light intensity. The correction factor, Ca. for this example was 0.98/0.90 and Light intensity for this experimental was included in the calculations for run was found to be 0.084 W/m2/nm as quantum yield to normalize results to shown in Equation 16. the actinometer.

Quantum yield 10 = (3.64 x 1015 photons) sees The overall reaction quantum yield was calculated using Equation 17. ( 3 x 1010 em/sec

(17) erg-sec x 1 Joule x 6.627 x 10-27 7 photon 10 erg Using data previously calculated:

1 Watt x x = 0.084 W/m2 3.04 x 10-11 mole/sec 1 Joule = 4>rxn 5.80 x 10-9 mole/sec (16) ( 1 mOle) Actinometer correction factor einstein (~)0.90

The sensitizer used for ini t i­ mole 0.0057 (18) ating photooxidation reactions in the == einstein

33 Table 8. Percent light absorbed in solutions of varying depth and methylene blue concentration.

Methylene Blue Average Solution Concent rat ion Absorbance Depth % Light Absorbed (mg/l) 1 cm cell cm 560 to 700 run

1 0.096 2 36

2 0.194 2 59

3 0.346 5 98

5 0.499 2 90

10 0.980 2 100

34 RESULTS

This investigation, among others absorbed, is a measure of the efficiency (Sargent and Sanks 1974, 1976 and Bulla of uti.lization of the light energy and Edgerley 1968) has shown the reac­ absorbed by the system. The efficiency tion rates from sensitized photooxida­ of the photooxidation reaction depends tions to be a function of several upon the rate of reaction relative to process parameters, i.e., pH, sensitizer the rates of all processes leading to concentration, light intensity and the deactivation of the excited state. dissolved oxygen concentration. First order rate kinet ics are commonly used, Triplicate analyses of each exper­ however, for reaction description under iment were performed for statistical a given set of reac tion cond itions, as relevance. Chauvenet's Criterion done in this analysis. (U.S. EPA 1978) was applied to the data to identify outliers. No data fit the Kinetic reaction rate constants criterion so all observed rate constants were calculated for each experiment and reaction quantum yields were. used in based on the disappearance of the parent further statistical analyses. For compound. The data were fitted to a clarity, the calculated means for first-order kinetics equation, i.e., the tripl icate analyses are presented. natural logarithm of the fraction of The supporting data for the mean values substrate remaining as a function of of the rate constants and quantum yields time. Reaction rates were taken as the are contained in Appendix A and B, negative slope of the linear form of respectively. For the statistical the first order equation. analyses see Appendix C of Naeger (1985). The first order rate constants were used to compare results of various Optimum Reaction Conditions treatments studied in order to establish optimum conditions for photolysis, to To determine optimum conditions for evaluate the effect of initial substrate the photooxidation of each caupound, the concentration and the effect of com­ effects of pH and methylene blue concen­ bining substrates on photolysis, and to tration on observed reaction rates were calculate reaction quantum yields examined. A comparison of treatment and half-lives of the organic compounds conditions was based on the magnitude of under investigation. the first order reaction rate constant, i.e., the greater the rate constant, the For aid in evaluating the effec­ more effective the treatment condition. tiveness of a given treatment condition, The same comparison was performed with the efficiency of a photochemical the obtained quantum yields as a means reaction, represented by the reaction of support ing the select ion of opt imum quantum yield, 4>rxn' was calculated treatment conditions for each compound. for each dye sensitized photooxidation experiment and is presented along with Acridine the observed first order reaction rate constants. The quantum yield, given as Sensitized photooxidation rate the number of moles of substrate re­ const,lOts for acridine are shown in acted per mole of photons (einstein) Table 9. Initial experiments were used

35 Table 9. First order kinetic decay constants (l/hour) for the photolysis of 5 mg/l aqueous solutions of acridine.

Reaction Rate Constants (l/hour) Methxlene Blue Concentrations, mg/l pH 10 5 2 1 0

3" 0.0609 0.0582 0.0233 0.0303 (+0.0343)* (+0.0297) (+0.0098) (+0.0703)

5 0.0824 0.1321 0.1293 0.0656 (+0.1552) (+0.1009) (+0.0350) (+0.0539)

7 0.0174 0.0194 0.0157 0.0132 w (+0.0158) (+0.0087) (+0.0020) (+0.0116) 0- 9 0.0032 0.0048 0.0030 0.0070 (+0.0056) (+0.0039) (+0.0096) (+0.0011)

11 0.0087 0.0092 0.0037 0.0060 (+0.0074) (+0.0105) (+0.0021) (+0.0122)

*Numbers in parentheses represent 95 percent confidence intervals. to evaluate the photoosensitization of concentration, for samples photolyzed in acridine with 2 and 5 mg/l methylene solutions of pH 5, 7, 9, and 11, owing blue. In solutions buffered at pH 3 and to the spread of the calculated rate pH 5, the rate of acridine decay was constant values within the 95 percent greater than in neutral and basic confidence intervals. However, in solutions. At a 95 percent confidence solutions buffered to pH 3, a dependency level, there was no significant differ­ on the concentration of methylene blue ence in the degree of treatment in pH 3 was shown. A greater treatment effi­ solutions containing 2 or 5 mg/l methyl­ ciency was exhibited in solutions ene -blue. There was, however, signifi­ containing 2 and 5 mg/l dye at pH 3 than cantly better acridine removal, based on solutions at the same reaction pH with a a greater reaction rate, using a lower and 1 mg/1 methylene blue. dye concentration of 2 mg/l in pH 5 solutions. A lower dye concentration, 1 Re act ion qua n t urn y i e 1 d s for the mg/l methylene blue, was analyzed in sensitized photolysis of acridine are acidic solutions in further acridine shown in Table 10. The efficiency of tests. Since t he rate of disappearance the sensitized photochemical reaction of acridine increased in pH 7,9, and 11 based on quantum yields closely paral­ solutions when the dye concentration leled the magnitudes of the reaction increased from 2 to 5 mg/l, a higher rate constants in that the higher rate concentration of methylene blue was constants were observed in solutions investigated under basic conditions. with the most efficient utilization of absorbed light energy. The photooxidation rate of acridine as a funct ion of pH for each methyl ene Re act ion qua n tum y i e 1 d s a s a blue concentration examined is shown 1n funct ion of pH obtained for the photo­ Figure 7. When 5 mg/l solut ions of lysis of acridine are shown in Figure 9. acridine were photolyzed, the rate of Similar to t1"!e dependency of the rate photolysis was dependent on reaction pH constants on reaction pH, the quantum for each dye concentration examined. In yield was dependent on pH for all dye solutions conta1n1ng no dye, reaction concentrat ions examined. At methyl ene rates were greatest when the solution blue concentrations of 1, 2, and 5 mg/l, was buffered to pH 5, while all other pH pH 5 solutions achieved the highest levels exhibi ted a statist ically equal quantum yield. In solutions containing degree of acridine removal. When 10 mg/1 dye, solutions with a reaction methyl ene blue was present as a sens i­ pH of 7 were statistically more ef fi­ tizer in concentrations of 1, 2, or 5 cient at util izing the absorbed light mg/l, pH 5 solutions produced the most energy, at the 95 percent confidence effective removal of the substrate. The level, than solutions buffered to pH 9 highest dye concentration used, 10 mg/l or 11. methyl ene blue, was appl ied only to pH 7, 9, and 11 solutions. The rate of The dependency of the react ion acridine removal was statistically quantum yield on the concentration of greater in solutions buffered to pH 7 methylene blue is shown in Figure 10. than the other pH levels under these The quantum yields were independent of cond it ions. methylene blue concentration at all pH levels examined except pH 5 when the Photooxidation rates as a funct ion quantum yield observed in 1 mg/l dye of methylene blue are shown in Figure 8 solutions was significantly higher than for each pH level analyzed. Unlike the the efficiency achieved in 2 mg/l solu­ dependency of rate constants on the pH tions. At a 95 percent confidence level, for a given methylene blue concentra­ solutions containing either 1 or 2 mg/l tion, the rate of removal of acridine methylene blue achieved a higher quantum was independent of methylene blue yield than solutions containing 5 mg/l.

37 LEGEND 0.14 Methylene Blue, 10mg/i Methylene Blue, 5mg/l ,fl Melhylene Blue, 2mg/l 0.12 I: \ Methylene Blue. I mg II -, I' \ L. Methylene Blue, Omg/l ::J o // \ .c I : \ I .\ (j) 0.10 I­ Z /I \ ~ I \ (j) Z 0.08 / /' A \ o I ./ \ \ () I :./ \ \ ~ I / \ \ () J/1,// ,',~. \ \ \ W 0.06 ($ : " \ \ o . .' '" \ \ 0:: \ W ... \ \ o ::' ... \. \ 0:: /: " \ \ o 0.04 : .: " \ \ l­ : ..... \ \ (j) rc G5 / •• \ ", '.\ LL " , 0.02 ,

O.OO~------'------~------.------.------r 3 5 7 9 II pH

Figure 7. First order kinetic decay constants for the photolysis of 5 mg/l acri­ dine solutions as a function of pH; comparison of methylene blue con­ centrations of 10, 5, 2, 1, and 0 mg/l.

38 0'" PH 3 pH :; r~'\ pH 7 pH 9 0.12 I \ pH II I \ \ / \ 0.10 ~ \ z \ i:! II) I z \ o 0.08 i u ?:i u ~ 0.06 cc W Q CC o 0.04 l- (/) g; lL 0.02 __ .-0- __ -e------<;>

__ ••• _ ... _ ...... ,e)

. ,····----.:::r.·.· .. ·...... (i) ...... ·0 0.00, , a 4 6 e 10 METHYLENE BLUE CONCENTRATION, mg/l

Figure 8. First order kinetic decay constants for the photolysis of 5 mg/l acri­ dine solutions as a function of methylene blue concentration; comparison of pH 3, 5, 7, 9, and 11 •

. 090

Me1h:rlene Slue. I mQII Methylene Blue, 2mq/l

Mtth)'le"e elut~ 5, fftqlt .~ .075 Mefhylene Blue.IOm(j/1 c~ ~.. '0.060 E 6 ..J UJ ;;: :::E .045 :::J I- Z < :::J 0.030 z o i= u« :%! .015

I 3 7 9 pH

Figure 9. Reaction quantum yields for the photolysis of 5 mg/l acridine solutions as a function of pH; comparison of methylene blue concentrations of 10, 5, 2, and 1 mg/l.

39 I" ,

Table 10. Reaction quantum yields (moles/einstein) for the photolysis of 5 mg/1 aqueous solutions of acridine.

Reaction Quantum Yields (moles/e instein) Methxlene Blue Concentrations, mg/l pH 10 5 2 1

3 0.0165) 0.0241 0.0156 (+0.0160)* (+0.0106) (+0.0101)

5 0.0221 0.0548 0.0866 (+0.0384) (+0.0400) (+0.0447)

7 ~ 0.0042 0.0051 0.0066 0 (+0.0043) (+0.0043) (+0.0014)

9 0.0008 0.0013 0.0014 (+0.0015) (+0.0011) (+0.0038)

11 0.0021 0.0025 0.0016 {+0.0020 (+0.0034) (+0.0010)

*Numbers in parentheses represent 95 percent confidence intervals. .090 pH 3 pH 5 '\. pH 7 c Q) .075 \ pH 9 til . c pH II W '­ \ til Q) E .060 \. o A -1 W \ )- ~ .045 \ :::> ..... z \ \ a .030 z o \ ..... o \

0- __ -----0- __ _ ---- : .....:::..:..:..:..::: =::.~::..:..:..:.:..:.:: :.:.:.:.:..:::::..:.::.:::~:: ... :

o 2 4 6 8 10 METHYLENE BLUE CONCENTRATION, mg/I

Figure 10. Reaction quantum yields for the photolysis of 5 mg/l acridine solutions as a function of methylene blue concentration; comparison of pH 3, 5, 7, 9, and 11.

41 Treatment efficiencies for the The same statistical comparison was removal of acridine are shown in Table made using the entire data set of 11. The values represent the percent of reaction quantum yields observed for the acridine photooxidized in a 48-hour photolysis of acridine. The results are period. Optimum reaction conditions shown in Table 13. Quantum yield values occurred in solutions buffered to pH 5 within the same shaded areas were found regardless of sensitizer concentration. to be statist ically equal; t he darker the shading the more efficient the To identify the optimum conditions utilization of absorbed light energy, fo r ac rid i ne pho todegrada t ion. the leading to a higher quantum yield. The entire data set was statistically entire data set of quantum yield values evaluated using Duncan's New Multiple was examined since the term quantifies Range Test. The data divided into four not only the rate of substrate decay but groups, and examination of Figures 7 and also the amount of light energy absorbed 8 qualitatively confirm the results of to achieve the given rate of decay. The this analysis. Table 12 is a represen­ quantum yield accurately represented the tation of the statistical equality among efficiency of each photochemical react ion rate constants determined for reaction which occurred, and supported the entire acrid ine data set. Rate the d if fe rences found among treatment constants contained within the same conditions based on rate constants. shaded areas were found to be statis­ tically equal; the darker the shading, By including the quantum yields as the more efficient the treatment. a parameter for selecting optimum treatment conditions for acridine Solutions at pH 5 with 1 and 2 photooxidations, differences are more mg/l methylene blue provided the optimum clearly defined. Based on reaction rate treatment conditions for the photodegra­ constants, the optimum process condi­ dation of acridine. Samples treated tions for the removal of acridine with no dye and with 5 mg/l methylene occurred in solutions at pH 5 with blue, both at a pH of 5, were statis­ either 2 or 1 mg/l methylene blue. tically equal to samples containing 2 or Although there was no significant 5 mg/l methylene blue buffered to pH 3. difference between two rate constants at The least efficient conditions were pH 3 the optimal level, there was a signifi­ samples with 1 mg/l methylene blue along cant difference, at the 95 percent with all samples buffered at pH 7, .9, confidence level, between the react ion and 11 at any dye concentration. quantum yields. Solutions buffered to

Table 11. Comparison of treatment efficiencies (%) for the photooxidation of 5 mg/l solutions of acridine during a 48 hour photolysis period.

Methx:lene Blue Concentration, mg/l pH 10 5 2 1 0 "-

3 96.0 95.8 7 8.5 88.2 5 99.9 99.7 100.0 96.1 7 58.8 56.6 57.3 36.9 9 14.3 21.3 14.4 29.2 11 35.8 26.0 16.8 27.4

42 Table 12. Shaded representation of statistical equality among the observed re­ action rate constants for the photolysis of acridine.

REACTION RATE CONSTANTS, II hour Methylene Blue Concentration, mg/I 10 52! o

3

5

7

9

II

Table 13. Shaded representation of statistical equality among the observed re­ action quantum yields for the sensitized photolysis of acridine.

RE ACTION QUANTUM YIELDS, moles/ Einstein Methylene Blue Concentration, mg/I pH 10 5 2

3

5

7

9

t 1

43 pH 5 and conta1n1ng 1 mg/1 methylene mg/l was the resultant rate constant blue were significantly better at independent of reaction pH. At a dye utilizing absorbed light energy than the concentration of 5 mg/l, the reaction solutions conta1n1ng 2 mg/l methylene rate was dependent on solution pH, with blue. Therefore, the optimum treatment a reaction pH of 7 being significant­ condition for the removal of acridine ly better at the 95 percent confidence would be in solutions at pH 5, which level than other pH conditions at th1S exhibited the highest photochemical dye concentration. efficiency. In the absence of methylene blue, Anthracene the reaction rate was dependent on the reaction pH over the entire pH range First order kinetic decay rate studied, increasing as the pH increased constants for the photolysis of 1 mg/l to pH 9, a fter which the reac t ion rate solutions of anthracene are shown in decreased slightly. Table 14. The samples containing no dye resulted in photolysis at least an order Anthracene reaction rates as a of magnitude higher than those samples function of methylene blue concentration containing dye. are shown in Figure 12. The rate of anthracene photolysis was significantly Photolysis rates as a function of better when no dye was present in the reaction pH for the degradation of test solutions. In the sensitized anthracene are shown in Figure 11. Only photooxidation process, there was no at a methylene blue concentration of 2 significant difference in reaction rates

Table 14. First order kinetic decay constants (l/hour) for the photooxidation of 1 mg/l solutions of anthracene.

Reaction Rate Constants (l/hour) MethIlene Blue Concentrations, m~/l pH 5 2 0

3 0.0424 0.0552 0.5207 (+0.0673)* (+0.0220) (+0.3116)

5 0.0796 0.0661 1.0860 (+0.1109) (+0.0449) (+0.2197)

7 0.1595 0.0751 1.3486 (+0.1125) (+0.0474) (+1.1320)

9 0.0601 0.0558 3.571 (+0.0392 ) (+0.0393) (+1.8074)

11 0.0494 0.0513 2.404 (+0.0149) (+0.0093) (+0.4203)

*Numbers 1n parentheses represent 95 percent confidence intervals.

44 3.50

Methylene Blue Omg/I 3.00 ...... Methylene Blue 2 mg/I -I ~ Methylene Blue 5 mg/I :J --- 0 .c -en 2.50 f- Z en~ z 2.00 0 0 ~ 0 W 1.50 0 tl: W 0 tl: 0 1.00 f- en tl: IJ.. 0.50

_------E:>-- ...... 8:...... _. ~ r: ."8 ..••.••••• -<3- ••••••••- ~ ""- .,.... O.OO1------T------,------,------·~·~·-·-·-...-·--=~~·~ :3 5 7 9 II pH

Figure 11. First order kinetic decay constants for the photolysis of 1 mg/l anthracene solutions as a function of pH; comparison of methylene blue

concentrations of 5 t 2t and 0 mg/l

45 3.50 --'-- pH /I _ .... - pH 9 pH 7 pH 5 3.00 I- pH 3 - C/) 2.50 I­ Z ~ C/) o2 2.00 U ~ U oW 1.50 a:: w o a:: o 1.00 I­ C/) a:: ii: 0.50

...0 o. oo+-----,-----=~==::···-~-~-~!..~.. -~-~-~ .. ~.'.~~~ .. ~.:2~: .. ~ o 2 3 4 5 METHYLENE BLUE CONCENTRATION (mg / I)

Figure 12. First order kinetic decay constants for the photolysis of 1 mg/l anthracene solutions as a function of methylene blue concentration; comparison of pH 3, 5, 7, 9, and 11.

46 between solutions containing 2 or 5 mg/l was applied to the rate constants methylene blue, except in pH 7 solutions observed for the photooxidat ion of containing 5 mg/l methylene blue. quinoline to identify if rate constants Further analysis was suspended and were significantly different from research efforts were concentrated on zero. Solutions buffered to pH 5 and exploring the mechanisms involved in the containing 0 or 5 mg/l methylene blue direct exci tat ion and photodegradation achieved reaction rates significantly of anthracene. greater than zero. In addition, rate Treatment efficiencies for anthra­ constants observed for pH 3 solut ions cene degradation are shown in Table IS. wi th no dye, and pH 7 solut ions wi th 5 The percent of anthracene removal for mg/l dye wert significantly greater sensitized photooxidat ions was achieved than zero. All other rate constants during a 24 hour irradiation period, were statistically equal to zero, while the percent removal for the direct indicating that quinoline was relatively photolysis represents a SO minute ir­ unreactive- toward singlet oxygen. radiation period. The photodegradation of quinoline The reaction quantum yields for the as a funct ion of pH is shown in Figure sensitized photooxidation of anthracene 14. When no dye was present, react ion are shown in Table 16. The optimal pH values of 3 and 5 were significantly sensitized photooxidation conditions better, at the 95 percent confidence est a b 1 ish ed for the de grad at ion 0 f level, than all other pH conditions anthracene based on first order reaction tested. In the presence of a sensi­ rates, and the largest overall quantum tizing dye, there was no significant yields, occurred in solut ions buffered difference l.n the effect iveness of to pH 7. compound removal at any pH level.

Figure 13 shows the reaction The photooxidation of quinoline was quantum yields as a funct ion of pH not improved through the use of methyl­ for each dye concentration examined. ene blue as a sensitizer. Due to the Similar to the react ion rate analysis slow react ion rates compared to photo­ results, only at a methylene blue lysis rates of acridine and anthracene, concentration of 2 mg/l was the reaction no further individual testing of the quantum yield independent of pH. At a compound was performed. dye concentration of 5 mg/l, the quantum yield was dependent on pH with solutions The photolysis of 5 mg/l solutions buffered to pH 7 achieving significantly of quinoline over a 72 hour period higher quantum yield values at the 5 resulted in low treatment efficiencies percent confidence level than other pH (Table 18), compared to those achieved conditions. for acridine or anthracene.

Neutral pH conditions appeared to In the presence of 5 mg/l methylene be optimal for the removal of anthracene blue, there was no significant differ­ via sensitized photooxidation based on ence in the photodegradation of quino­ statistical analysis of both reaction line based on first order rates of rate constants and quantum yields. reaction at any pH level. This same conclusion held true for the reaction Quinoline quantum yields as shown in Figure 15 when the quantum yields are plotted as a The first order kinetic decay function of pH. Only solutions buffered constant s for the photooxidation of 5 to pH 5 or 7 achieved quantum yields mg/l of quinoline with no dye and in the which were significantly greater than presence of 5 mg/l methylene blue are zero. Likewise, the sensitized reaction shown in Table 17. The standard t-test rates observed for pH 5 and 7 solutions

47 Table 15. Comparison of treatment efficiencies (%) for the photooxidation of 1 mg/l solutions of anthracene with dye for a 24 hour irradiation period and without dye for a 50 minute irradiation period.

Percent Removal (%) Methylene Blue Concentrations, mg/l pH 5 2 0 24 hour photolysis 50 minute photolysis

3 63.4 87.6 48.4 5 81.3 90.2 65.4 7 100.0 95.3 74.2 9 97.3 87.3 100.0 11 94.1 87.7 100.0

Table 16. Reaction quantum yields (moles/einstein) for the photooxidation of 1 mg/l aqueous solutions of anthracene.

Reaction Quantum Yields (moles/ einstein) Methylene Blue Concentrations, mg/l

pH 5 2

3 0.0025 0.0059 (+0.0037)* (+0.0036)

5 0.0053 0.0071 (+0.0094) (+0.0062)

7 0.0102 0.0080 (+0.0092) (+0.0062)

9 0.0050 0.0058 (+0.0019) (+0.0031)

11 0.0031 0.0054 (+0.0004) (+0.0003)

*Numbers in parentheses represent 95 percent confidence intervals.

48 ·012

c: .010 Q) III -c: W III "Q) .008 0 E.. 0 -I W ./' >- .006 ~ ~ ....:::> z 0 .004 Z ....0 (,)

.0001r------r-~--_.------,_----_,r_----_ 3 5 7 9 II

Figure 13. Reaction quantum yields for the photolysis of 1 mgtl anthracene solu­ tions as a function of pH; comparison of methylene blue concentrations of 5 and 2 mgtl.

49 Table 17. First order kinetic decay constants (l/hour) for the photolysis of 5 mg/l solutions of quinoline.

Reaction Rate Constants (l/hour) Methylene Blue Concentrations, mg/l pH 5 0 3 0.0037 0.0036 (+0.0071)* (+0.0016) 5 0.0026 0.0034 (+0.0020) (+0.0007) 7 0.0022 0.0015 (+0.0018) (+0.0029) 9 0.0030 0.0015 (+0.0053 ) (+0.0030) 11 0.0008 0.0003 (+0.0026) (+0.0009) *Numbers in parentheses represent 95 percent confidence intervals .

.0035

\ \ - .0030 I .... :J \ 0 .c: \ .0025 U) I- Z \ ;: \ U) Z .0020 \ 0 () ~ () W ,0015 Cl a: w -- Methylene Blue. 5 mg / I Cl a: -- Methylene Blue I a mg / I 0 ,0010 \ I- \ U) a: \ lL. \ .0005 \. ~

.00004------~------~----_,------_r 5 7 9 " pH Figure 14. First order kinetic decay constants for the photolysis of 5 mg/l quino­ line solutions as a function of pH; comparison of methylene blue con­ centrations of 5 and 0 mg/1.

50 Table 18. Removal efficiencies (%) for the photolysis of 5 mg/1 solutions of quinoline during a 72 hour photolysis period.

Percent Removal Methylene Blue Concentrations, mg/1 pH 5 0

3 23.5 26.3 5 15.4 23.9 7 16.4 9.6 9 24.4 13.0 11 8.0 9.1

0.0040 c: Q) en -c: W 0.0030 ...... en Q) 0 E Q - ...J 0.002.0 w >- ::e ::> z~ 0.0010 0 z 0 0.0000 ~ (,)

- 0.0010 3 5 7 9 II pH

Figure 15. Reaction quanturr yields for the phytolysis of 5 mg/1 quinoline solu­ tions as a function of pH in a 5 mg/l methylene blue solution (95 percent confidence limits shown for each d~ta point).

5J were the only pH conditions which and sensitized photooxidation conditions produced sens it ized rates of reac tion were investigated for anthracene, choos­ significantly greater than zero. ing test solutions of 2 mg/l methylene blue buffered to pH 7 and pH 9 solutions Naphthalene wi thout dye. Treatment c ond i t ion s utilized as a function of substrate Due to the characteristically concentration were solutions buffered to volatile nature of naphthalene, photo­ pH 5 with 2 mg/l methylene blue. oxidation studies involving 1 mg/l solut ions of naphthalene coul d not be Observed reaction rate constants performed until the rate of volatil i­ from the examination of the ef fect of zation was established. Initial experi­ initial anthracene concentration are ments indicated that a short react ion shown in Table 21, and these rates are time would be necessary since an in­ plotted in Figure 16. The rate of creasingly larger fraction of the direct photooxidation with a mg/l compound was lost to volatilization as methylene blue increased initially with reaction time was lengthened (Table 19). an increase in anthracene concentration The volatilization studies were con­ from 0.5 to 1.0 mg/l. However, the rate duc ted under the same experimental dropped significantly as reactant conditions used for photolysis studies. concentration increased. Solutions in the reactors were placed under photobench lamps, and glass plates used to minimize evaporation were placed No significant difference in treat­ over the reactors. Aluminwn fo it was ment effectiveness was found between then used to completely cover the initial anthracene concentrations with 2 reactors and eliminate the possibility mg/l methylene blue, indicating that the of photolysis. rate of sensitized photolysis is inde­ pendent of initial concentration. In order to minimize the effect of volatilization, all photolysis experi­ Figure 17 shows the sensitized me n t s we rep e r f 0 rIll e din 1 e sst han 6 photooxidation rate of acridine as a hours. For each experiment, an addi­ function of initial substrate concentra­ tional reac tor was inc luded to monitor tion. No significant difference in the amount of naphthalene lost to reaction rates was observed, at the 95 correct for volatilization at each time percent confidence level, for five interval sampl ed. The resul t s of the initial acridine concentrations, indi­ sensitized and direct photooxidation of cat ing that ac rid ine photolys is was naphthalene are shown in Table 20. The independent of initial concentration. sometimes uneven surfaces of the re­ actors on which the glass plates rested, resulted in inconsistent rates of volatilization. Because of the diffi­ Table 19. Percent of naphthalene lost culty in monitoring volatilization and due to volatilization under the seemi ngly unreact ive nature of photolysis reaction condi­ the substrate, no further testing of tions with light eliminated; naphthalene was performed. mean values from duplicate analyses. Initial Substrates Concentration Reaction Time (hours) % Volatilized Optimwn reaction treatment condi­ tions for anthracene and ac ridine were o 6 22.6 chosen for the analysis of the effect of 18 43.0 initial substrate concentration on pho­ 30 61.3 tooxidation reaction rates. Both direct

52 Table 20. First order kinetic rate constants O/hour) for the photolysis of 1 mg/l aqueous solutions of naphthalene. ~

Reaction Rate Constants (1 hour) Methllene Blue Concentrations, mg/l 5 0 pH Uncorrected Corrected Uncorrected Corrected

3 0.3215 -0.0021 0.0767 +0.0015 (+0.0005) (+0.0041) 5 0.2691 -0.0001 0.1498 -0.0006 (+0.0016) (+0.0058) 7 0.2773 -0.0001 0.0223 -0.0477 (+0.0053) (+0.0049) 9 0.2440 +0.0011 0.1002 +0.0036 (+0.0032) (+0.0005) 11 0.3290 +0.0026 0.1397 +0.0064 (+0.0014) (+0.0020) *Numbers in parentheses represent 95 percent confidence intervals.

Table 21. First order kinetic decay constants (l/hour) from the examination of the effect of initial anthracene concentration on the rate of photolysis.

Reaction Rate Constants (l/hour) Anthracene Me thyl ene Blue Methylene Blue Concentration Concentration Concentration mg/l 2 mg/l o mg/l pH 7 pH 9

0.5 0.0544 1.2875 (+0.0955)* (+0.7233)

1.0 0.0597 3.6570 (+0.0666) (+1.0910)

2.0 0.0302 0.5608 (+0.0339) (+0.6171)

5·9 0.0302 0.6264 (+0.0184) (+0.5796 )

10.0 0.0238 0.4427 (+0.0184) (::..0.1594) *Numbers in parentheses represent 95 percent confidence intervals.

53 .....II 4.0

_.- Methylene Blue 0 mg I I pH 9

Methylene Blue 2 mg II pH 7 T.\ 3.5 / . I . \ '- :l 1'1 0 .c I \ - 3.0 1 t- \ Z I I ~ I en I z 2.5 1 I 0 u I I I ~ I i u w 2. I \ 0 1 1 ~ 1 t- , w I I Z 1.5 ::x:: I I 0:: I w J 0 I 0:: ' . 0 I. I I ent- 0:: I [i: Jr-.- ._._.-.-e-._. O. -- ...... --.-- .--.-- '-'-E)

o 2 4 6 8 10 INITIAL SUBSTRATE CONCENTRATION, mg/ I

Figure 16. First order kinetic decay constants as a function of initial anthracene concentration; comparison of methylene blue concentrations of 2 and 0 mg/l.

54 --I I... ::J 0 0.10 ..c ...... I- Z ~ 0.08 (J) Z 0 U

~ 0.06 U W C u r:: w 0.04 Z ~ 0::: W 0 0::: 0.02 0 I- (J) 0::: lL. 0.00 0 :3 6 9 12 15 INITrAL SUBSTRATE CONCENTRATION, mg/l

Figure 17. First order kinetic decay constants as a function of initial acridine concentration in a 2 mg/l methylene blue solution (95 percent confi­ dence limits shown for each data point).

55 The dependency of reaction quantum contained in the mixture of subst rates yields on the initial concentration of with and without dye were statistically acridine and anthracene in solution with equal, based on the standard t-test, to dye was significantly different than the the rate constants observed when the dependency of first order reaction rate compounds were photolyzed individually. constants on initial substrate concen­ However, the reaction rate constants tration. The rate constants, k C , obtained for anthracene when contained observed for both compounds were in­ in a mixture with and without dye were dependent of initial substrate concen­ sign i ficant ly les s, at the 95 percent tration when photolyzed in solutions confidence level using the standard containing the sensitizing dye. In t-test, than the rate constants achieved contrast, the quantum yields were when the compound was photolyzed in­ dependent on the initial concentration dividually. of the compound. Figures 18 and 19 show the observed reaction quantum yields as Similar resul ts were found when a a function of initial concentration comparison of observed quantum yields for anthracene and acridine, respective­ was made, as shown in Table 23. Quino­ ly. The quantum yield increased with line and acridine achieved reaction increasing substrate concentration due quantum yields when in a mixture similar to the util ization of the initial rate to those achieved when the compounds constant, KC, given as the number of were individually photolyzed. Anthra­ moles of substrate reacted per second, cene exhibited a significant decrease in in calculations involving quantum yield. quantum yield when the Icompound was The first order reaction rate constant, photolyzed in the presence of acridine k C , was independent of initial substrate and quinoline. concentrat ion since the constant only represents the rate as a function of Substrate Half-Lives t lme,. l.e., . h-our 1 . Determination of substrate half­ Th e qua n tum y i e 1 d val u e s for lives is a useful means of predicting sensitized photooxidations of both the persistence of chemicals in the compounds increased wi th increas ing enviromnent. The calculated half-lives concentration and although the reaction of organic substrates based on sens i­ ra tes showed no change wi th c oncen­ tized photolysis experiments represent a tration, the mass of compound photo­ necessary parameter for the design oxidized per second in~reased. of sensitized phototreatment lagoons. Half-lives of the substrates used in Substrate Mixtures this study were calculated using Equa­ tion 19 and are presented in Tables 24, Foote (1968) found that some PNA 25, and 26. compounds are capable of act ing as In 2 sensitizers in photochemical experi­ (19) ments. This potent ial react ion was k C investigated by observing the effect of the combination of anthracene, quinoline where and acridine on the rate of photolysis of each of the compounds. The results, tl/2 half-life of compound, days shown in Table 22, compare the reaction rate of the compound when individually k C = first order reaction rate con­ photolyzed wi th the rate of degradation stant, Ilday observed in the compound mixture. Half-lives ranged from 12 minutes The react ion r ate cons tant ob­ (anthracene, direct photolysis, pH 9) to served for quinoline and acridine when 96 days (quinoline, direct photolysis,

56 0.021

0.018 c: Q)

U) -c: IJJ ...... 0.015 U) Q) -0 E.. 0 -l 0,012 IJJ >- ~ :::> I- z 0.009 < :::> 0 z 0 0.006 I-- u < IJJa::: 0.003

0.000 -t------.-----r-----,...------.----- o 2 4 6 8 10 INITIAL SUBSTRATE CONCENTRATION, mg/l

Figure 18. Reaction quantum yields as a function of initial anthracene concentra­ tion in a 2 mg/l methylene blue solution.

57 .12

c:: .10 lU en -c:: W "-en .!!! 0 .08 E 0 - ...J LIJ >- .06 ~ ::::> t- Z 0.04 Z 0 t- O

O~------r------~------~------~------~ :; 6 9 12 15 INITIAL SUBSTRATE CONCENTRATION, mg/I

Figure 19. Reaction quantum yields as a function of initial acridine concentration in a 2 mg/l methylene blue solution .

. 58 Table 22. A comparison of the first order decay reaction rate constants O/hour) for substrate mixture (acridine,S mg/l; anthracene, 1 mg/l; quinoline 5 mg/l) with the photolysis rates of each compound treated individually.

Compound Methylene Blue pH Reaction Rate Constants Concentration (I/houd mg/l Mixture Individually

Quinoline 2 7 0.0028 0.0022 Quinoline 0 7 0.0040 0.0015 Anthracene 2 7 0.0362 0.0751 Anthracene 0 7 0.0573 1.3486 Acridine 2 7 0.0150 0.0157 Ac ridine 0 7 0.0186 0.0132

Table 23. A comparison of quantum yields (moles/einstein) for substrate mixtures (anthracene, 1 mg/l, acridine and quinoline, 5 mg/O with the quantum yields of each compound photolyzed individually.

Compound Methylene Blue pH Reaction Quantum Yields Concentrat ion (mole/einstein) mg/l Mixture Individually

Quinoline 2 7 0.0016 0.0010 Anthracene 2 7 0.0030 0.0080 Acridene 2 7 0.0061 0.0061

Table 24. Calculated half-lives for the photolysis of anthracene under different treatment conditions.

Half-lives (days) Methylene Blue Concentrations, m~/l pH 5 2 0

3 0.68 0.52 0.055 5 0.36 0.44 0.027 7 0.18 0.38 0.021 9 0.48 0.52 0.008 11 0.58 0.56 0.012

59 Table 25. Calculated half-lives for the photolysis of acridine under different treatment conditions.

Half-lives (days) Methylene Blue Concentrations, mgtl pH 10 521 o

3 0.47 0.50 1.24 0.95 5 0.35 0.22 0.22 0.44 7 1.66 1.49 1.84 2.19 9 8.94 6.02 9.53 4.13 11 3.33 3.14 7.74 4.81

Table 26. Calculated half-lives for the photolysis of quinoline under different treatment conditions.

Half-lives (days) Methylene Blue Concentrations. mgtl pH 5 0

3 7.91 8.08 5 11.34 8.49 7 13.15 19.13 9 9.62 19.55 11 34.51 96.27

pH 11). The sensitized or direct use of av ai 1 ab Ie 1 ight energy, the photolysis of anthracene at any pH level sensitizer should absorb as much of the resulted in half-lives of less than 1 light I S emission spectrum as possible. day. Acridine photolysis was found to The degree to which the light source be extremely dependent on reaction pH emission spectrum and the sensitiz­ and sensitizer concentration with er absorbance spectrum overlap is a half-lives ranging from 5 hours to 9 measure of total energy absorbed. days. Quinoline has a much longer Figure 20 depicts the effectiveness of half-life than the other compounds methylene blue as a sensitizer for the studied and would be expected to be more light emission source used. As shown, persistent and difficult to treat. the light intensity output overlaps the entire absorbance spectra resulting in Actinometry light energy being available at all absorbing wavelengths of methylene Only the energy of the wavelengths blue. absorbed by the sens1t1z1ng dye serves to generate the triplet sensitizer and The actinometer system, incor­ in turn singlet oxygen. For efficient porating a methylene blue sensitized

60 III I. ,

1.00 -- Incident Light Intensity 200 ....••. Methylene Blue Absorbance

0.80 160 E c ." (II"'- . E "'- (J) 0.60 120 VI +- 0 +- c 0 0 . ..c . $ \... 0 C'I VI . ~ 0.40 . 80 E ...... 0.20 .. 40 .. . .' 0.00 , , .. , ... , ., ...... •• , .. ' '" ::=:=:--. I o 400 450 500 550 600 650 700 750 800 Wavelength. nanometers

Figure 20. Methylene blue absorbance spectrum compared with the light source emission spectrum. photoreaction, was used to monitor light intensity, but also for the calculation intensity for the maximum absorbing of the overall quantum yields. This wavelength of methylene blue (570 to 700 information provides support for the nm). By summing the light intensities determination of the effectiveness of a emitted by the photobench lamps at given treatment condition. wavelengths between 570 nm and 700 nm, as determined from the data obtained Toxicity Testing from the Optronics Spectroradiometer an average light intensity of 0.095 W/m i /nm Sampl es of anthracene and acridine was established as the effective output were analyzed for toxicity prior to of light per wavelength available for photolysis, at half the reaction time sensitized photoreactions. and after photolysis, using the Beckman Microtoxm acute toxicity test. Optimum The act inometer was run prior to photooxidation conditions for each the s tart of each experiment. The compound were used as the photolysis effective light intensity available per reaction conditions in the toxicity wavelength was calculated using 650 nm analysis. Table 27 shows the mean as the median wavelength in the methyl­ values from duplicate analyses using the ene blue absorbance spectrum and by Microtoxm acute toxicity test. applying Equation 10: The numeric values assigned to the toxicity of a compound represent the (10) percent of sample in a dilution which caused a 50 percent reduct ion in 1 ight output by the photoluminescent bacteria. The intensity measurements as well as The lower the reported toxicity value, the reaction rate of the actinometer are on a scale of 100, the more toxic is the contained in Appendix B. solution, since less sample is necessary to produce a toxic response from the Light emission. for ind ividual runs bacteria, i.e., a 50 percent reduct ion varied from 0.058 to 0.124 W/m2/nm, 1n light output. however, a decrease in 1 ight intens i ty with time was not found, as the inten­ Anthracene was found to be nontoxic sity readings appeared to vary randomly. in all samples tested. Acridine, An average 1 igh t intens i ty for the however, exhibited a toxic response entire experiment was 0.092 W/m2/nm. prior to photolysis, and the degree of toxicity increased during photolysis Includ ing an ac t inome ter in the indicating that the photoproducts experimental design allowed for not only produced were more toxic than the parent the determination of incident light compound.

62 Table 27. Results of the toxicity analysis utilizing the Microtox acute toxicity test.

Methylene Blue Reaction Compound Concentration pH Time Toxic i ty* (mg/1) (hours)

Anthracene 2 7 o NT Anthracene 2 7 12 NT Anthracene 2 7 24 NT Anthracene o 9 o NT Anthracene o 9 0.33 NT Anthracene o 9 0.83 NT Acridine 2 5 o 18.5 Acridine 2 5 24 13. 1 Acridine 2 5 48 13.0

*NT = Non-toxic

63 DISCUSSION

Photochemical reactions are Optimum Reaction Conditions initiated by the absorption of light energy by a molecule and are terminated Given that the quantum yield values with the disappearance of that molecule, observed for the sensitized photooxida­ or with its return to ground state. In tion of each compound closely paralleled the course of a photochemical reaction, the results obtained for the analysis of the excitation energy of the electron reaction conditions using the observed may be released in several ways, leading kinetic rate constants, only the to chemical changes such as formation of reaction rate constants will be used to excited molecules, intramolecular discuss optimum reaction conditions. rearrangement or free radical formation. Essential to a discussion of photo­ The optimum reaction conditions for chemical reactions is an understanding the photodecomposition of the organic of the pathways of excitation and substrates have been established based deactivation possible for a light on the photooxidation conditions when a absorbing molecule. A Jablonski diagram maximum reac tion rate was achieved, (Figure 21) is a schematic representa­ accompanied by a maximum quantum yield t ion of the red is tr ibut ion 0 f energy value. The optimum reaction pH and result ing from the absorbance 0 flight sensitizer concentration for each energy by a molecule (So), The excited compound are presented in Ta bl e 28. states initially produced by absorption Optimum reaction conditions for each of a photon (SI, S2) are almost always compound occurred at different pH values singlet states. A triplet state (TI) is and with varying amounts of methylene the result of an excitation to a singlet blue present. state accompanied by a reversal of the spin of the electron; its energy will be Anthracene photodimerization lower than that of the corresponding excited sinGlet state, since part of the The maximum reaction rates observed energy has been used in reversing the in this study occurred during the direct electronic spin. photolysis of anthracene and were at least several orders of magnitude great­ Although it was not the purpose of er than other photolysis conditions an­ this study to concentrate on the energy alyzed. The absorption spectrum for a I redistribution mechanisms which occurred mg/l solution of anthracene is shown in when the organic compounds 0 f interes t Figure 22. Two s i3ni ficant absorbance were photolyzed, the pathways represen­ peaks occurred at 375 nm and 398 nm. ted in Jablonski's diagram are funda­ Murov (1973) and Tuno (1978) re:;:lOrted mental to the understanding of photo­ that excitation of anthracene molecules chemical reactions. Explanation of the to a singlet state.· can occur from the results required investigation into absorption of light of sufficient energy electronic transit ion pathways, parti­ in the 375 nm range. In the excited cularly intersystem crossing,81+ TI, state, anthracene can form a dimer with T1 + So, and internal conversion, a g r 0 un d s tat e ant h r ace n e mo 1 e cuI e 81 + So' (Neckers 1967, Cowan and Drisko 1976).

65 Internal conversion

Legend

§~§ - Vibrational levels -----i...... - Absorption or emission of radiation ~ - Nonradiative deactivation -- -... -Intersystem crossing and internal conversion So - Singlet ground state

SI Sz - First and second excited singlet states t respectively TI - Lowest triplet energy state

Figure 21. Jablonski's diagram.

Table 28. Optimum reaction conditions for the photodegradation of 5 mg/l solutions of acridine and quinoline and 1 mg/l solutions of anthracene.

Compound Photolxsis Process Direct Sensitized Dye Concentration pH Rate pH Rate (mg/!)

Anthracene 9 3.57 0.1595 7 5 Acridine 5 0.0656 0.1293 5 1 Quinoline 5 0.0034 0.0025a ND 5 aRepresents an average rate for all pH levels ND - No significant difference ( = 0.05) in pH treatments

66 .0.05

0.04

W 0 Z

0.0 I

O.OO-r------r------,r------.------,------.------.------~------_r--- 320 330 340 350 360 370 380 390 400 WAVELENGTH, nm

Figure 22. Absorbance spectrum for a 1 mg/l anthracene solution.

The photodimerization of anthracene was anthracene decreases (Neckers 1967), is one of the first photochemical systems strong evidence supporting the singlet to be extensively investigated. The mechanism. mechanism involves the following reaction steps: Dimerization occurs across the 9, 10 carbon atoms creating the structure An + hv .... Anl* (20) shown in Figure 23.

An + AnI* + hv"" dimer (21) To investigate whether the reaction rates observed for the direct photolysis An + AnI* .... 2An ( 22) of anthracene could be attributed to its dimerization, a glass filter was placed The e:l<;,fited singlet state, represented over a solution of anthracene that was as Anl, is considered the reactive then photolyzed. The filter effectively state for dimerization, rather than an blocked the penetration of light below excited triplet. The fact that as the 420 nm, thereby inhibiting direct concentration of anthracene in solut ion excitation of anthracene. The results, increases, the quantum yield of dimeri­ shown in Table 29, 'indicate that when zation also increases, while the fluo­ the anthracene absorbing wavelengths rescence efficiency (Sl"" So) of excited of light were effectively blocked,

67 hv

Anthra cene Dimer

Figure 23. Photodimerization of anthracene.

Table 29. Change in the concentration of anthracene with time when photolyzed without a filter and with a 420 nm filter.

Time Concentration (mg/l) Time Concentration (mg/l) Hour wi thout fii ter Hour with 420 nm filter

0 1.08 0 1.30 0.167 0.80 0.25 1.30 0.333 0.45 0.50 1.26 0.500 0.20 1.00 1.16 0.833 0.00 2.00 1.30

essentially no reaction occurred, Sens + hv -+ Sens* (23) 1 suggesting that reaction rates observed when the I ight was not filtered repre­ sens~ -+ Sens; (24) sented a direct absorbance of I ight by anthracene and excitation to a singlet Sens; + 3l:g0 2 -+ 16 g02 + Sens (25) state, initiating the photodimerization reaction. The sensitizer absorbs sufficient light energy to raise the molecule to its Effect of sensitizing dye first excited singlet state, Sens~. Intersystem crossing, which is a rad 1- The sensitizer involved 1n the ationless transition from an excited photooxidation process reacts with singlet state to a triplet state, molecular oxygen by transferring its proceeds by a coupl ing between the energy from the triplet excited state. vibrationally nonexcited Sens~. and The mechanism proposed for the energy the isoenergetic vibrational triplet transfer is represented in the following state. This is followed by vibrational equations: relaxation to vibrational sublevel zero

68 of the triplet state. Thus, inter­ In this process, the excited system crossing is an internal con­ singlet anthracene molecule could vers ion wi th a change in the d i rec­ have been quenched by ground state tion of spin (Mousseron-Canet and methylene blue instead of reacting with Mani 1972). Finally, the triplet ground state anthracene molecules to sensitizer transfers its energy to form a dimer. Anthracene would be molecular oxygen, forming a reactive returned to ground state and in turn, singlet oxygen species, with the methylene blue would be excited to the triplet sensitizer returning to ground more reactive triplet energy state. state. The proposed mechanisms for energy Anthracene reactions. Sensiti::ed transfers are given in the following photochemical reaction rates involv.ng equations: anthracene were greatest when the highest dye concentration examined, 5 An + hv -+ AnI* (20a) mg/l methylene blue, was utilized. This condition occurred in solutions buffered AnI* -+ An3 * (20b) to an optimum pH of 7. An3* + Sens -+ Sensl* + An (26) Since methylene blue does not absorb light effectively below 460 nm Sens + hv -+ Sensl* (23) anthracene molecules were capable of absorbing light energy in the wavelength Sens *1 -+ Sens *3 (24) region where methyl ene blue was in­ capable of sensitizing a photoreact ion. Sens3* + 3 E g02 -+ 1.6. g0 2 + Sens (25) The excitat ion energy from the absorp­ tion of light from 350 to 420 nm was The greatest rate of photooxidation of sufficient to elevate anthracene mole­ anthracene would be achieved under cules to the singlet state from which conditions which give the greatest dimerization could occur. Dimerization triplet sensitizer concentration thereby reactions occurred rapidly in the increasing the production of singlet absence of dye. However, the rate of oxygen. Quenching is a diffusion reaction was less in dye solutions than controlled process, and as such the rate when no dye was present so another of triplet sensltlzer formed from the mechanism must have operated which quenching of singlet anthracene would be competed with dimerization. expected to increase as the concentra­ tion of methylene blue was increased. Quenching is a bimolecular process ln which an excited state is converted Acridine reactions. Sensitized to the ground state by transferring its photooxidation experiments involving energy to the quenching molecule, and it acridine showed a strong dependence on in turn is excited to a higher energy methylene blue concent ration in solu­ state. For quenching to occur, the tions huffered to optimum pH conditions quencher must possess a lower energy of 3 or 5. Acridine solutions buffered requirement for excitation than the at neutral and alkal ine pH were nearly energy of the excited molecule. Anthra­ unreactive toward singlet oxygen and cene molecules possess 76 kcal and 42 reaction rates were independent of kcal energy in their singlet and triplet methylene blue concentrations. states, respectively. Methylene blue requires 34 kcal energy to be raised to The maximum reaction rates for its excited state, and as such, is the -sensitized photooxidation of acri­ capable of accepting energy from the dine occurred when solutions con­ excited anthracene molecules in the ~ained the least amount of methylene system. blue. Acridine possesses photochemical

69 characteristics similar to those of examined, there was no significant anthracene in that it can absorb 1 ight difference in the reaction rate con­ energy in a wavelength region where stants between solutions containing a methylene blue cannot. Its triplet sensitizer and those without dye energy is greater than the energy of present. methyl ene blue and the quenching mecha­ nism proposed to occur in photooxida­ Effect of pH tions involving anthracene, is photo­ ahemically allowed. Similar results The effect of pH on the rate of were not obtained for optimum sensitizer photochemical reactions was examined concentration, however. over a pH range from 3 to 11 at 2-pH unit intervals. An optimum reaction pH It is unclear what photochemical was established for each compound tested mechanism was involved when examining and the results are shown in Table 28. the results of the effect of dye concen­ The optimum reaction pH for dye sensi­ tration on the photolysis of ac ridine. tized and direct photolysis of ac ridine It is suggested in the literature was pH 5. Quinoline also achieved that many electron-rich heterocyclic maximum direc t photolysis rates in low organics, such as acridine, are capable pH conditions of 3 and 5. In contrast, of quenching singlet oxygen via a anthr aeene photoehemica lly reacts at charge-transfer mechanism (Wasserman and more rapid rates in solut ions buffered Murray 1979). This reaction involves to neutral and alkaline pH levels. Most interaction of the electron-poor singlet reports in the I iterature (Sargent and oxygen molecule with electron donor Sanks 1976, Bellin and Yankus 1968, molecules to give a charge-transfer Bonneau et al. 1975) cite neutral to complex, which can then dissociate to alkaline pH conditions as optimal for donor and ground state oxygen. photochemical reactions, as was found in the photolysis of anthracene in this The reduced reaction rates in study. acridine solutions containing higher dye concentrations under optimum pH condi­ The low pH conditions utilized when tions could be attributed to a charge­ maximum reaction rates were observed for transfer quenching of tr iplet methyl ene the photooxidation of quinoline and blue by ground state acridine similar to acridine were explained when a compari­ the quenching of singlet oxygen by a son was made of pKa values for these charge-transfer reaction. Methylene compounds with those compounds cited in blue exists in solution as predominantly literature. The pKa values for quino­ a cationic species capable of participa­ line and acridine are 4.9 and 5.5, tion in the proposed charge-transfer respect ively. Bellin and Yankus (1968) reaction with acridine. This reaction reported that sensitized photooxida­ is diffusion controlled and as such, the tions of the amino acids tyrosine and rate of quenching of triplet methylene histidine proceed more efficiently at blue would be expected to increase as alkaline pH and that photooxidation the concentration of methylene blue was reactions are only rarely observable at increased, leading to decreased reaction acid pH. The pKa of tyrosine is 8.40 rates in solutions containing higher while that of histidine is 9.17. dye concentrations. Bonneau et al. (1975) observed the rate of tryptophan photooxidation to be most Quinoline reactions . The effect of efficient at a pH greater than 9 when sensitizer concentration on the rate of sensitized with methylene blue. The pKa photoox id at ion of quinol ine was in­ of tryptophan corresponds with the significant since the rate of decomposi­ results of alkaline pH producing higher tion was not enhanced in the presence of reaction rates reported to be 9.38. methylene blue. At each pH level Finally, Sargent and Sanks (976)

70 suggested alkaline pH conditions for Initial Substrate Concentration the optimum rate of cresol or phenol photooxidation. Both compounds exist as Anthracene and acridine were 50 percent ionized species at pH levels an a 1 y zed t 0 de term i net he e f f e c t 0 f of 10.1 for cresol and 9.89 for phenol. ini tial substrate concentrat ion on the rate of photooxidation. Figures 16 and 17 show the rate of substrate disappear­ It is unclear what causes the ance as a function of substrate concen­ optimum pH for photooxidation to occur tration for anthracene and acridine, at exactly the pKa of the compound. respect ively. It is apparent that compounds possessing high pKa value will achieve optimum Anthracene reaction rates of reactions in alkaline pH solutions and that a low pKa value will Anthracene was photolyzed utilizing effect the optimum photoreaction for a the optimum conditions for both sensi­ compound in low pH solutions. tized and direct photodegradation. The rate of reaction for the direct photo­ The pKa of the photochemical lysis of anthracene increased as the sensitizer also appears to govern initial concentration increased from 0.5 the reactivity of the substrate. Bellin mg/l to 1.0 mg/l. The maximum rate con­ and Yankus (1968) found that the sensi­ stant was achieved at an anthracene con­ tizer, rose bengal, exists in solution centration of 1.0 mg/l. As the anthra­ as an anion due to its low pKa value of cene concentration increased from the 4.3, and that rose bengal photosensi­ opt imum, the rate constant decreased. tized the degradation of high pKa This result was also reported by Sargent compounds, histidine and tryrosine, and Sanks (976) with cresol, i.e .• an more efficiently than the high pKa dye, initial increase in reaction rate as the methylene blue. If the opposite condi­ cresol concentration increased up to an tion is found to be true, in that a high optimum concentration. As the cresol pKa, cationic, dye sensitlZes low pKa concentration increased from that compounds more efficiently, then the optimum, the first-order reaction rate choice of a sensitizer for photochemical constant decreased, but the total weight experiments must include information of cresol oxidized increased. The regarding the pKa of the chemical theory given by these authors for the substrates as well as the sensitizing decrease in the rate of cresol oxidation dyes. beyond the optimum concentration was that cresol oxidation products became Although the choice of sensitizer, involved in secondary reactions that based on pKa values, effects the re­ affected the reaction rate. activity of the substrate, it does not appear to alter the fact that the Due to the tendency for anthracene optimum pH for reaction will occur at to dimerize, a self-interaction model, the pKa of the substrate. Photochemical such as vertical stacking, best explains experiments involving methylene blue its decrease in reaction rate beyond the (pKa 11.6) and high pKa compounds, optimal concentration. The work of Ts'o attained maximum react ion rates in and Chan (1964) has demonstrated that a solutions buffered to the pKa of the number of different types of molecules photolyzed substrate (Bellin and Yankus form complexes of various degrees of 1968, Sargent and Sanks 1976). Maximum self-association depending upon their methylene blue sensitized photoreaction total solut ion concentrat ion. Gonzalez rates for low pKa compounds, acri­ and Langerman (1977) stud ied the self­ dine and quinol ine. were also achieved association of flavin mononucleotide ins 01 uti 0 n s b u f fer ed tot he p Ka 0 f molecules in solution and proposed a these compounds. vert ical stacking configurat ion of the

71 molecule upon dilution based on nuclear An + (An)n-l -+ (An)n magnet ic resonance data. Al though = verifiable evidence was not produced, Gonzalez and Langerman (1977) re­ vertical stacking of anthracene 1S ported that at high solute concentra­ supported by the anthracene quantum tions, vert ical stacking occurred and yield data collected. that configurations ranged from monomers up to polymers of 50 stacked molecules. The nature of the forces driving At the higher anthracene concentrations, the association have been described as 2, 5, and 10 mg/l, self-interaction by hyd rophobic interact ion and sur face vertical stacking is expected to have energy reactions (Kauzmann 1959). increased. The decrease in reaction The hydrophobic interaction occurs since rate observed at the higher concentra­ anthracene is a hydrophobic molecule tions is postulated to be due to the which will tend to self-associate when addition of the nonpolar extraction in aqueous sol ut ions. The sur face solvent, methylene chloride, to the energy model proposes that the important photolyzed samples, thereby reducing the source of interaction energy is the hydrophobic interactions responsible for surface energy required to form cavities the stacking of anthracene molecules. around the molecule in the system. In When the stacks broke apart, the mon­ an association reaction, the cavities omers of anthracene were released into around the two reactant species have a the extract solution, and were subse­ larger total sur face area than the quently measured by gas chromatography. single cavity around the product, so that the tendency to minimize the The rapid rate of reaction in surface energy literally squeezes the solutions which contained 0.5 and 1 mg/l molecules together. anthracene was observed because. the predomi nant reac t ion was the photo­ Gonzalez and Langerman (1977) chemical dimerization of anthracene. reported that at concentrations of The bonded, photochemically formed dimer solute well below their solubility, does not break apart when added to the monomers and dimers were the most nonpolar extraction solvent since the prevalent forms of flavin mononucleotide bonds of dimerization are not due to in solut ion. At the lower concentra­ hydrophobic interaction as with the tions of anthracene, 0.5 and 1.0 mg/l, vert ically stacked polymers of anthra­ the solution likely contained pre­ cene. The photochemical d imer was not dominantly monomers and dimers. A detected under the gas chromatographic concentration of 1 mg/l exhibited a conditions utilized and thus resulted significantly faster reaction rate than in apparent high rates of parent com­ the observed rate for 0.5 mg/l solutions pound disappearance. since the rate of dimerization is diffusion controlled, and the higher Not all of the anthracene molecules anthracene concentration will exhibit a in the solutions of higher concentration higher rate of dimerization. As the without dye participated in the vertical initial concentration of anthracene stacking mechanism given that the increased from 1 mg/l, the tendency react ion rate obtained at these concen­ to sel f-associate increased due to the trations were significantly higher than increase in the number of anthracene the rates for the sensitized photo­ molecules present. The various reaction oxidat ion of anthracene. The data equil ibria are characterized by the suggest that photochemically formed equations of the form: dimers were being produced as well as complexes formed by vertical stacking. An + An -+ (An)2 (25) When the effect of initial anthracene ( 26) concentration on the rate of sensitized

72 photooxidation was investigated, no reaction. Secondly, there must be a significant difference was found between substance present with an energy re­ rates of reaction. In these solutions, quirement for excitation lower than the the sensitized photooxidation process energy available from the sensitizer. was the primary mechanism for the The absorbance spectrum for each com­ disappearance of anthracene, suggesting pound contained in the mixtures is shown that the self-association by vertical in Figure 24. The wavelength region of stacking was inhibited by the presence interest in this experiment ranges from of the charged, cationic sensitizing 350 nm, where the glass cover plates dye, and the interaction of this dye block 1 ight penetration, to 420 nm, with singlet oxygen. where the most actively absorbing compounds in the mixture no longer Acridine reactions effectively absorb visible light.

The sensitized photodf!gradation of Quinoline cannot perform as a acridine showed no significant differ­ photochemical sensitizer since the ence in the rate of reaction at the compound does not absorb light in the various initial concentrations examined. wavelength region of interest. However, Unlike anthracene, acridine did not acridine and anthracene absorb light exhibit a tendency to self-associate energy and could possibly act as sen­ most likely due to the charge of the sitizers since the first criteria, molecule at the nitrogen position. This the absorption of visible light, 1S result 1S supported by the work of satisfied . Sarma et al. (1968) in that self­ association of pyridine rings by The second criteria is that a vertical stacking did not occur due to substance be present whose first excited the positive charge on the pyridine state lies below the energy level of the nitrogen. excited sensitizer so that energy transfer can occur. Anthracene receives Effect of Mixing Substrates sufficient energy from the absorption of light in the 375 to 400 nm wavelength The effect of combining substrates range to be raised to an ex'cited singlet on the rate of photolysis was investi­ (Turro 1978) reqU1r1ng 76 kcal energy gated for solutions with sensitizing dye (Murov 1973). The efficiency of inter­ present and in the absence of a dye. system crossing from Sl + Tl is 75 Experiments without a dye present were percent so the predominant excitei performed to determine whether one or species of anthracene will be T1 more of the organic compounds in the with 42 kcal of energy available. Since mixture was capable of acting as a the energy of formation for the excited photochemical sens1t1Zer. The experi­ state of quinoline and acridine requires ment was repeated with 2 mg/l methylene 62 kcal and 45 kcal, ri:!spectively, blue added to the solutions to investi­ anthracene would be unable to transfer gate the possible competition among its excited energy to these compounds substrates and molecular oxygen for the requiring more energy. Molecular oxygen energy from the triplet sensitizer, and would however, be capable of accepting to determine if singlet oxygen favors energy from the excited anthr,cene the oxidation of one compound over molecule ihus forming either lEg (37 another. kcaU or 6g (22 kcaU singlet oxygen species. In order for a compound to be an effective sensitizer when a sensitizing Acridine, showing a strong absorb­ dye is not present, it must first be ance peak in the wavelength region of capable of absorbing visible light interest, requires 45 kcal of energy to energy to initiate the photochemical be raised to its excited state. An

73 0.400 --- Acridine 5mg/l --- Anlhracene I mg I I ....••.. Quinoline 5 mg II

W 0.320 U «Z CO 0.240 0::o if) Q) « 0.160

0.080

~------" O.OOO~~~--~--~~--~~~~~~~------.------~-----.------'------r------, 320 330 340 350 360 370 380 390 400 410 420 WAVELENGTH, nm

Figure 24. Absorbance spectrum for 5 mgtl acridine solutions, 5 mgtl quinoline solutions and 1 mgtl anthracene solutions.

excited acridine molecule could not rates obtained when the organ1cs were transfer energy to a molecule of quino­ photolyzed individually is shown 1n line since quinoline requires more Table 22. Rates of photoreact ion for energy, 62 kcal, than acridine is acridine and quinoline when in the capable of transferring. Anthracene and mixture without dye were similar to the molecular oxygen would be capable of react ion rates observed when the com­ accepting the energy from acridine since pounds were photolyzed with the dye the energy requirement for excitation individually. If the photodegradat ion for both compounds would be satisified observed for each compound was due to with the energy received from an excited oxidation via singlet oxygen, then it acridine molecule. Acridine could also was likely that one of the compounds in be involved ina photochemical process the mixture without dye was able to termed energy migration (Barltrop and sensitize the reaction and initiate the Coyle 1978) in which the donor and formation of singlet oxygen. Quinoline acceptor molecule are identical and is was not acting as the sensitizing represented by Equation 30: molecule given that it cannot absorb visible light energy above 350 nm. Ac* + Ac -+ Ac + Ac* ( 28) Acridine appeared to sensitize A comparison of observed reaction photochemical reactions as effect­ rates for the photolysis of a mixture of ively as methylene blue as evidenced by organic substrates with the react ion the reaction rates observed for the

74 photolysis of quinoline. The rate of When anthracene was photolyzed in a photodegradation of quinoline in a mixture with methylene blue, the rate of mixture without methylene blue, with reaction was significantly less than acridine acting as the sensitizer, was when photolyzed indivdually with methyl­ statistically equal to the rate of ene blue. The lower react ion rate was decomposition achieved when quinoline due to acridine absorbing light energy was photolyzed individually in the below 420 nm more effect ively than presence of methylene blue. The same anthracene. The primary photoreac­ held true for the photolysis of acridine tion was consequently oxidation via and anthracene as their reaction rates sing let oxygen sensitized by me thylene in a mixture without dye were equal blue and acridine. When indiv idually to the rate of photooxidation achieved photolyzed with methylene blue, anthra­ with methylene blue present when cene can absorb I ight where methyl ene photolyz~d individually. blue cannot, and dimerization occurs, resulting in an increase of the overall Anthracene was the only compound reaction rate. which exhibited a decrease in the rate of reaction when combined with other Analysis of Toxicity organic substrates. When anthracene was photolyzed individually in the absence Of the two organic compounds of methylene blue, dimerization occurred analyzed for toxicity, only acri­ at a rapid rate leading to the high dine exhibited a toxic response when react ion rate. However, when ac ridine exposed to bacteria in the Microtox~ was included in the mixture, anthracene acute toxicity test. Anthracene was was unable to compete for the available found to be non-toxic both prior to and visible light energy since acridine, at following photolysis, in solution with the experimental concent rat ion of 5 methylene blue and in the absence of mg/l, absorbs 1 ight more strongly methylene blue. (Figure 24). The results of the analysis of The reaction rate for the photo­ toxicity are contained in Table 27. lysis of anthracene in a mixture Acridine exhibited a toxic response when without methylene blue was statis­ in solution prior to photolysis. The t ically greater, at the 95 percent degree of toxicity increased when the confidence level, than the photo­ solution was treated with light for 24 lysis rate achieved when methylene blue hours. After photolysis was completed, was added to the mixture. In these following 48 hours of irradiation, the mixtures without dye, the disappear­ degree of toxicity was not significantly ance of anthracene was attributed to changed from the 24 hour results. acrid ine-sensit ized photooxidat ion in addition to photodimerization, although The photoproducts formed during the dimerization occurred at a much slower photooxidation of acridine are more rate. toxi c than t he parent compound. In addition, the stability of these com­ It was shown earlier that acridine pounds was suggested by the fact that was capable of sensitizing photo­ the degree of toxicity did not decrease chemical reactions with an efficiency wi th cont inued irrad iat ion. After 24 equal to methylene blue. If the re­ hours, 95 percent of the acridine had action rate of anthracene in a mixture been photodegraded and following 48 wi thout dye was greater than wi th dye, hours of irradiation acridine was not the difference would equal the rate of detected in solution, yet the continued dimerization when anthracene was coo­ treatment with light was unable to peting with acridine for the available reduce the toxicity by further degrada­ I ight energy. tion of the photoproducts.

75 Prediction of Reaction Rates photooxidation. Quantum yield values obtained when a compound is individually Establishing the reaction quantum photolyzed will not necessarily equal yield based on laboratory kinetic data the yield obtained when the compound is provides a means for estimating environ~ included in a mixture. Reaction path­ mental photolysis rates. The reaction ways for degradation of the compound may quantum yield of cempl ex molecules in differ when combined with other com­ solut ion is usually independent of pounds due to competing processes wavelength and light intensity (Turro involved in the excitation or deact iva­ 1978). Therefore, once the overall tion of substrate molecules. One such reaction quantum yield is determined for process, capable of altering the rate of a speci fic cempound, the react ion rate photodegradat ion and subsequently the for the photolysis of the compound quantum yield of a substrate when under different conditions may be cembined in a mixture, is the absorbance predicted using Equation 8. Equation 8 of 1 igh t energy by compounds compet ing is applicable for environmental photo­ for light energy, for excitation and lysis if conditions of the actinometers initiation of photochemical reactions. of choice resemble the reaction condi­ tions used for the photolysis of the In this study, anthracene achieved organic compound. a quantum yield of 0.008 molel einstein when photolyzed individually with From the results of this study, methylene blue. When the same concen­ environmental photolysis rate predic­ tration of anthracene was photolyzed in tions can be made for each of the a mixture of substrates, the quantum compounds investigated, under any light yield decreased by over 60 percent. The conditions, by establishing the rate of decrease in yield was attributed to the reaction for the actinometer under the absorbance of I igh t energy by 0 ther light conditions to which the compound compounds in the mixture in the same will be subjected. The overall reaction wavelength region at which anthracene quantum yield for the compound of could absorb light and sensitize a interest is independent of light in­ photoreact ion. It therefore, becomes tensity and wavelength and will remain important to evaluate the absorbance constant as will the quantum yield for characterist ics of the waste prior to the actinometer. If the compound is to bench scale analysis. undergo sensitized photolysis ut il izing the same sensitizer as the actinometer, Since photolysis data for in­ Equation 8 can be used to calculate the div idually treated compounds cannot only unknown, the rate constant for be utilized if competing reactions the sensitized photooxidation of the within the mixture occur, bench scale specific compound. Correction for analysis must be performed to determine differences in volume and methylene blue the quantum yield for each component of concentration between actinometer and the mixture. To do this, the sample sample should be made if necessary as must be characterized both qualitatively per the example calculation presented and quantitatively, so that the dis­ earlier. appearance of each component due to photooxidation can be monitored with If, however, the environmental time. photolysis rate prediction involves a complex mixture of organic cempounds, An estimation of optimum pH condi­ the individual quantum yield data frem tions for utilization in bench scale this study could not be utilized. analyses can be obtained through exami­ Quantum yield is not only a measure of nation of pKa values for each waste photochemical efficiency, but also a component. Results from this study measure of processes compet ing wi th indicate that optimum reaction pH is

76 established at the pKa of the substrate c being photolyzed. K Ka

71 ENGINEERING SIGNIFICANCE

The sc 1 enc e 0 f p hotochemi s t ry has intensity and wavelength, the yield been stud ied for hundreds of years. It values obtained for a specific compound is only in the last decade that re­ can be utilized to predict the rate of searchers have begun to apply the photodegradation of that compound fundamental principles of photochemical in a treatment 1 agoon under natural reactions to the treatment of toxic and sunlight conditions. The following hazardous organic compounds. This sensitized phototreatment lagoon design research has demonstrated the utility of exemplifies the usefulness of these photooxidation for the successful treat­ quantum yield data. ment of hazardous waste constituents. Design of a Sensitized Sargent and Sanks (1974, 1976) and Photooxidation Lagoon Watts (1983) provided the most engineer­ ing-oriented approach in the application Determination of flowrate of photooxidation to waste treatment by establishing optimum sensitized photo­ Oil shale retorting operations are oxidation conditions for several organic expected to produce 50 ,000 barrels of compounds. Reac t ion cond i t ions and oil per day, if and when commercializa­ design parameters established by these tion begins (Yen 1978). Nowacki (1980) authors are however, compound specific. estimates that a modified in-situ retort will produce 0.52 barrels of wastewaters This report provides an analysis per barrel of oil produced. The follow­ of photochemical reaction data that ing equation was used to determine the suggests trends in reactivity based on flow of wastewater requiring treatment pKa values of both sensitizer and for the removal of organics from a typi­ substrate, initial substrate concentra­ cal typical modified in-situ retort tion and 1 ight absorbance characteris­ operation. tics of the substrates. Estimates of photochemical reaction rates and optimum reaction conditions for a variety of compounds can now be made based on photochemical trends established in this 55 gal 3.785 liter research. x bbl x gal

A chemical actinometer is essential 3 x 1m 5,413 m3 /day (31) in photochemica: experiments and was 1000 liter = included in this experimental design to monitor 1 ight intensity emitted by the where photobench. The act inometer provides a means for calculating quantum yield Q = wastewater flowrate, m3/day values for each of the organic compounds studied, and with these quantum yield Design considerations data, design of hazardous waste treat­ ment systems using photooxidation The design of a phototreatment reactions becomes possible. Since the lagoon must be based upon the rate­ quantum yield is independent of 1 ight limiting substrate within the mixture of

79 compound s . The 0 rg ani c compounds k = first order reaction rate utilized In this study were character­ constant, day-l ized as typical components of oil shale C = effluent substrate concentra­ retort wastewater. The rate limiting tion, mg/l compound In the mixture was quinoline, exhibiting the slowest reaction rate Co influent substrate concentra­ constant, 0.0028 hour-I, and smallest t ion, mg/l quantum yield, 0.0016 mole/einstein. Watts (1983) has shown that a Table 3 contains a chemical charac­ minimum of 1.0 mg/l dissolved oxygen was terization of typical retort wastewater required to maintain photochemical from the modified in-situ retort pro­ reactions via singlet oxygen. Sufficient cess. The pH of the waste stream is oxygen can be supplied from atmospheric approximately 8 and since the alkalinity oxygen to maintain the required oxygen is extremely high, in excess of 92,000 concentration without the need for mg/l, pH adjustment would be costly due mechanical aeration. to the alkalinity creating a high buffer capacity. Therefore, the pH of the Photooxidation reactions are waste stream will not be adjusted in the independent of temperature from 5°C to design presented. 40°C (Bulla and Edgerley 1968, Watts 1983) and as such, the wastewater The results of this study showed temperature entering the lagoon system that t he observed quant urn yie 1 d and would need no adjustment. first order reaction rate constants from the photolysis of quinol ine were inde­ Actinometry pendent of reaction pH. It was assumed for this design that the quantum yield A methylene blue-tryptophan acti­ for the photolysis of quinoline in a nometer was run under natural light mixture at pH 8 would not be signifi­ conditions so that an estimate of the cantly different than the quantum yield rate of q uinol ine photooxidat ion in observed when in a mixture buffered to sunlight can be made and utilized in pH 7. The quantum yield data utilized lagoon design. for the lagoon design would be 0.0016 mole/einstein, the value achieved in a 2 The actinometer was run on a cloudy mg/l dye solut ion wi th ac rid ine and day so that the predicted rate of anthracene, buffered to pH 7. reaction for quinoline would be based on the lowest light intensity available. To maintain low cost of operation, The rate of tryptophan decay in the the methylene blue concentration actinometer was expected to be greater suggested is 2 mg/l. The lagoon would in natural sunlight than under labor­ be designed to achieve 90 percent atory lamps, so to insure that the rate removal of quinoline. A plug flow was not limited, the tryptophan concen­ reactor design would be utilized tration was doubled. The actinometer by applying the design equation (Metcalf contained 3 mg/l methylene blue in and Eddy 1979). a 450 ml solution which was buffered to pH 9. Table 30 shows the results of the actinometer photolysis. v = ~ Un C/C ) (32) k 0 Determination of rate constants where The first order reaction rate con.... V = volume, m3 stant, k a , was 0.9491 hour-I. Th is rate was converted to an init ial rate Q = flowrate, m3/day constant using Equation 33:

80 Table 30. Results from tryptophan-methylene blue actinometer performed under natural light c.ondi t ions on a cloudy day at 42"N lat ltude.

Time Tryptophan (Hours) Concentration InC/Co (mg/l)

o 43.45 o 0.0833 40.47 -0.071 0.1667 36.91 -0.163 0.2500 34.42 -0.233

Ka 0.9491 1 hour kC 2.43 x 10-11 mole 3600 sec = hour x 3600 sec = --xsec hour

liter x 2.13 x 10-4 mole x 0.45 lit-er x x 1 liter 3.88 x 10-5 mole 0.20 liter

= 2.52 x 10-8 mole (33) 0.0113 0.2706 sec = hour = "';;";'d';;;'a';"y';"';" (35)

The predicted rate of quinoline Design volume photooxidat ion on a cloudy day can be determined using the quantum yield Using the plug flow reactor design obs erved in the 1 abora tory, 0.0016 equation, and assuming a 90 percent mole/einstein, when quinoline was reduction of substrate, the design combined in a mixture and applying volume is: Equat ion 17. V m3 day In 0.5 = 5413 daY 0.2703 5.0 (17) = 46,111 m3 (36)

The correction factor, Ca , for a 2 Design area mg/l methylene blue solution, normalized to a 3 mg/l methylene blue act inometer The area of the lagoon is calcu­ 1S 0.98/0.59 as taken from Table 8. lated using Equation 37.

Solving for the only unknown, KC, A = V/D (37)

0.0016 = KC where 2.52 x 10-8 mole/sec A = area, acres 1 mole 0.98 (34) einstein 0.59 D = depth, cm

The initial rate constant, KC, is 2.43 To determine minimum lagoon depth, x 10-11 mole/sec and is converted to the extinction coefficient for a 2 mg/l the first order rate constant, kC, methylene blue solution is required, and using Equation 35. is given as:

81 _ A mtnlmum of 10 cm for effective light (38) e: - MBJ/, penetration and increasing depth to 61 cm, to account for evaporation, wind where action, eddy diffusion and thermal mixing. e: = extinction coefficient, M-Icm-l Solving Equation 41 results in the area of the lagoon: A average methylene blue absorb- ance from 570 to 700 nm 100 cm A = 46,000 m3 1 x 50 cm x 1 m MB = molar concentration of methyl- ene blue, mole/I 1 ft2 1 acre x x J/, = path length, cm 0.0929 m2 43,560 ft2 = 18.6 acres (41) For a 2 mg/l solution of methylene blue the extinction coefficient is calculated using Equation 39. Based on the results of this research, a photooxidation treatment e: = ____0_._1::-9_4_---:--:- __ x 1 cm lagoon 61 cm deep, covering 19 acres 5.35 x 10-6 mole/liter with a retention time of 8.5 days would be sufficient to remove 90 percent = 36,268 M-lcm-l (39) of the most difficult to treat waste component. Lagoon design can be altered The depth of the lagoon is deter­ for light intensity changes or changes mined by the depth of effect ive light in waste composition. penetration. Complete light absorbance occurs when the system reaches 2.0 Effluent concentrations absorbance units. The minimum depth is then calculated using Equation 40. Effluent concentrations of acridine and anthracene can be predicted by 36 ,268 M-I cm-l = __---.;;;2~ •..;..0__:_-- estimating the initial rate constant of 5 .35 x 10-6 MO) each compound using Equation 34 with the reaction quantum yield observed for each (40) compound and the natural light actinom­ eter data. The estimated initial Solving for the depth, )/" achieves a rate constants for anthracene and minimum depth of 10.3 cm. acridine are 4.55 x 10-11 mole/sec and 9.25 x 10-11 mole/sec, respectively. For a facultative lagoon system, Conversion of the initial rate constant depth can range from 2 to 5 feet (Met­ to a first order reaction rate constant calf and Eddy 1979). Sufficient mixing is performed by following Equation 35. is expected from wind action, eddy Finally, the percent removal of acridifle diffusion and thermal mixing to occa­ and anthracene under the lagoon des ign sionally bring wastewater from the condi t ions is calculated using Equat ion deeper part of the lagoon to the sur­ 36. Substituting the design volume, face. In addition, water loss due to flow rate values and the first order evaporation will require a lagoon depth react ion rate constant. allows for the greater than 10 cm for the most effi­ determinat ion of the only unknown, the cient utilization of light. removal rate achieved for each compound.

A 2 foot (61 cm) lagoon depth was Anthracene is expected to achieve chosen for this design based on a 100 percent removal under the design

82 condition presented, while 99.9 percent need to be implemented to remove of acridine would be removed. acridine prior to photooxidation.

Note that photoproducts from Chemical costs the photolysis of acridine exhibited greater toxicity than the parent Methylene blue costs for the ~iven compound as indicated by the results lagoon design are based on the estimate from this study. The proposed utiliza­ presented by Watts (1983) of $13.00/kg tion of a phototreatment lagoon is as a and adjusted to January 1985 prices pretreatment process prior to biological (Chemical Engineering 1985) as oxidation. The increase in toxicity $12.50/kg. The followins equation was attributed to the photoproducts of us ed to determine the annual chemical acridine would not affect the operation cost for the proposed lagoon. of a biological treatment process if the photoproduct s are found to be more 2 mg/l x 1.43 mgd x 8.34 Ib x $12.60 biodegradable than the refractory parent 106 gal kg compound. If, however, the photo­ products are shown to be as recalcitrant as acridine, in addition to representing a higher de3ree of toxicity. an alternative pretreatment process would = $49,862/year

83 SUMMARY AND CONCLUSIONS

Photooxidation studies were con­ of methylene blue, as well as providing ducted to determine the sensitivity necessary data for the calculat ion of of four selected organic compounds to reaction quantum yields. Although not a oxidation via singlet oxygen produced measure of iubstrate reactivity, quantum from the absorption of light energy. yields were utilized as supporting data Kinetic rate studies involving the to the results obtained and conclusions degradation of the substrates were drawn from comparisons of reaction rate performed and the observed first order constants. Quantum yield values were reaction rate constants were used as a utilized in calculations for prediction measure of the compounds' reactivity and of environmental reaction rate constants susceptibility to photooxidation. which are necessary for engineering Optimum photoreaction conditions for design of photooxidation treatment each compound were estab lished based on lagoons. sensitizer concentration and reaction pH. F rom this research, the fo llowing conc lu.dons can be drawn: The effect of initial subs trate concentration on the rate of react ion 1. Optimum pH for sensitized was investigated using acridine and photooxidation reactions can be pre­ anthracene as photochemical substrates. dicted based on the pKa of the compound. Optimum reaction conditions determined Methylene blue-sensitized photooxidation in earlier experiments were used as the of acridine and quinoline was most react ion condi t ions for this phase of efficient at a reaction pH of 5, the pKa testing. of each compound.

Anthracene, acridine, and quino­ 2. The optimum pH for anthracene line, were included in the analysis of photolysis, in the presence of methylene the effect of combining substrates on blue, occurred at pH 7. In the absence the rate of reaction to determine of methylene blue, photodimerizat ion of whether these organic substrates could anthracene was theorized from experi­ ac t as sens it ize rs in photochemica 1 mental evidence. Optimum react ion pH react ions. for photodimerization occurred at pH 9.

Toxicity analyses were performed to 3. The rate of sens itized photo­ determine if photoproducts from the oxidation of acridine and anthracene are oxidation of anthracene and acridine independent of the initial substrate were actually more toxic than the parent concentration. Although the first order compound. rate of reaction does not change with an increase in substrate concentration, the An actinometer, utilizing trypto­ mass of substrate oxidized per unit time phan and methylene blue in a sensi­ increases with concentration. tized photoreaction, was run routinely throughout the experimentation. The 4. The rate of direct photolysis actinometer results provided a means of of anthracene is dependent on the determining the average incident light initial concentration of anthracene intensity for the absorbing wavelengths and will increase with increasing

85 concentration up to an optimum rate, of photodimerizat ion and photooxidation after which increase in concentration decreased significantly. tend s to cau se s el f-as soc i at ion by vertical stacking and a decrease in 7. Absorbance spectra of com­ react ion rate. pounds to be degraded by photooxidation become critically important to the 5. Acridine, showing strong understanding of overall reactions absorbance of light energy from 350 to when substrates are capable of com­ 420 nm, is capable of sensitizing petitively absorbing visible light photoreact ions involving anthracene and energy and sensitizing photooxidation quinoline as efficiently as methylene reactions. blue. 8. Acridine was toxic prior to and 6. Anthracene absorbs sufficient following photolysis, with the degree of energy from light wavelengths of 350 to toxicity increasing for photolyzed 420 nm to sensitize a self-associated samples. Unidentified reaction products photodimerization reaction, ift addition from the photooxidation of acridine to sensitizing photooxidation reactions appear to be more toxic than the parent via singlet oxygen. However, in the compound as indicated by the Beckman presence of compounds capable of Microtox Toxicity Analyzer. Anthracene successfully competing with anthracene was non-toxic both before and after for absorbance of light energy, the rate photolysis.

86 RECOMMENDATIONS FOR FURTHER STUDY

Further research is necessary to systems can be designed without sensi­ expand the photochemical concepts t izing dye. and engineering applications developed 1n this study. 4. Prediction of reaction rates and optimum photooxidation conditions 1. Prediction of optimum reaction based on a substrates I physical and pH based on the pKa of the substrate chemical properties should be further should be further verified by examining investigated. An effort should be made compounds wi th a wide range of pKa to es tab !ish the correlat ions between values. ring structure, energy of triplet formation and sensitized photooxidation 2. The efficiency of the sensi­ rates. tized photooxidation process based on dye pKa values with respect to that of 5. A correlation between the rate the substrate being photolyzed should be of photooxidation and the amount of established. available energy absorbed by a system per wavelength of light should be 3. An actinometer should be investigated. developed for use in calculating design parameters, such as quantum yield and 6. The nature and biological reaction rate constants, for systems of recalcitrance of photooxidation products substrate mixtures without sensitizing of selected compounds of environmental dyes present so that lagoon treatment significance should be studied further.

87 REFERENCES

Acher t A. J. t and 1. Rosenthal. 1977. Coomes, R. M. 1979. Carcinogenic Dye-sensitized-photooxidation--a testing of oil shale materials. new approach to the treatment of Twelfth Oil Shale Symp. Proc., organic matter in sewage effluents. Co lorado Schoo 1 0 f Mi nes . pp. Water Res. 11:557-562. 311-320.

Altschuller, A. P' t and J. J. Bufalini. Cowan, D. 0., and R. L. Drisko. 1976. 1971. Photochemical aspects of air Elements of organic photochemistry. pollution: A review. Environ. Plenum Press, New York, New York. Sci. and Tech. 5:39-64. 586 p.

Barltrop, J. A., and J. D. Coyle. 1978. Crittenden, B. D., and R. Long. 1976. Principles of photochemistry. John The mechanism of formation of Wiley and Sons. Chichester, polynuclear aromatic compounds in England. 213 p. combustion systems. pp. 209-223. In: R. Freudenthal and P. W. Jones Bellin, J. S., and C. A. Yankus. 1968. (eds.). Carcinogenesis-a compre­ Influence of dye binding on the hens ive survey. Vo 1. 1. Po ly­ sens it ized photooxidat ion of amino nuclear aromatic hydrocarbons­ acids. Arch ives of Biochem. and chemistry, metabolism, and carcino­ Biophysics 123:18-28. genesis. Raven Press, New York, New York. Bonneau, R., R. Pottier, O. Bagno, and J. Joussot-Dubien. 1975. Dinneen, G. u. 1974. Oil shale and its pH dependence of singlet oxygen potential utilization. Symposium production in aqueous solutions Proceedings: Environmental Aspects using thiazine dyes as photo­ of Fuel Conversion Technology. sensitizers. Photochem. Photobiol. EPA-650/2-74-l18. St. Louis, 21: 159-163. Missouri. pp. 341-352.

Brauer, H. D., W. Drews, R. Schmidt, G. Dulin, D., and T. Mill. 1982. Develop­ Gaugl i tz, and S. Hubig. 1982. ment and eva luat ion of sunlight Heterocoerdianthrone: A new act inometers. Envi ron. Sci. and actinometer for the visible (400- Tech. 16:815-820. 580 nm) range. J. of Photochem. 20:335-340. Federal Register. Guidelines estab- 1ishing test procedures for the Bulla, C. D., and E. Edgerley, Jr. analysis of pollutants. U.S. 1968. Photochemical degradation of Environmental Protection Agency. refractory organic compounds. 44(233) December 3. pp. 69514- JWPCF 40:546-556. 69516.

Calvert, J. G., and J. N. Pitts. 1966. Foote, C. S. 1968. Mechanisms of Photochemistry. John Wiley & Sons, photosensitized oxidation. Science New York, New York. 899 p. 162:963-970.

89 Fox, J. P. 1980. Water related impacts Heelis, P. F., B. J. Parson, G. O. of in-situ oil shale processing. Phillips, and J. F. McKellar. LBL-6300. Lawrence Berkeley 1981. The flavin sensitized photo­ Laboratory, Berkeley, California. oxidation of ascorbic acid; A con­ tinuous and flash photolysis study. Fox, J. P., D. E. Jackson, and R. H. Photochem. Photobiol. 33:7-13. Sakaji. 1980. Potential uses of spent shale in the treatment of oil Hicks, R. E., R. F. Probstein, 1. Wei, shale retort waters. Thirteenth D. S. Farrier, J. Lotwala, and T. Oil Shale Symp. Proc., Colorado E. Phillips. 1980. Wastewater School of Mines. Golden, Colorado. treatment and management at oil pp. 311-320. plants. Thirteenth Oil Shale Symp. Proc., Colorado School of Mines. Fruchter, J. S., and C. L. Wilkerson. Golden, Colorado. pp. 321-334. 1980. Characterization of oil shale effluents. Prepared for the Houban-Herin, N., C. M. Calberg-Bacq, J. U.S. Department of Energy, pre­ Piette, and A. Van de Vorst. 1982. liminary draft. PNL-SA-88l6. Mechanisms for dye-mediated photo­ Pacific Northwest Laboratory, dynamic action: Singlet oxygen Richland, Washington. production, deoxyguanosine oxida­ tion and phage inactivating effi­ Gonzalez, M. J., and N. Langerman. ciencies. Photochem. Photobiol. 1977. A the rmodynamic d esc r ip­ 36:297-306. tion of the self-association of flavin monomucleotide. Archives Israelsen, C. E., V. D. Adams, J. C. of Biochem. and Biophysics. Batty, D. B. George, T. C. Hughes, 180: 75-81. A. J. Seierstad, H. C. Wang, and H. P. Kuo. 1980. Use of saline water Gurnani, S., M. Arifuddin, and K. T. in energy development. Water Augusti. 1966. Effect of visible Resources Planning Series. Utah light on amino acids. I. - Trypto­ Water Research Laboratory, UWRL/P- phan. Photochem. Photobiol. 80/04. Utah State University, 5:495-505. Logan, Utah. 128 p.

Harrison, R. M., R. Perry, and R. A. Jones, B. M., R. H. Sakaji, and C. G. Welling. 1975. Review paper. Daughton. 1982. Physico-chemical Polynuclear aromatic hydrocarbons treatment me thods for oil shal e in raw, potable, and waste waters. wastewater: Evaluation as aids to Water Res. 9:331-346. biooxidation. Fifteenth Oil Shale Symp. Proc., Colorado School of Hartstein, A. M., and B. M. Harney. Mines. Golden Colorado. pp. 1981. The Department of Energy's 581-597. oil shale RD & D program: An over­ view. pp. 1-12. In: H. C. Stauffer Kauzmann, W. 1959. Some factors in the (ed.). Oil shale, tar sands and interpolation of protein denatura­ related materials. ACS Sympos ium tion. pp. 1-64. In: C. B. Anfinsen Series 163. Washington, D.C. and M. L. Anson (eds.). Advances in protein chemistry. Academic Hatchard, C. G., and C. A. Parker. Press, New York, New York. 1956. A new sensitive chemical actinometer. II. Potassium ferri­ Kearns, D. R. 1971. Physical and oxalate as a standard chemical chemical properties of singlet actinometer. Royal Soc. Proc. molecular oxygen. Chem Rev. (London). Series A235:518-536. 71 : 395-427.

90 Kearns, D. R., R. A. Hollins, A. U. magnetic resonance spectrometric Khan, R. W. Chambers, and P. studies of carcinogenic polynuclear Radlick. 1967a. Evidence for the aromatic hydrocarbons in tobacco participation of 1 g and 1 g and marijuanna smoke condensates. oxygen in dyesens it ized photo­ Anal. Chem. 48:405-416. oxidation reactions. 1. Jour. Amer. Chem. Soc. 89:5455-5456. Lee, M. L., M. V. Novotny, and K. D. Bartle. 1981. Analytical chemis­ Ke arns, D. R., R. A. Ho 11 i ns, A. U. try of polycyclic aromatic com­ Khan, and P. Radlick. 1967b. pounds. Academic Press, New York, Evidence for the part icipation of New York. 462 p. 1 g and 1 g oxygen in dyesen­ sitized photooxidation reactions. Leighton, P. A. 1961. Photochemistry II. Jour. Amer. Chem. Soc. 89: of air pollution. Academic Press, 5456-5457. New York, New York. 300 p.

Kingsbury, G. L., R. C. Sims, and J. B. McLaren, A. D., and D. Shugar. 1964. White. 1979. Multimedia envjron­ Photochemistry of proteins and mental goals for environmental nucleic acids. A Pergamon Press assessment; Vol. IV. MEG charts Book. MacMillan Company, New York, and background information sum­ New York. 449 p. maries (catagories 13-26). EPA- 600/7-79-l76b. Research Triangle Maase, D. L. 1980. An evaluat ion of Park, North Carolina. 494 p. p,>lycyclic aromatic hydrocarbons from processed oil shales. PhD Klieve, J. R., G. D. Rawlings, and J. R. Dissertation, Utah State Univer- Hoeflein. 1981. Assessment of sity, Logan, Utah. oil shale retort wastewater treat­ me nt and cont ro 1 tech no logy: Maase, D. L., and V. D. Adams. 1983. Phases I and II. EPA 600/7-81-081. P,)lycyclic aromatic hydrocarbons - Cincinnati, Ohio. are they a problem in processed oil shales? UWRL/Q-83/07, Utah Kochevar, I. E., R. B. Armstrong, J. ~ater Research Laboratory, Utah Einbinder, R. R. Walther, and L. C. State University, Logan, Utah. Harber. 1982. Coal tar photo­ toxicity: Active compounds and Mabey, W. R., T. Mill, and D. G. Hendry. action spectra. Photochem. Photo­ 1982. Photolysis in water. pp. bioI. 36 :65-69. 49-102. In: Laboratory protocols for evaluating the fate of organic Larson, R. A., T. L. Bott, L. L. Hunt, chemicals in air and water. EPA- and K. Rogenmuser. 1979. Photo­ 600/3-82-022. Athens, Georgia. oxidation products of a fuel oil and their ant imicrobial act ivity. Malaney, G. W., P. A. Lutin, J. J. Environ. Sci. and Tech. 13:965-969. Cibulka, and L. H. Hickerson. 1967. Resistance of carcino­ Lee, M. L., and R. A. Hites. 1976. genic compounds to oxidation by Characterization of sulfur-contain­ activated sludge. JWPCF 39:2020- ing polycyclic aromatic compounds 2029. inc arb 0 n b 1 a c k . A n a 1. C h em • 48:1890-1893. Mercer, B. W. 1980. Environmental control technology for oil shale Lee, M. L., M. Novotny, and K. D. was t ewa t e r . Bat t e lie, Pac i f i c Bartle. 1976. Gas chromatography/ Northwest Laboratory. Report to mass spect rome t r ic and nuc lear DOE. DOE/EV-0081.

91 Metcalf and Eddy, Inc. 1979. Waste- of photochemistry. pp. 1-21. In: water engineering: treatment, dis­ J. N. Pitts, W. A. Noyes, and G. S. posal, reuse. McGraw Hill, New Hammond (eds.). Advances in photo­ York, New York. 920 p. chemistry. Vol. 1. John Wiley and Sons, Inc. New York, New York. Mousseron-Canet, M., and J. C. Mani. 484 p. 1972. Photochemistry and molecu­ lar reactions. Israel Program Plimmer, J. R. 1971. Principles of for Scientific Translations. photodecomposition of pesticides. Jerusalem, Israel. 268 p. pp. 279-289. In: Degradation of synthetic organic molecules in the Murov, S. L. 1973. Handbook of photo- biosphere. Nat ional Academy of chemistry. Marcel Dekker, Inc., Science, Washington D. C. New York, New York. 272 p. Porter, G., and F. Wilkinson. 1961. Naeger, Betty-Ann. 1985. Engineering Energy transfer from the triplet treatment of hazardous wastewaters state. Royal Soc. Proc. (London). ut iliz i ng dye-sens it ized photo­ Series A264:1-18. oxidation. M.S. thesis, Utah State University, Logan, Utah. 201 p. Rao, T. K., J. L. Eppler, M. R. Guerin, J. J. Schmidt-Collerus, and L. Neckers, D. C. 1967. Mechanistic Leffler. 1979. Biological moni­ organic photochemistry. Reinhold toring of oil shale products and Publishing Corporation, New York, effluents using short-term genet ic New York. 320 p. analysis. In: Oil Shale Sympo­ sium: Sampliftg, analysis and Neff, J. M. 1979. Polycyclic aromatic quality assurance. EPA-600/9-80- hydrocarbons in the environment: 022. Philadelphia, Pennsylvania. Sources, fates and biological ef­ fects. Applied Science Publishers, Rao, T. K., R. W. Larimer, C. E. Nix, LTD. Essex, England. 262 p. and J. L. Eppler. 1981. Muta­ genicity testing of aqueous mate­ Nowacki, P. (ed.). 1980. Health rials from alternate fuel produc­ hazards and pollution control in tion. L. P. Jackson, and C. C. synthetic liquid fuel conversion. Wright, (eds.). Amer. Soc. for Noyes Data Corporation. Park Testing Materials. Special Techni­ Ridge, New Jersey. 511 p. cal Publication 720. Philadelphia, Pennsylvania. Nowacki, P. (ed.). 1981. Oil shale technical data handbook. Noyes Raphaelian, L. A., and W. Harrison. Data Corporation. Park Ridge, New 1981. Organic constituents in pro­ J e rs ey • 309 p. cess water from the in-situ retort­ ing of oil from oil shale kerogen. Pelroy, R. A., and M. R. Petersen. Argonne National Laboratory. 1979. Use of Ames Test in evalua­ AN/PAG-5. Argonne, Illinois. tion of shale oil fractions. Envi­ ron. Health Perspect. 30:191-203. Redente, E. F., W. J. Ruzzo, C. W. Cook, and W. A. Berg. 1981. Re­ Pfeffer, F. M. 1974. Pollutional torted oil shale characteristics problems and research needs for and reclamation. pp. 128-138. In: an oil shale industry. EPA-600/2- K. K. Peterson (ed.). Oil shale, 74-067. Ada, Oklahoma. 86 p. t he en vir 0 nm en t a I c hall eng e s . pitts, J. N., F. Wilkinson, G. S. Colorado School of Mines Press, Hammond. 1963. The vocabulary Golden, Colorado.

92 Santodonato, J., and P. H. Howard. Snipes, W., G. Keller, J. Woog, T. 1981. Azaarenes: sources, Vickray, R. Deering, and A. Keith. distribution, environmental impact 1979. Inactivation of lipid­ and health effects. Hazard containing viruses by hydro­ assessment of chemicals. Current phobic photosensitizers and developments, Vol. 1. 1 :421-440. near-ultraviolet radiation. Academic Press, New York, New Photochem. Photobiol. 29:785- York. 790.

Sargent, J. W., and R. L. Sanks. 1974. Spikes, J. D., and R. Straight. 1967. Light energized oxidation of Sensitized photochemical processes organic wastes. JWPCF 46: 2547- in biological systems. Ann. Rev. 2554. Phys. Chem. 18:409-440.

Sargent, J. W., and R. L. Sanks. 1976. Steel, R. G., and J. H. Torrie. 1980. Dye-catalyzed oxidation of indus­ Principles and procedures of trial wastes. Jour. Environ. statistics. A biometrical ap­ Engin. Div., Am. Soc. Civil Eng. proach. McGraw-Hill Book Company, 102:879-895. New York, New York. 633 p.

Sarma, R. H., P. Dannies. and N. O. Tabek, H. H., S. A. Quave, C. I. Mashni, Kaplan. 1968. Investigations of and E. F. Bar t h • 1981 . B i 0- inter- and intramolecular inter­ degradability studies with organic actions in flavin-adenine and priority pollutant compounds. dinuc leot ide by proton magnet ic JWPCF 53:1503-1518. resonance. Biochemistry. 7:4359- 4367. rs'o, P. O. P., and S. I. Chan. 1964. Interact ion and associat ion of Scheier, A., and D. Gominger. 1976. A bases and nucleosides in aqueous preliminary study of the toxic solutions. II. Association of ef fect s of irradi ated vs. non­ 6-methylpurine and 5-bromouridine irradiated water solub Ie fract ions and treatment of mUltiple equi­ of No. 2 fuel oil. Bull. Environ. libria. Amer. Chem. Soc. 86:4176- Contam. and Toxicol 16:595-603. 4180.

Schmeltz, I., and D. Hoffmann. 1976. Turro, N. J. 1978. Modern molecular Formation of polynuclear aromatic photochemistry. The Benjamin! hydrocarbons from combust ion of Cummings Publishing Co., inc. organic matter. pp. 225-239. In: Menlo Park, California. 628 p. R. Freudenthal and P. W. Jones (eds.>. Carcinogenesis--a compre­ U. S. Department of the Interior. hensive survey. Vol. 1. Poly­ 1973. Final environmental state­ nuclear aromatic hydrocarbons­ ment for the prototype oil shale chemistry, metabolism, and carcino­ leasing program. Vol. I: Regional genesl.s. Raven Press, New York, impacts of oil shale develop­ New York. ment. No. 2400-00785. Washington, D.C. Smith, E. J. 1971. Apparatus for p hot och emica 1 experiment at ion. U. S. Environmental Protection Agency. pp. 3-42. In: J. M. Fitzgerald 1976. Rationale for recommended (ed.). Analytical photochemistry list of priority pollutants. Memo and photochemical analysis-solids, to director of Effluent Guide­ solutions, and polymers. Marcel lines Division. November 19. Dekker, Inc., New York, New York. Washington, D.C.

93 U. S. Environmental Protection Agency. Utah State University, Logan, 1978. Introduction to environ­ Utah. mental statistics-self-instruction course. SI473. #8. Research Yen, T. F. 1978. Oil shale of the Triangle Park, North Carolina. 4 United States-a review. pp. p. 1-17. In: T. F. Yen (ed.). Science and technology of oil Wagner, S., W. D. Taylor, A. Keith, shale. Ann Arbor Science. Ann and W. Snipes. 1980. Effect Arbor, Michigan. of acridine plus near ultra­ violet light on escherichia Young, R. H., K. Wehrly, and R. L. coli membranes and DNA in vivo. Marten. 1971. Solvent effects Photochem. Photobiol. 32:771- in dye-sensitized photooxidation 779. reactions. Jour. Am. Chern. Soc. 93:5774-5779. Wasserman, H. H., and R. W. Murray. 1979. Singlet oxygen. Academic Zepp, R. G., and D. M. Cline. 1977. Press, New York, New York. 684 p. Rates of direct photolysis in aquatic environment. Environ. Sci. Watts, R. J. 1983. The development of and Tech. 11:359-366. design criteria for the construc­ tion of sensitized photooxidation Zepp, R. G., N. L. Wolfe, G. L. Baugh­ lagoons for the treatment of toxic rna n , and R. C. Ho 11 is. 1977 • and biologically recalcitrant in­ Singlet oxygen in natural water. dustrial wastes. PhD Dissertation, Nature 267:421-423.

94 Appendix A

Supporting Data Reaction Rate Constants

95 III

Table 31. Results from triplicate analyses of 5 mg/l acridine solutions; reaction rate constants from sensitized and direct photooxidations.

Met h Y 1 e n e B 1 u e Con c e n t rat ion, m g / 1 10 5 2

pH Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

3 -0.0702 -0.0674 -0.0450 -0.0443 -0.0650 -0.0650

5 -0.0808 -0.0207 -0.1456 -0.0979 -0.1770 -0.1214

7 -0.0138 -0.0247 -0.0136 -0.0219 -0.0202 -0.0152 -0.0166 -0.0150 -0.0156 \D 0\ 9 -0.0058 -0.0023 -0.0016 -0.0049 -0.0063 -0.0032 -0.0072 +0.0004 -0.0023

11 -0.0072 -0.0121 -0.0067 -0.0112 -0.0121 -0.0044 -0.0045 -0.0028 -0.0039 I" ,

Table 31. Continued.

Met h y len e B 1 u e Con c e n t rat ion , m gIl 1 0

pH Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

3 -0.0278 -0.0218 -0.0204 -0.0252 -0.0049 -0.0608

5 -0.1419 -0.1319 -0.1141 -0.0409 -0.0815 -0.0744

7 -0.0151 -0.0079 -0.0167 '",.... 9 -0.0067 -0.0075 -0.0068

11 -0.0043 -0.0115 -0.0021 III lil

Table 32. Results from triplicate analyses of 1 mg/l anthracene solutions; reaction rate constants from sensitized and direct photooxidations.

Met h y 1 e n e B 1 u e Con c e n t rat ion , m g / 1 5 2 0

pH Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

3 -0.0372 -0.0717 -0.0183 -0.0536 -0.0647 -0.0472 -0.5849 -0.3762 -0.6011

5 -0.0341 -0.0814 -0.1233 -0.0655 -0.0844 -0.0483 -1.1790 -1.0760 -1. 0030

7 -0.1095 -0.1977 -0.1713 -0.0841 -0.0880 -0.0532 -1.1819 -1.8641 -0.9997

9 -0.0768 -0.0973 -0.0663 -0.0539 -0.0410 -0.0725 -3.64 -4.26 -2.81

\0co 11 -0,0536 -0.0520 -0.0425 -0.0541 -0.0470 -0.0527 -2.28 -2.597 -2.336 Table 33. Results from triplicate analyses of 5 mg/l quinoline solution; reaction rate constants from sensitized and direct photooxi­ dations.

Met h Y 1 e neB 1 u e Con c e n t rat ion, m g / 1 5 o

pH Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

3 -0.0069 -0.0015 -0.0026 -0.0042 -0.0036 -0.0029

5 -0.0031 -0.0016 -0.0029 -0.0037 -0.0032 -0.0033

7 -0.0030 -0.0017 -0.0019 -0.0027 -0.0003 -0.0015

9 -0.0054 -0.0013 -0.0023 -0.0027 -0.0003 -0.0014

11 +0.0003 -0.0017 -0.0011 +0.00006 -0.0003 -0.0007

Table 34. Observed reaction rate constants from the photolysis of a mixture containing anthracene,acrLdine and quinoline.

Methylene Blue Compound pH Concentration Run 1 Run 2 Run 3 mg/l

Anthracene 7 2 -0.0310 -0.0371 -0.0404

Anthracene 7 0 -0.0680 -0.0460 -0. OS", 9

Acridine 7 2 -0.0146 -0.0191 -0.0113

Acridine 7 0 -0.0260 -0.0183 -0.0115

Quinoline 7 2 -0.0035 -0.0028 -0.0022

Quinoline 7 0 -0.0073 -0.0026 -0.0020

99 Table 35. Observed reaction rate constants from experiments investi­ gating tne eff-ect of initial anthracene concentration on the rate of photolysis.

Initial Methylene Blue Methylene Blue Substrate Concentration Concentration Concentration 2 mg/1 o mg/1 mg/1 pH 7 pH 9

Run 1 Run 2 Run 3 Run 1 Run 2 Run 3

0.5 -0.0938 -0.0170 -0.0524 -1.572 -1.3003 -0.9901

1.0 -0.0510 -0.0384 -0.0898 -3.826 -4.029 -3.116

2.0 -0.0271 -0.0256 -0.0378 -0.3116 -0.8084 -0.5624

5.0 -0.0500 -0.0282 -0.0249 -0.4143 -0.5887 -0.8763

10.0 -0.0153 -0.0290 -0.0270 -0.3976 -0.4144 -0.5162

Table 36. Observed reaction rate constants from experiments investigating the effect of initial acridine con­ centration on the rate of photolysis.

Initial Substrate Methylene Blue concentration Concentration 2 mg/l mg/1 pH 5

Run 1 Run 2 Run 3 Mean Conf. Limit (95%)

LO -0.0951 -0.0835 -0.0883 -0.0890 :to.0145

2.0 -0.0101 -0.0723 -0.0826 -0.0856 :to.0372

5.0 -0.0783 -0.0968 -0.0808 -0.0853 :!:0.0249

10.0 -0.0787 -0.0788 -0.0790 -0.0788 :!:0.0004

15.0 -0.0834 -0.0911 -0.0909 -0.0885 :!:0.Ol09

100 Appendix B

Supporting Data Reaction Quantum Yields

101 II'III

Table 37. Summary table from triplicate analyses of the photolys is of 5 mg/l acr id ine solut ions; average incident light intensity, reaction rate constants from actinometry, reaction rate constants from photolysis, absorbance correction factor, and reaction quantum yields.

Methylene Blue Average Absorbance React ion a C pH Concentra t ion Run II Inc ident Light K K Correction Quantum mg/l Intensity mole/sec mole/sec Factor Yield W/m2/nm mole/ eins tein

-9 ll 3 1 1 0.085 5.88xl0 4.31xlO- .98/.36 0.0200 -9 ll 3 1 2 0.089 6.12xl0 3.38xlO- .98/ .36 0.0151 -9 ll 3 1 3 O.-lOS 7.22xlO 3. 17xlO- .98/.36 0.0119 -9 10 5 1 1 0.085 5.88xlO 2.20xl0- .98/.36 0.1019 -9 1O 5 1 2 0.089 6.12xl0 2.05xlO- .98/.36 0.0911 -9 10 -c 3 0.105 7.22xlO 1.77x10- .98/.36 0.0668 N 5 1 -9 11 3 2 1 0.086 5.96xl0 6.88xl0- .98/.59 0.0192 -9 10 3 2 2 0.091 6.26xlO 1.0IxlO- .98/.59 0.0268 -9 1O 3 2 3 0.093 6.39xlO 1.01xlO- .98/.59 0.0262 -9 10 5 2 1 0.086 5.96xl0 1. 52xl0- .98/.59 0.0424 -9 10 5 2 2 0.091 6.26xl0 2.75xlO- .98/ • 59 0.0729 -9 IO 5 2 3 0.093 6.39x10 1. 88xlO- .98/.59 0.0490 -9 ll 7 2 1 0.086 5.96xl0 2.58xlO- .98/.59 0.0072 -9 ll 7 2 2 0.091 6.26x10 2.33xlO- .98/.59 0.0062 -9 11 7 2 3 0.093 6.39xlO 2.42xl0- .98/.59 0.0063 I. ,

Table 38. Summary table from triplicate analyses of the photolys is of 5 mg/l acridine solutions; . average incident light intensity. reaction rate constants from actinometry. reaction rate constants from photolysis, absorbance correction factor. and reactiDn quantum yields.

Hethylene Blue Average Absorbance Reaction a C pH Concentration Run n Incident Light K K Correction Quantum mg/l Intens ity mole/sec mole/sec Factor Yield W/m2/nm mole/einstcbl

-9 ll 9 2 1 0.086 5.96xlO 1.l2xlO- .98/.59 0.0031 -9 13 9 2 2 0.091 6. 26xlO 6.21xlO- .98/.59 0.0002 -9 12 9 2 3 0.093 6.39x10 3.57xlO- .98/.59 0.0009 -9 l2 11 l 1 0.086 5.96xlO 6.98xlO- .98/.59 0.0019 -9 12 11 2 2 0.091 6.26xlO 4.35xlO- .98/.59 0.0012 -9 l2 o 11 2 3 0.093 6.39xlO 6.05xlO- .98/.59 0.0016 IJ.) 9 1O 3 5 1 0.073 5.06x10- 1.09xlO- .98/.90 0.0234 9 1O 3 5 2 0.107 7.34xlO- 1.05xlO- .98/.90 0.0155 9 ll 3 5 3 0.104 7.14xlO- 6.98xlO- .98/.90 0.0107 -9 1O 5 5 1 0.073 5.06xlO 1. 25xlO- .98/.90 0.0270 -9 11 5 5 2 0.107 7.34xlO 3.2lx10- .98/.90 0.0048 9 1O 5 5 3 0.104 7. 14xlO- 2.26xlO- .98/.90 0.0345 9 11 7 5 1 0.073 5.06x10- 3.24xl0- .98/.90 0.0070 9 11 7 5 2 0.107 7.34x10- 3. 14x10- .98/.90 0.0047 -9 11 7 5 3 0.104 7.14xlO 2.36x10- .98/.90 0.0036 I..

Table 39. Summary table from triplicate analyses of the photolys is of 5 mg/l acr id ine solut ions; average incident light intensity. reaction rate constants from actinometry, reaction rate constants from photolysis. absorbance correction factor. and react fun quantum yields.

Methylene Blue Average Absorbance React ion a C pH Concentrat ion RUn 1/ Inc ident Light K K Correction Quantum mg/l Intens ity mole/sec mole/sec Factor Yield W/m2/nm molel eins tein

-9 12 9 5 1 0.073 5.06x10 7.60xlO- .98/.90 0.0016 -9 12 9 5 2 0.107 7.34x10 9.78x10- .98/.90 0.0015 -9 12 9 5 3 0.104 7.14x10 4.97xlO- .98/.90 0.0008 9 11 11 5 1 0.073 5.06x10- 1.74xlO- .98/.90 0.0038 -9 11 11 5 2 0.107 7.34x10 1. 88x10- .98/.90 0.0028 0 -9 12 .j:'- 11 5 3 0.104 7.14xlO 6.83x10- .98/.90 0.0011 -9 11 7 10 1 0.085 5.88xlO 2.14x10- .98/1. 00 0.0036 -9 ll 0.0061 7 10 2 0.089 6.12x10 3.83xlO- .98/1.00 -9 11 7 10 3 0.105 7.22xlO 2.11xlO- .98/1.00 0.0029 -9 12 0.0015 9 10 1 0.085 5.88xlO 9.00xlO- .98/1.00 -9 12 9 10 2 0.089 6.l2xlO 3.57xlO- .98/1.00 0.0006 -9 12 0.0003 9 10 3 0.105 7.22xlO 2.48x10- .98/1.00 -9 11 0.0019 11 10 1 0.085 5.88xlO 1.12xlO- .98/1.00 -9 11 0.0030 11 10 2 0.089 6.12xl0 1.88x10- .98/1.00 -9 -11 0.0014 11 10 3 0.105 7.22x10 1.04xl0 .98/1.00 Il Ii II

Table 40. Summary table from triplicate analyses of the photolysis of 1 mg/l anthracene solutions;. average incident light intensity, reaction rate constants from actinometry, reaction rate constants from photolYSiS, absorbance correction factor, and reaction quantum yields.

Methylene Blue Average Absorbance Reaction a C pH Concentration Run II Incident Light K K Correction Quantum mg/l Intensity mole/sec mole/sec Factor Yield W/m2/nm mole/ eins tein

-9 ll 3 2 1 0.122 8.39xlO 1. 67xlO- .98/.59 0.0054 -9 11 3 2 2 0.107 7.36xlO 2.02xl0- .98/.59 0.0075 -9 ll 3 2 3 0.124 8.56xlO 1. 47xlO- .98/.59 0.0047 -9 11 5 2 1 0.122 8.39xl0 2.04xl0- .98/.59 0.0066 -9 11 5 2 2 0.107 7.36xlO 2.63x10- .98/.59 0.0098 0 -9 11 \JI 5 2 3 0.124 8. 56xlO 1. 51x10- .98/.59 0.0048 -9 11 7 2 1 0.122 8.39x10 2.63x10- .98/.59 0.0085 -9 11 7 2 2 0.107 7.36xlO 2.75xlO- .98/.59 0.0102 -9 11 7 2 3 0.124 8.56x10 1. 66x10- .98/.59 0.0053 -9 11 9 2 1 0.122 8.39x10 1. 68xlO- .98/.59 0.0055 -9 11 9. 2 2 0.107 7.36xlO 1. 28x10- .98/.59 0.0047 -9 11 9 2 3 0.124 8.56xl0 2.26x10- .98/.59 0.0072 . -9 ll 11 2 1 0.122 8.39xlO 1. 69xlO- .98/.59 0.0055 -9 ll 11 2 2 0.107 7.36x10 1. 47xlO- .98/.59 0.0054 -9 ll 11 2 3 0.124 8.56x10 1. 65xlO- .98/.59 0.0052 :11 " ,

Table 4l. Summary table from triplicate analyses of the photolysis of 1 mg/l anthracene solutions; average incident light intensity. reaction rate constants from actinometry, reaction rate constants from photolysis, absorbance correction factor, and reaction quantum yields.

Methylene Blue Average Absorbance Reaction a C pH Concentrat ion Run II Inc ident Light K K Correction Quantum mg/l Intens ity mole/sec mole/sec Factor Yield W/m2/nm mole/einstein

-9 ll 3 5 1 0.090 6.20xlO 1. 61xlO- .98/.90 0.0020 -9 ll 3 5 2 0.084 5.80xlO 2. 24xlO- .98/.90 0.0042 -9 12 3 5 3 0.065 4.48xlO 5.72xlO- .98/.90 0.0014 -9 ll 5 5 1 0.090 6.20xlO 1.06xlO- .98/.90 0.0019 -9 11 5 5 2 0.084 5.80xlO 2.54xlO- .98/.90 0.0048 -0 -9 11 0\ 5 5 3 0.065 4.48xl0 3.85xl0- .98/.90 0.0094 -9 11 7 5 1 0.090 6.20xlO 3. 42xlO- .98/.90 0.0060 -9 11 7 5 2 0.084 5.80xl0 6. 17xl0- .98/.90 0.0116 -9 ll 7 5 3 0.065 4.48xlO 5. 35xlO- .98/.90 0.0130 -9 ll 9 5 1 0.090 6.20xlO 2.40xlO- .98/.90 0.0042 -9 ll 9 5 2 0.084 5.80xlO 3.04xlO- .98/.90 0.0057 -9 ll 9 5 3 0.065 4.48xlO 2.07xlO- .98/.90 0.0050 -9 ll 11 5 1 0.090 6.20xlO 1. 67xlO- .98/.90 0.0029 9 11 11 5 2 0.084 5.80xlO- . 1. 62xlO- .98/.90 0.0031 -9 11 11 5 3 0.065 4.48xlO 1. 33xl0- .98/.90 0.0032 I :!l 11 "

Table 42. Sununary table from triplicate analyses of the photolys is of 5 mg/l quinoline solut ions; average inciDent light intensity, reaction rate constants from actinometry, reaction rate constants from photolysis, absorbance correction factor, and reaction quantum yields.

Methylene Blue Average Absorbance Reaction a C pH Concentration Run II Incident Light K K Correction Quantum mg/l Intensity mole/sec mole/sec Factor Yield W/m2/nm mole/ eins te in

-9 11 3 5 1 0.090 6.22xlO 1.49xlO- .98/.90 0.0026 -9 12 3 5 2 0.058 4.01xlO 3.23xlO- .98/.90 0.0009 -9 12 3 5 3 0.079 5.45xlO 5.60xlO- .98/.90 0.0011 -9 . 12 5 5 1 0.090 6.22x10 6.68x10- .98/.90 0.0012 -9 12 5 5 2 0.058 4.01xlO 3.45xlO- .98/.90 0.0009 -9 12 -0 6.25xlO- 0.0013 "'-I 5 5 3 0.079 5.45xlO .98/.90 -9 12 7 5 1 0.090 6.22xlO 6.46xlO- .98/.90 0.0011 -9 12 7 5 2 0.058 4.01x10 3.66xlO- .98/~90 0.0010 -9 12 7 5 3 0.079 5.45xlO 4.09xlO- .98/.90 0.0008 -9 11 9 5 1 0.090 6.22xl0 1.16x10- .98/.90 0.0020 -9 12 9 5 2 0.058 4.01xlO 2.80xlO- .98/.90 0.0008 -9 12 9 5 3 0.079 5.45x10 4.95x10- .98/.90 0.0010 -9 13 11 5 1 0.090 6.22xlO 6.46xl0- .98/.90 0.0001 12 11 5 2 0.058 4.01x10-9, 3.66xlO- .98/.90 0.0010 -9 12 11 5 3 0.079 5.45xlO 2.37xlO- .98/.90 0.0005 I, '1,·1I,

Table 43. Summary table from triplicate analyses examining the effect of combining substrates on the yields and reaction rates; mixtures were photolyzed in 2mg/l methylene blue solutions and buffered to a pH of 7.

Substrate Average Absorbance React ion a C Substrate Concentrat ion Run II Incident Light K K Correction Quantum mg/l Inte~s ity mole/sec mole/sec Factor Yield W/m /nm mole/ eins tein

-9 ll 1 0.092 6.34xlO 2.67xlO- .98/.59 Acr id ine 5.0 0.0059 ll 2 0.092 6.34xlO-9 2.96xlO- .98/.59 Acridine 5.0 0.0078 ll 0.092 6.34xlO-9 1. 75xlO- .98/.59 Acridine 5.0 3 0.0046 12 1 0.092 6.34xlO-9 9.68xlO- .98/.59 Anthracene 1.0 0.0025 ll -9 1. 16xlO- .98/ . 59 Anthracene 1.0 2 0.092 6.34xlO 0.0030 -9 ll 3 0.092 6.34xlO 1. 26xlO- .98/.59 <:> Anthracene 1.0 0.0033 00 12 IJO.092 -9 7.45xlO- .98/.59 Quinoline 5.0 1 6.34xlO 0.0020 12 -9 6.12xlO- .98/.59 Quinoline 5.0 2 0.092 6.34xlO 0.0016 12 '6. 34xlO-9 4.76xlO- .98/ .59 Quinoline 5.0 3 0.092 0.0013 Table 44. Summary table from triplicate analyses examining the effect of initial anthracene concentration on the reaction rates and yields; average incident light intensity. reaction rate constants from actinometry. reaction rate constants from photolysis. absorbance correction factor, and reaction quantum yields in 2 mg/l methylene blue solutions.

Anthracene Average Absorbance Reaction a C pH Concentration Run II Incident Light K K Correction Quantum mg/l Intensity mole/sec mole/sec Factor Yield W/m2/nm mole/ einstein

-9 11 7 0.5 1 0.092 6.34x10 1.46x10- .98/.59 0.0038 -9 12 7 0.5 2 0.092 6.34x10 2.65xlO- .98/.59 0.0007 -9 12 7 0.5 3 0.092 6.34x10 8.18x10- .98/.59 0.0021 -9 ll 7 1.0 1 0.092 6.34x10 1. 59xlO- .98/.59 0.0042 -9 11 7 1.0 2 0.092 6.34xl0 1. 20x10- .98/ . 59 0.0031 0 -9 11 10 7 1.0 3 0.092 6.34xlO 2.80x10- .98/.59 0.0073 -9 11 7 2.0 1 0.092 6.34xlO 1.69xlO- .98/.59 0.0044 -9 ll 7 2.0 2 0.092 6.34xlO 1.60xlO- .98/.59 0.0042 -9 ll 7 2.0 3 0.092 6.34xlO 2.36xlO- .98/.59 0.0062 -9 11 7 5.0 1 0.092 6.34xlO 7.80x10- .98/.59 0.0205 -9 11 7 5.0 2 0.092 6.34x10 4.40xlO- .98/.59 0.0115 -9 11 7 5.0 3 0.092 6.34x10 3.89x10- .98/.59 0.0102 9 11 7 10.0 1 0.092 6.34x10- 4.78x10- .98/.59 0.0125 -9 11 7 10.0 2 0.092 6.34x10 9.05xlO- .98/.59 0.0237 -9 11 7 10.0 3 0.092 6.34x10 8.43xlO- .98/.59 0.0221 Table 45. Summary table from triplicate analyses examining the effect of initial acridine concentration on the reaction rates and yields; average incident light intensity, reaction rate constants from act inometry. reaction rate constants from photolys is, absorbance correct ion factor. and react ion quantum yields in 2 mg/l methylene blue solutions

Acridine Average Absorbance Reaction a C pH Concentration Run (I Inc ident Light K K Correction Quantum mg/l Inte~s ity mole/sec mole/sec Factor Yield W/m /nm mole/ einstein

-9 ll 5 1.0 1 0.092 6.34xl0 2.95xlO- .98/.59 0.0077 -9 11 5 1.0 2 0.092 6.34xlO 2.59x10- .98/.59 0.0068 -9 ll 5 1.0 3 0.092 6.34xl0 2.74xl0- .98/.59 0.0072 -9 ll 5 2.0 1 0.092 6.34xl0 6.32xl0- .98/.59 0.0166 -9 ll 5 2.0 2 0.092 6.34xl0 4.49xlO- .98/.59 0.OH8 2.0 0.092 -9 ll -0 5 3 6.34xlO 5.13xl0- .98/.59 0.0134 -9 10 5 5.0 1 0.092 6.34x10 1. 22xlO- .98/.59 0.0318 -9 1O 5 5.0 2 0.092 6.34xlO 1.50x10- .98/.59 0.0394 -9 1O 5 5.0 3 0.092 6.34x10 1.25xl0- .98/.59 0.0329 -9 1O 5 10.0 1 0.092 6.34xlO 2.44xl0- .98/.59 0.0640 -9 1O 5 10.0 2 0.092 6.34xlO 2.45xlO- .98/.59 0.0640 -9 10 5 10.0 3 0.092 6.34xl0 2.45x10- .98/.59 0.0643 -9 lO 5 15.0 1 0.092 6.34x10 3.88xlO- .98/.59 0.1018 -9 1O 5 15.0 2 0.092 6.34x10 4.24xl0- .98/.59 0.1111 . -9 1O 5 15.0 3 0.092 6.34x10 4.23xlO- .98/.59 0.1109