Air Pollution DYNAMIC IRRADIATION CHAMBER TESTS OF AUTOMOTIV EXHAUST U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service ------- DYNAMIC IRRADIATION CHAMBER TESTS OF AUTOMOTIVE EXHAUST Merrill W. Korth Engineering Research and Development Section Robert A. Taft Sanitary Engineering Center U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service Division of Air Pollution Cincinnati 26, Ohio November 1963 ------- The ENVIRONMENTAL HEALTH SERIES of reports was estab- lished to report the results of scientific and engineering studies of man's environment: The community, whether urban, subur- ban, or rural, where he lives, works, and plays; the air, water, and earth he uses and re-uses; and the wastes he produces and must dispose of in a way that preserves these natural resources. This SERIES of reports provides for professional users a central source of information on the intramural research activities of Divisions and Centers within the Public Health Service, and on their cooperative activities with State and local agencies, re- search institutions, and industrial organizations. The general subject area of each report is indicated by the two letters that appear in the publication number; the indicators are AP - Air Pollution AH - Arctic Health EE - Environmental Engineering FP - Food Protection OH - Occupational Health RH - Radiological Health WP Water Supply and Pollution Control Triplicate tear-out abstract cards are provided with reports in the SERIES to facilitate information retrieval. Space is provided on the cards for the user's accession number and additional key words. Reports in the SERIES will be distributed to requesters, as sup- plies permit. Requests should be directed to the Division iden- tified on the title page or to the Publications Office, Robert A. Taft Sanitary Engineering Center, Cincinnati 26, Ohio. Public Health Service Publication No. 999-AP-5 ------- FOREWORD The problem of vehicle exhaust as an air pollutant has been under intensive study by government and private research agencies for several years. Basic to these studies is the deter- mination of the types of pollutants contained in vehicular exhaust, the photochemical reactions that occur when exhaust is dis- charged into the atmosphere, and the products responsible for various air pollution effects. Photochemical reactions are being studied in detail by the use of 'smog1 chambers, in which vehicular exhaust diluted with air is irradiated to simulate the effects of sunlight in the atmos- phere. This report describes irradiation chamber tests con- ducted by the Division of Air Pollution of the Public Health Service. The work is performed by personnel of the Division's Laboratory of Engineering and Physical Sciences at the Robert A. Taft Sanitary Engineering Center. Preliminary tests were conducted at the Center beginning in February 1960. The dynamic irradiation tests described in this report cover the period from November I960 to May 1961. This work represents one of a series of irradiation studies whose primary object is the determination of photochemical reactions under a selected variety of laboratory conditions representative of urban atmospheres. Future reports will describe additional test projects, all of which are part of the over-all Public Health Service program of research in environ- mental health. ------- CONTENTS Page ABSTRACT vii SUMMARY 1 INTRODUCTION 1 TEST FACILITY AND PROCEDURES 2 Dynamometer System and Test Engine 3 Exhaust Transfer and Dilution System 4 Dilution-Air Purification System .... 6 Irradiation Chambers 7 Exposure Facilities 10 Analytical Procedures 1Z TEST PARAMETERS 13 Exhaust Concentration 13 Average Irradiation Time 14 Fuel Content 14 Other Test Conditions 14 CHEMISTRY OF IRRADIATED EXHAUSTS 16 The NO - NO2 Reaction Process 16 NO — NO2 Photo-oxidation Reactions 16 NO2 — Free-Radical Reactions 16 CHEMICAL EFFECTS 18 NO2 18 Oxidant 21 Hydrocarbons 21 Formaldehyde 24 Other Aldehydes 25 BIOLOGICAL EFFECTS 26 Plants 26 Animals 28 Bacteria 28 SUMMARY OF RESULTS 29 REFERENCES 33 APPENDIX A: RAW DATA FOR IRRADIATION CHAMBER REACTION PRODUCTS 35 APPENDIX B: COMPUTER PROGRAM FOR REDUCTION OF OXIDES OF NITROGEN DATA 47 ------- ABSTRACT As part of an intensive study by government and private agencies the U. S. Public Health Service has built an irradiation chamber facility for investigation of irradiated auto exhaust under mixing conditions similar to those in the atmosphere. The facility consists of a programmed continuous-cycling chassis dynamometer, an exhaust dilution system, a dilution- air purification system, two irradiation chambers, and ex- posure facilities for evaluation of bacteria kill, plant damage, and various effects on small animals. Of the three variables studied during the first test series, the exhaust concentration at the start of irradiation appeared to produce the most significant effects. Fuel composition had a lesser influence. Very little difference was noted in the effects produced at two different average irradiation times. ------- DYNAMIC IRRADIATION CHAMBER TESTS OF AUTOMOTIVE EXHAUST SUMMARY A dynamic irradiation chamber facility was designed and built for investigations of irradiated auto exhaust under condi- tions of continuous mixing. The facility consists of a pro- grammed chassis dynamometer, an exhaust dilution system, a dilution-air purification system, two irradiation chambers, and various exposure facilities. Three variables were considered in this first series of tests: (1) initial exhaust concentration (approximately 13 ppm carbon and 35 ppm carbon), (2) average irradiation time (85 and 120 minutes), and (3) fuel composition (14% and 23% olefins). The effects of varying these test parameters were determined by use of appropriate test criteria including NC>2 formation rate, oxidant production, total hydrocarbon losses and reac- tion of specific species, aldehyde production, plant damage, and bacteria kill. Of the three variables studied, the exhaust concentration at the start of irradiation appeared to produce the most signi- ficant effects. Fuel composition had a lesser influence on some of the test criteria; very little difference was noted in the effects produced at the two average irradiation times. INTRODUCTION Air masses over urban areas continually undergo varying degrees of mixing of new pollutants with existing pollutants. The degree of mixing depends on atmospheric turbulence and the location and movement of parcels of air with respect to pollutant sources. For study of the atmospheric photochemical oxidation of dilute automotive exhaust under conditions that simulate continuous uniform atmospheric mixing of new with 1 ------- 2 IRRADIATION CHAMBER TESTS old pollutants, a dynamic irradiation system has been developed. In this dynamic system, dilute non-irradiated exhaust is con- tinually introduced into the irradiation chamber and dilute irra- diated exhaust is continually withdrawn. TEST FACILITY AND PROCEDURES The test facility used in this study consists of five major components: an automatically cycled chassis dynamometer to control the production of exhaust gases under simulated driving conditions, a two-stage exhaust-transfer and dilution system to dilute the raw exhaust gases to the specified concentrations, a dilution-air purification system, dynamic irradiation chambers for the irradiation of the dilute exhaust gases, and various ex- posure facilities (Figure 1; See also Figures 2-5, 8, 9). AUTOMOBILE DYNAMOMETER UNIT PROPORTIONAL STREAM SPLITTER EXHAUST DILUTION ASSEMBLY (oY-J [L)= i =n ® ^ IL N=— i rr DILUTION-AIR piiRiFir ATIDW i. « 00 V IRRADIATION CHAMBERS TO EXPOSURE FACILITIES AND CHEMICAL ANALYSIS Figure I. Auto exhaust irradiation facility. ------- facility and Procedures DYNAMOMETER SYSTEM AND TEST ENGINE To achieve reproducibility of engine operating conditions and hence of exhaust composition for a 10-hour period (the nor- mal period required for dynamic irradiation chamber runs), a controlled hydraulic-type chassis dynamometer was used with the test vehicle engine (Figure 2). Figure 2. Dynamometer with test vehicle. To assure that dynamometer operation would simulate actual road conditions, the dynamometer was modified to allow simu- lation not only of acceleration, uphill driving, and cruise, in which the engine powers the vehicle (engine-load), but also of deceleration and downhill driving, in which the inertia of the vehicle drives the engine (engine-driven). The modification en- tailed coupling a slave-engine assembly through an overrunning clutch to the power absorption roll to allow contolled motoring ------- IRRADIATION CHAMBER TESTS for the engine-driven condition. Horsepower absorption, pre- set on the dynamometer, established the engine-load conditions. Deviations in dynamometer power absorption and slave-engine power input are on the order of 1. 0%. To provide a continuous, rapid, and reproducing cycle for the test vehicle, a program- ming device for control of the output of the test vehicle engine and the slave engine was necessary. The device that •was de- veloped utilized a cam-operated servo mechanism, which con- trolled throttle position for both the test vehicle engine and the slave engine. Dynamometer horsepower absorption was pre- set prior to each test run. The test vehicle was a 1957 six-cylinder Chevrolet station wagon. Measurements of engine speed, manifold vacuum, car- buretor air flow, and throttle position, made on the test vehicle during actual road operations, were used as the basis for es- tablishing the power requirements for the programming cycle. Time distribution for the programming cycle was based on the frequency distribution of engine modes reported in earlier studies. The programming cycle thus developed produced ex- haust with a ratio of hydrocarbons to oxides of nitrogen of ap- proximately 12:1 (hydrocarbons expressed as ppm total carbon). Following is the sequence of engine conditions. Vehicle Speed, Time, Manifold Vacuum, Power, Condition . . mph mm in. Hg np Engine-load 18 2.2 18.3 2.0 Engine-driven 24 1. 1 23.5 Engine-load 38 2.2 15.5 9.5 Engine-driven 33 1. 1 24.5 Engine-load 36 2.2 16.0 8.0 Engine-driven 29 1.1 24.0 Idle mixture in the carburetor was adjusted daily to compensate for the day-to-day variations in exhaust emissions. The ignition system was checked weekly to assure peak condition of the spark plugs and distributor points. EXHAUST TRANSFER AND DILUTION SYSTEM The exhaust gases generated by the vehicle on the dynamo- meter were transferred to the irradiation chambers through the dilution system by means of a proportional stream-splitter, which separated a constant proportion (approximately 20%) of the total exhaust flow (Figure 3). It was required that the sample and waste gas streams discharge to equal pressure zones, and that the pressure drop be identical through both the sample and ------- Facility and Procedures 5 waste exhaust gas lines. Since the dilution-air system operated at a positive pressure and since the waste portion of the auto exhaust was discharged to atmosphere, it was necessary to pro- vide a point of essentially atmospheric pressure for introduction of the exhaust sample into the dilution-air system. This was done by introducing the exhaust sample into the primary dilution stage at a venturi throat held at atmospheric pressure. SAMPLE TO DILUTION SYSTEM TO SCAVENGE SYSTEM EXHAUST FROM AUTOMOBILE Figure 3. Proportional stream-splitter. Dilution ratios of 400:1 to 1400:1 were used to establish the hydrocarbon concentrations in the irradiation chamber at the levels designated for irradiation studies. To meet these con- ditions, the dilution system was constructed as a two-stage unit, with each stage capable of operating at flow rates of 100 to 250 cubic feet per minute. For this series of tests, dilution in the first stage was held constant at ratios of 19:1 to 21:1 while the second-stage dilution was maintained at 18:1 (for high concentration of exhaust) and at 66:1 (for simulated atmospheric concentrations). ------- 6 IRRADIATION CHAMBER TESTS To prevent contamination of the dilute automotive exhaust with unfiltered air, the dilution air was supplied under sufficient positive pressure to move the gases through the dilution system, the irradiation chamber, and the exposure facilities. Movement of the exhaust gases at relatively low pressure into the first stage of the dilution system, which was nominally at higher pressure, and transfer of the required portion of the gases from the first dilution stage to the second dilution stage were accomplished by introducing the gas into zones of low pressure created at venturi throats. Rapid mixing also was accomplished by the high velocity and turbulence created by introduction at each venturi throat. DILUTION-AIR PURIFICATION SYSTEM The dilution air, with which the exhaust sample is mixed before it is introduced into the irradiation chamber, should be as free as possible of contaminants that would interfere with the photochemical reactions within the chamber. The tempera- ture and humidity of this air should approximately duplicate those conditions found in typical urban atmospheres. To achieve these conditions, the dilution air was passed through an air conditioning system, which included particulate removal and charcoal filters, a cooling and dehumidifying coil, a heating coil, and a humidifier (Figure 4). PARTICIPATE FILTER CHARCOL FILTER OUTSIDE AIR FROM FAN Ji ' 'RAIN —* HEATI HEATING COILS TEMPERATURE CONTROLLER Figure 4. Dilution-air purification system ------- Facility and Procedures The particulate filter was a Cambridge Model ID-1000 ab- solute filter with 1000-cfm capacity. This filter has an efficien- cy of 99. 95% with 0. 3-micron particles. The charcoal filter was a Barneby-Cheyney Model 7-FM cell charged with 45 pounds of activated charcoal. Periodic sampling has shown that a rela- tively constant hydrocarbon content of less than 2 ppm carbon, which is essentially methane, remains in the clean air. A 7-1/2-ton-capacity mechanical refrigeration system pro- vided dehumidification and cooling. This system, together with steam heating coils and a steam-heated humidifier, maintained the diluting air at 75°F ± 3°F and 50% ± 5% relative humidity. IRRADIATION CHAMBERS Two 335-cubic-foot chambers were constructed for the dy- namic irradiation facility. The chambers were designed to operate as ideal dilution volumes into which the raw exhaust gases, diluted to the designated concentration, were continu- ously introduced as an equal quantity of irradiated gases was withdrawn. The chambers were constructed of aluminum with Mylar windows to allow for the irradiation of the dilute chamber gases (Figure 5). Figure 5. Irradiation chamber. GPO 806—304—2 ------- « IRRADIATION CHAMBER TESTS For this test series, rapid mixing, necessary for the dyna- mic chamber to perform as an ideal dilution volume, \vas pro- duced in each chamber by the jet action of two tube-axial fans, which caused approximately 50% of the gas at the fan to be ac- ct-lerated ahead and mixed, under turbulent conditions, within the chamber. Since chamber operation approached the perform- ance of an ideal dilution volume, any element of input gas was rapidly and completely mixed into the entire chamber vol- ume and any element of the output gas was representative of the entire volume. Under such conditions the average irradia- tion time for all molecules of gas in the chamber was equal to: Ta I (1 e-kt) (1) Where: Ta average irradiation time (exposure time) k a v q chamber flow rate v chamber volume t time after irradiation begins The exposure time is therefore determined only by the flow rate through the chamber and the time since start of irradiation. Irradiation of the gas within the chamber was supplied by illumination from sources external to the chamber through windows of 3-mil-thick Type D Mylar. The illumination was provided by two banks of 70 fluorescent tubes each, mounted in two cavity reflectors. The 96-inch T-8 tubes were operated on T-1Z ballasts, which supplied a 25% overload to increase the incident irradiation intensity. Two types of tubes, black light and warm white deluxe, were used in equal proportions to approximate solar radiation between the solar cut-off (about Z900 angstroms) and 3700 angstroms. The choice of Mylar as window material was a compromise among many factors, in- cluding spectral transmission, strength, surface adsorption, and diffusion. Strength and surface characteristics of the Mylar were found to be adequate. As shown in Figure 6, the spectral transmission curve for the Mylar is smooth, with a cut-off point between 3100 and 3200 angstroms, which is somewhat higher than the solar cut-off. The spectral distribution and intensity in the near-ultraviolet range produced in the chamber, as shown in Figure 7, approximated that of noonday sunlight^ below 3700 angstroms. ------- IOO 800 2900 3000 3100 3200 3300 340O 350O 36OO 37OO 3800 390O 40OO WAVE LENGTH, angstrom units Figure 6. Spectral transmission curve for 3-mil Mylar. 1000 CHAMBER LIGHT 70 WHITE LAMPS 70 BLACK LAMPS CAVITY REFLECTOR "MYLAR CUTOFF" 25 % OVERLOAD SUNLIGHT AND SKYLIGHT 3000 3200 3400 3600 3800 40OO WAVE LENGTH, angstrom units Figure 7. Light intensity. ------- 10 IRRADIATION CHAMBER TESTS To minimize surface effects, the chamber was made as large as possible within laboratory space limitations to reduce the surface-to-volume ratio. The interior surface was com- prised of approximately 168 square feet of Mylar and 132 square feet of aluminum; the surface-to-volume ratio for Mylar was 0. 50 and for aluminum, 0. 39. These materials have been found to have minimal effect on the major components present in the exhaust gases or produced in the irradiated gases. -1 The diluted exhaust gases were continuously fed to the dyna- mic chamber through the inlet port located at the rear of the lower recirculating fan. Simultaneously the irradiated exhaust gases were discharged through a distribution system to the plant, bacteria, and animal exposure facilities and to a manifold for chemical sampling and analysis. EXPOSURE FACILITIES Plant Exposures — Effluent from the irradiation chamber was piped to a lighted plant exposure chamber about 16 cubic feet in volume (Figure 8). The chamber was equipped with a system of fluorescent lamps to maintain the sensitivity of the plants. The light level was about 1000 foot-candles, high enough to keep the stomates open. Potted plants grown under greenhouse con- ditions were exposed. For most exposures two varieties of tobacco, pinto bean at three stages of growth, and petunia •were used. Bacteria Exposures — A rotary impactor technique was used to evaluate the effect-of irradiated exhaust on bacteria (Figure 9). In this technique bacteria cultures of E. coli "were plated in logarithmically increasing concentrations on a membrane filter strip, which was attached to the outer surface of a 1-1/2 inch- diameter metal cylinder. As the cylinder rotated at a constant angular velocity, a thin line of irradiated exhaust gases im- pinged on the surface of the membrane. After exposure, the membranes were incubated, dried, and stained, and the degree of kill was indicated by contrast between the color of the stained living colonies and the lack of color of those areas where the bacteria were killed. Animal Exposures -- During a few runs investigations were made of the effects of short-term exposure of animals to syn- thetically produced smog. Rats and mice were exposed in •whole-body chambers, and spontaneous running activity was evaluated. Guinea pigs were exposed through face masks; measurements included tidal volume, respiratory rate, and pulmonary flow resistance. Animals were sacrificed for various biochemical measurements. Most of the animal ex- posures were made during special chamber runs and are not a part of this test series. ------- Figure 8. Plant exposure chamber. ------- IRRADIATION CHAMBER TESTS Figure 9. Bacteria exposure equipment. ANALYTICAL PROCEDURES Concentrations of chemical constituents were monitored in (1) the raw exhaust from the automobile, (2) the exhaust-gas mixture after dilution, and (3) the contents of the irradiation chamber before and during irradiation. Carbon monoxide, car- bon dioxide, and gross hydrocarbons (as hexane)were measured in the raw exhaust gas by means of nondispersive infrared analyzers, in which a raw exhaust sample was drawn through an ice-bath condenser to remove interfering water vapor. The total hydrocarbon concentration in both the nonirradiated and irradiated exhaust gas mixture after dilution was measured with a flame-ionization-type detector, which responds in ppm total carbon atoms. NO and NO2 were measured by a continu- ous-recording colorimetric instrument that uses a modified ------- Parameters 13 Saltzman reagent. Because long response times are inherent in the instrument, a computer program was used to convert the instrument response to instantaneous values for NO and NO£. Details of this procedure are given in Appendix B. Carbon monoxide concentrations in the irradiated gases were measured by means of a long-path nondispersive infrared analyzer. Oxi- dant concentrations were measured by a continuous-recording coulometric instrument (Mast) that uses a neutral potassium iodide solution. Gas chromatography was used for detailed analyses of hydrocarbons containing 2 to 5 carbon atoms.'* Wet-chemical analytical methods were employed for measure- ment of olefin, formaldehyde, and acrolein, ^ and for confirma- tory measurements for NO and NO2- ^ A few mass-spectro- graphic and infrared analyses were also made. TEST PARAMETERS Three variables that affect the photochemical reaction sys- tem were studied in this test series. EXHAUST CONCENTRATION Effects of irradiation were evaluated under three conditions of concentration: 1. 'Atmospheric' hydrocarbon levels of approximately 11-15 ppm carbon. These concentrations represent ambient air levels during severe air pollution situations. 2. Concentrated hydrocarbon levels of approximately 32- 40 ppm carbon. 3. 'Equivalent1 atmospheric concentration levels achieved by diluting the chamber contents produced during irra- diation at concentrated levels. Variations in chemical reaction were evaluated for conditions 1 and 2. Plant exposures were made and evaluated at conditions 1 and 3. Effects on bacteria were evaluated at all three condi- tions. ------- 14 IRRADIATION CHAMBER TESTS AVERAGE IRRADIATION TIME Average irradiation times of 85 and 120 minutes were selec- ted to represent equilibrium conditions in the chamber after an extended period of irradiation. From the relationship previously discussed (Equation 1), average irradiation periods of 85 and 120 minutes at equilibrium (t infinite) require flow volumes of 3. 9 and 2. 8 cubic feet per minute respectively. At these flow volumes, for irradiation periods (i. e. time after irradiation starts) of 1.5, 2. 0, and 2. 5 times the average irradiation time selected, the actual average irradiation times are 78, 86, and 92 percent of the selected average irradiation time, respec- tively. FUEL CONTENT Previous work indicated that the olefin content of fuels could possibly influence the air pollution potential of irradiated exhaust gases. For this reason two fuels were chosen for testing in this series, one containing 14% olefins and one containing 23% olefins. These fuels were obtained from special stocks pre- pared by the Western Oil and Gas Association for automotive exhaust research. Both fuels had approximately the same aro- matic content: the low-olefin fuel, fuel 3, gave a bromine num- ber of 30, and the higher-olefin fuel, fuel 5, gave a bromine number of 50. The physical specifications of the fuels were held approximately equal: both had research octane numbers of 100; ASTM distillation ranges varied by not more than 5% at any point; Reid vapor pressures were 9. 3 and 9. 4; and API gravities were 56. 0 and 56. 5. Physical and chemical proper- ties of the fuels are shown in Table 1. OTHER TEST CONDITIONS Irradiation intensity in the region from 2900 to 3600 ang- stroms was held at levels believed to approximate summer noon- day sunlight in Southern California. All tests were conducted at a constant volumetric ratio of hydrocarbon to oxides of nitrogen of approximately 12:1 (hydrocarbon expressed as ppm carbon). The reaction system used in this test series was dynamic in that the introduction of the dilute nonirradiated exhaust gases ------- Parameters 15 Table 1. PHYSICAL AND CHEMICAL, PROPERTIES OF TEST FUELS Properties API gravity, degrees Reid vapor pressure, Ib/in. ** Distillation, °F Initial End point Recovery, volume % Sulfur (total), weight % Sulfur (RSH), ppm Gum (existent), mg/lOOml Gum (accelerated, 4-hr), mg/lOOml Bromine no. (electrometric), g/100 g Tetraethyl lead, ml/ gal Fluorescent indicator analysis (as received), volume % Saturates Olefins Aromatic s Octane no. Motor, F-2 Research, F-l Fuel No. 3 56.0 9.3 92 406 98 0. 013 0. 0002 1. 1 1.4 31 0.53 47. 1 13.9 39.0 88.9 100. 1 Fuel No. 5 56.5 9.4 95 400 97.5 0.042 0.0004 2.3 3.2 49 1.38 38.8 22.9 38.3 87.8 100. 1 and the withdrawal of an equal quantity of irradiated exhaust gases was continuous. In contrast, previous studies have utilized a static system, °> ' in which a dilute mixture of exhaust gases was charged to the chamber, the gases were irradiated, and the effects were evaluated after a selected irradiation period. The two techniques generally produce the same series of reactions. The chief difference is that in the dynamic system the incoming nonirradiated gases and the irradiated gases present in the chamber approach kinetic chemical equilibrium, whereas in the static system the chemical reaction of the single charge is allowed to approach completion. The extent of mixing of community atmospheres is both complex and extremely variable. On the average it is probably ------- !& IRRADIATION CHAMBER TESTS greater than that which occurs in a static system and less than O the continuous, uniform mixing produced in the dynamic irradia- tion chamber. Probably specific cases occur in which the mixing actions of both static and dynamic systems are paralleled in the atmosphere. Because of the intricate movements of parcels of air, it is possible that mixing conditions much like those produced in both systems could occur simultaneously at different locations within a community. These two test condi- tions represent extremes of mixing, and results obtained in both series of chamber studies probably span the average mixing conditions of the atmosphere. CHEMISTRY OF IRRADIATED EXHAUSTS THE NO NO2 REACTION PROCESS The oxidation of NO to N©2 is one of the significant systems in the photochemical reaction complex. Studies to date indicate that this complex reaction system consists of two fundamental reaction series: the NO NO-, photo-oxidation reactions and the NO2 — free-radical reactions. NO NO2 PHOTO-OXIDATION REACTIONS Nitric oxide oxidizes to nitrogen dioxide in the presence of organic compounds under ultraviolet radiation at a rate far high- er than the thermal rate of oxidation of nitric oxide by molecular oxygen. This rapid conversion of nitric oxide to nitrogen di- oxide appears to require a chain reaction involving peroxyalkyl (RO2) and peroxyacyl (RCOj) free radicals. These radicals apparently are generated extremely rapidly by initial steps in- volving atomic oxygen attack on hydrocarbon species or photoly- sis of nitrogen dioxide. Investigations of the photochemistry of model systems have shown that many types of organic compounds can participate in the photo -oxidation of nitric oxide to nitrogen dioxide and the for- mation of ozone in the photochemical reaction complex. Olefins, many aromatic hydrocarbons, some higher-molecular-weight paraffins, and aldehydes already have been shown to participate. Lower-molecular-weight paraffins (including methane, ethane, propane, and the butanes), acetylene, and benzenes do not ap- pear to participate to any marked degree in these reactions. NO2 - FREE-RADICAL REACTIONS The free-radical species that participate in the photo-oxida- tion of nitric oxide also may react with nitrogen dioxide to form ------- Chemistry 17 organic nitrogen compounds as the nitrogen dioxide concentration increases during the irradiation. Alkyl nitrates and peroxyacyl nitrates have been identified and measured as products in static chamber experiments. Under many of the experimental conditions used in this study a substantial depletion of nitrogen dioxide was observed after the nitrogen dioxide reached its maximum concentration. A limited number of infrared measurements also indicated the presence of alkyl nitrates and peroxyacyl nitrates. The presence of these substances accounts for at least part of the decrease in the ni- trogen dioxide during the irradiations. Under dynamic irradiation conditions, photochemical reac- tion begins when the irradiation is started (Figure 10, Table Al). -120 -90 -60 -30 0 30 60 90 120 150 180 210 240 270 300 330 TIME, minutes Figure 10. Chemical relationships for a typical chamber reaction. It is postulated, as discussed above, that initially the NO-NO2 photo-oxidation reactions predominate. This results in a rapid decrease in NO concentration, with a correspondingly rapid in- crease in NO2 concentration. Some of the hydrocarbons present also decrease significantly in this stage of the reaction. This reaction continues, depending on the conditions of irradiation, for 30 to 120 minutes until the NO2 concentration reaches a maxi- mum and the NO concentration approaches a minimum. At about ------- 18 IRRADIATION CHAMBER TESTS this point the NO2 free-radical reactions become important, and the NO2 concentration drops as the NO2 reacts with inter- mediate exhaust reaction products. This condition continues un- til equilibrium is approached between the NO NO2 photo-oxida- tion reactions and the NO2 free-radical reaction. An increase in the level of oxidant and other reaction products and a further decrease in hydrocarbon concentration accompany the NO2 free-radical reactions. As similarly observed in the static ir- radiation systems, ozone appears only when the NO2 reaches peak concentration and the NO approaches its minimum equili- brium concentration. CHEMICAL EFFECTS NO; The rate of formation of NO2 is an important index for the characterization of irradiated atmospheres because it is a mea- sure of reactivity of the over-all system, readily obtained with available monitoring instrumentation. Table 2 lists the average and individual NO2 formation rates observed for all combina- tions of the test variables. These data are pictured in Figure 11. It is apparent that within the scope of the study, initial ex- haust concentration had the greatest effect on NO2 formation rate. Fuel composition had a lesser effect; average irradiation time (AIT) had little or no effect. Table 2. NO2 FORMATION AND DISAPPEARANCE Test conditions Exhaust level Atm Atm Atm Atm Cone Cone Cone Cone Fuel No. 3 3 5 5 3 3 5 5 AIT, min 85 120 85 120 85 120 85 120 Average actual exhaust gas cone, ppm C 14.3 12.2 14.8 14.4 33.0 36.0 39.0 35.5 NO2 Formation rate, pphm/min Individual runs 1.76, 1.25, 1.50 1.45,1. 67 1.88,2.22 2. 10, 1.95 2.42,2.90 2. 61,2. 36 3. 59,3.80 2.68,2.74 Avg 1.5 1.5 2. 1 2.0 2.6 2.5 3.7 2.7 Average NO2 disappearing, % 41.5 60.5 60.0 63.0 27. 5 35.5 39.0 26.5 Comparing all runs made at atmospheric hydrocarbon levels with those at concentrated levels shows a mean increase in aver- age NO2 formation rate from 1. 8 (range 1. 25-2. 22) to 2. 9 pphm/ minute (range 2. 36-3. 80). Changing from a low (14%) to a high (23%) olefin fuel caused an average increase from 2. 0 (range 1. 25-2. 9) to 2. 6 pphm/minute (range 1. 88-3. 80). Increasing average irradiation time from 85 to 120 minutes caused a ------- Chemical Effects 19 SHADED PLANE REPRESENTS 120-MINUTE AVERAGE IRRADIATION TIME UNSHADED PLANE REPRESENTS 85-MINUTE AVERAGE IRRADIATION TIME. FUEL 5 (23 % OLEFIN) FUEL 3 (14% OLEFIN) 10 20 30 40 50 HYDROCARBON CONCENTRATION LEVEL, ppmC Figure I I. NO2 formation rate as influenced by hydrocarbon concentration, fuel type, and irradiation time. decrease from 2.5 (range 1. Z5-3.80) to 2.2 (range 1. 45-2.74). This decrease was the result of unusual test conditions; as will be shown later essentially no change in NC>2 formation rate is attributable to changes in average irradiation time. The aver- age differences caused by changes in concentration and fuel composition are both significant at the 1% level determined by an analysis of variance. As shown in Figure 11, the effects of increasing initial hydrocarbon concentration and changing from fuel 3 to fuel 5 appear to be additive. This suggests that the initial olefin con- centration may be the most important factor in determining the NC>2 formation rate, since both concentration and fuel influence olefin content. Table A2 gives the average concentration of four-carbon and higher olefins at the start of irradiation for each test situation. These data are plotted in Figure 12 against NC>2 formation rate and show a good relationship between olefin concentration at the start of irradiation and NC>2 formation rate. Consideration of the importance of olefin concentration in the initial exhaust mixture also helps explain the apparent de- crease in NC>2 formation rate as the average irradiation time (AIT) increased from 85 to 1ZO minutes in the concentrated series with fuel 5 (Figure 11). Although the total hydrocarbon concentrations were similar, the exhaust mixture for the runs at 85 minutes AIT contained an average of 6. 57 |J.g/l of C^.+ olefins and that used in the runs at 120 AITaveraged 3. 82 p.g/1. ------- 20 IRRADIATION CHAMBER TESTS 8.0 o +u* z O O Z O O 7.0 — 6.0 5.0 4.0 3.0 t 2.0 1.0 o 120 MIN AIT • 85 MIN AIT 0 1.0 2.0 3.0 4.0 5.0 6.0 NO2 formation rate, pphm /min. Figure 12. Relationship of olefin concentration to NO2 formation rate. The apparent difference in NO£ formation rate (3. 7 vs Z. 7 pphm/min) is probably not due to changes in AIT but to initial olefin concentration. Adjustment of these points to similar olefin concentrations brings them into line and leads to the con- clusion that AIT has no influence on NO? formation rate. The experimental conditions that led to the low olefin levels for the 120-minute AIT cannot be explained. ------- Chemical Effects 21 Analysis of the NOo free-radical reaction data does not allow conclusions as detailed as for the NC>2 formation rate. The rate of NC>2 disappearance cannot be established accurately, since limitations in the instrumentation prevented the precise determination of the time of peak NC>2 concentration. Therefore the percentage of NO2 disappearing (defined as the percent de- crease in NC>2 concentration from peak to equilibrium) has been used as a measure of the NC>2 free-radical series of reactions described earlier. Analysis of the data in Table 2 indicates that NC>2 disappearance is greater for the atmospheric series of tests than for the concentrated series. Average values drop from 56% disappearance for the atmospheric series to 32% for the concentrated series. This difference is significant at the 1% level. The differences caused by fuel composition (41 and 47%) and average irradiation time (42 and 46%) are not statisti- cally significant. OXIDANT One of the principal reactions associated with the NO to NO2 photo-oxidation is the formation of oxidant (Figure 10). The coulometric instrument used to monitor ozone is sensitive to NO2 and other oxidants as well as to ozone. To secure a better indication of ozone levels, corrections were made for the principal interference, NO2- The corrected values for oxidant, shown in Figure 13, are primarily ozone, although they include, to a very limited degree, the response of other oxidants. Two conclusions are apparent from the oxidant data: (1) av- erage irradiation time has no influence on oxidant production and (2) the higher initial hydrocarbon concentration level inhibits the production of oxidant for as much as 1/2 hour at the begin- ning of the irradiation. Although variations in oxidant level were observed with changes in exhaust concentration and fuel composi- tion, these followed no consistent pattern and provide no basis for general conclusions. By contrast, reports of static irradia- tion chamber tests" indicate that ozone formation increased with exhaust concentration as a function of the one-half power of the increase in hydrocarbon level. Additional work is needed to re- solve the apparent difference between the results of static and dynamic tests. HYDROCARBONS The disappearance of hydrocarbons during irradiation is in accord with the NO2 free-radical reactions hypothesis, i. e. the involvement of NO2 with intermediate hydrocarbon reaction products. The disappearance of hydrocarbons also could be the result of reactions involving ozone and oxygen atoms to form oxygenated products. In all tests the reduction in organics as ------- 22 IRRADIATION CHAMBER TESTS 60 CURVE IRRADIATION TIME 85 120 85 120 85 120 85 0 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 TIME, minutes Figure 13. Oxidant concentrations corrected forNO2 response. measured by the flame-ionization analyzer was not appreciable until the NC>2 concentration peaked (Appendix A, Figures Al through A8). At this point the reduction increased sharply and continued until chamber equilibrium was reached. The over-all reduction ranged from 15 to 20% for all test conditions. A consistent difference was noted in the reactivity of the various classes of hydrocarbons as measured by the percent decrease in concentration of these individual species. These reactivities were not influenced by changes in exhaust concen- tration, fuel composition, or irradiation time, however. Olefins were studied by both wet chemistry and gas chroma- tography. Four-carbon and higher molecular weight olefins were measured by a wet chemical procedure, the p-dimethyl- aminobenzaldehyde method of Altshuller and Sleva. ^ Aroma- tics and two- through five-carbon paraffins and olefins, acety- lene and methyl acetylene were analyzed by gas chromato- graphy.4' H Measurements made for acetylene indicated no significant reaction rate for this compound, but methyl acetylene appeared to be slightly reactive. Tables AZ A5 give data on concentra- ------- Chemical Effects 23 tion of four- and five-carbon paraffins for the various test con- ditions. Table 3 summarizes losses of these paraffins during irradiation. It appears that on the average four- and five-carbon paraffins did not react at significant rates. Table 3. PERCENT DECREASE FOR SPECIFIC HYDROCARBONS Compound Di olefin Butadiene Terminally bonded olefine Butene-1, isobutene Pentene-1 Z-Methylbutene Avg Internally bonded olefins c-Butene-2 t-Butene-Z c-Pentene-2 t-Pentene-2 2-Methylbutene-2 Avg Paraffins Isobutane n-Butane Isopentene n-Pentene Avg Fuel No. 3 85 Atm 68 62 47 71 62 100a 100a 90 90 98 96 10 14 13 12 12 Min Cone 62 60 55 67 61 100a 100a 86 81 87 85 4 5 6 8 6 120 Minb Atm 73 75 67 77 73 100a 100* 100a 100a 100a 100a 13 3 10 10 9 Fuel No. 5 85 Atm 67 52 36 75 58 100a 100a 100a 100a 100a 100a -4 3 -3 -3 -2 Min Cone 67 71 57 88 71 100a 100* 96 95 93 95 4 -13 -5 -3 -4 120 Min Atm 75 67 63 88 73 100a 100a 100a 100a 100a 100a 0 2 5 5 3 Cone 65 68 60 68 65 100a 100a 100a 100a 100a 100a -5 -2 0 6 0 aValues after irradiation are below the limits of detection; although percent decrease is approximate, it usually approaches 100 percent. bNo values are given for concentrated levels, since data were inadequate for analysis. Olefins — Although all olefins measured showed a significant decrease in concentration during the test runs, the decrease was not influenced by the test variables. The relative concen- tration of the individual olefins at the start of irradiation was dependent on fuel composition and did vary to some extent with initial exhaust concentration. These data'are shown in Tables A2 A5. Individual four- and five-carbon olefins reacted at rates that were dependent upon the location of the double bond (Table 3). The terminal olefins showed an average decrease of approxi- mately 65% after irradiation periods of 170 minutes for the 85- minute AIT series and 240 minutes for the IZO-minute AIT series. At the end of these same irradiation periods, all internally bonded olefins showed almost complete reaction. The final concentrations for internally bonded olefins were usually below the limits of detection of the analytical methods. Bu- tadiene-1, 3 behaved similarly to the terminally bonded olefins. Decreases of 12% and 50% were found for ethylene and propylene, respectively, in a few gas-chromatographic measurements ------- 24 IRRADIATION CHAMBER TESTS of these compounds in the atmospheric series. Static irradiation chamber tests9' 12 also show a high degree of reactivity for terminally bonded olefins and essentially complete disappearance of internally bonded olefins. Aromatics — Aromatics were studied in a few runs by gas chromatography. i1 The chromatographic column used did not provide resolution for complete separation of benzene from other reaction products. Work with the mass spectrometer, 13 con- firmed by pure component studies, indicated that benzene is non- reactive. All other aromatics studied showed some degree of reactivity after an irradiation period equal to two times the av- erage irradiation time. The results of three runs, all made with fuel 3 in the concentrated series, are shown in Table 4. These few tests indicated an average decrease of about 10% for toluene and ethylbenzene, about 30% for xylene and the 3- and 4-ethyltoluenes, and about 50% for styrene. Table 4. PERCENTAGE CHANGE IN AROMATIC HYDROCARBONSa AITb 120 120 85 Avg Toluene 5 -13 6 8 Ethyl - Benzene 0 -17 -16 -11 m- and p- Xylene -21 -28 -27 -25 o-Xylene -29 -29 -27 -28 Styrene -40 -70 c -55 3- and 4- Ethyltoluene -21 -39 c -30 aPercent change after irradiation periods of 170 and 240 minutes for 85 and 120 minute AIT, respectively. ^Average irradiation time, minutes. GConcentrations of these components too low to determine changes in composi- tion. These results indicate that the nature of substitution on the benzene ring controls the reactivity rate. More detailed evalua- tion of this possibility is not warranted by the limited experimen- tal data available. FORMALDEHYDE Formation of formaldehyde in the irradiation system de- pended primarily on initial exhaust concentration, as shown in Figure 14. Increasing initial exhaust concentration by 2 to 2-1/2 times generally resulted in a proportional increase in formaldehyde after 240 minutes of irradiation, although the ini- tial formation rates appear about the same for each exhaust con- centration. These average data also indicate that fuel composi- tion may influence slightly the production of formaldehyde. Average irradiation time did not appear to influence formalde- hyde production. GPO 8O6—304-3 ------- Chemical Effects 25 CURVE I IRRADIATION I FUEL I CONG. I 3 CONC 3 CONC 5 CONC 5 CONC 20 40 60 80 100 120 140 160 180 200 220 240 260 280 TIME, minutes Figure 14. Formaldehyde concentrations. OTHER ALDEHYDES Aldehydes other than formaldehyde were measured in a few runs and were found both in the nonirradiated and the irradi- ated exhaust mixtures. Irradiation brought about a five-fold in- crease in their concentration; that is, only about 20% of the aldehyde came directly from the auto exhaust. Acrolein, analyzed by the method of Cohen and Altshuller, 5 behaved gen- erally like formaldehyde; the concentration was initially low, increased during irradiation, and reached a maximum after approximately two average irradiation times. In several analy- ses a 25 to 30% decrease in acrolein concentrations, noted after the maximum was reached, indicated further reaction of the acrolein. Aldehydes showed the same general relationships to initial exhaust concentration and fuel composition as the NC>2 formation rate — indicating a. dependency on initial olefin concentration. This is not unexpected, since it is the attack of ozone and atomic oxygen on the olefins and aromatics that produces aldehydes. In this study, however, the aromatic content of the fuels was held ------- 26 IRRADIATION CHAMBER TESTS constant. Increasing the concentration level of exhaust in the chamber and changing from fuel 3 to fuel 5 both resulted in an increase in the aldehydes produced. The reaction that produces acrolein, primarily from the butadiene-1, 3 is typical of the aldehyde-producing reactions: CH2 = CH-CH=CH2 + (0)—*CH2 = CH-CHO + CH2O + other products. In two instances total aldehydes were also determined by the 3-methyl-Z-benzothiazolone hydrazone test described by Sawicki et al. -^4 Table 5 shows the relationship between formaldehyde^ and total aldehydes for the tests with fuel 3 at concentrated exhaust levels and 85-minute AIT. Approximately 60% of the total aldehydes were formaldehyde in these runs. Table 5. ALIPHATIC ALDEHYDE CONCENTRATION21 (ppm) Run Total aldehydesb Formaldehyde0 % Formaldehyde 1 2. 1. 2 1. 1. 25 90 61 95 1. 11 1.28 0.96 1.02 49 67 60 52 Average 57 Concentration after equilibrium is attained. bTotal aldehydes corrected for formaldehyde; absorptivity of 40 is used for all aldehydes other than formaldehyde. °As determined by the chromotropic acid method. ^ BIOLOGICAL EFFECTS PLANTS Plants were exposed to two types of irradiated gases: (1) directly to chamber gases irradiated at the atmospheric exhaust levels, and (2) to chamber gases irradiated at the high exhaust concentration, but diluted to atmospheric levels before plant exposure. Exposures started within 15 minutes after the cham- ber lights were turned on, and usually continued for 4 hours. Following exposure, plants were returned to the greenhouse and held for observation over a period of time. ------- Biological Effects 27 Symptoms of severe injury were sometimes noticeable within a few hours. Some injury patterns were not visible until 48 hours after exposure. The patterns of injury, including the microscopic patterns observed with thin sections, varied with species and age of tissue. Prior- and post-exposure culture conditions had some effect on injury pattern. Broad classes of injury patterns are thought to relate to specific phytotoxicant classes. At this time, however, tech- niques of exposure and classification of injury patterns have not been developed sufficiently for formulation of a multiple- class injury scale. A number of general conclusions are possible concerning the effect of the three variables on plant damage: (1) fuel 5 produced a greater total injury than fuel 3 and also a different distribution of type of injury, and (Z) higher exhaust concentrations at time of irradiation also produced a different distribution of type of injury than the atmospheric series, but tended to decrease the total in- jury. (It should be noted that the higher exhaust concentrations were diluted to atmospheric levels after irradiation and before plant exposure. ) For reporting gross injury a scale of 0 (no in- jury) to 4 (total injury of sensitive tissue) was used. The av- erage index for all plants is shown in Table 6 for each fuel at each irradiation time and concentration. Table 6. AVERAGE GROSS PLANT INJURY3- Irradiation Irradiation concentration, Fuel time, min PPm C 12-15 85 32-40c 3 „ 12-15 120 32-40C 12-15 32-40c 5 „ 12-15 120 32-40C Plant injury 1. 0 0.4 0.5 0.2 2. 1 1.8 2. 2 1.0 aAverage of all plants exposed at each condition. t>Scale: 0 (no injury) to 4 (total injury of sensitive tissue). cPlants exposed after dilution of exhaust to equivalent atmospheric carbon levels. ------- 28 IRRADIATION CHAMBER TESTS ANIMALS Although animal exposures were made during only a few of the irradiation chamber runs discussed in this report, the facil- ities were used in special tests designed specifically for animal studies. The conditions of these tests did not necessarily con- form to those used for the chemical, plant, and bacteria studies discussed in this report and are not a part of this test series. These special investigations with small animals, which were generally exploratory studies, are summarized in detail in a report prepared by the Laboratory of Medical and Biological Sciences, Division of Air Pollution. 16 BACTERIA The effects of the irradiated gases on bacteria (E. coli) are shown in Table 7. Table 7. EFFECT OF CONCENTRATION ON BACTERIA KILL Fuel No. 3 5 Irradiation time, min 85 120 85 120 Irradiation level Atm Cone Atm Cone Atm Cone Atm Cone Bacteria kill, relative unitsa Exposure at irrad. level0 2.0 9.0 4.0 6.5 1.5 9.0 2.5 8.0 Exposure at atm levelc 2.0 1.5b 4.0 1.0b 1. 5 2. Ob 2. 5 2.0b aComplete kill equals 9; exponential scale. ^Exposure made at the concentration at which irradiation occurred. cExposure made after dilution of the high-concentration exhaust to equiva- lent atmospheric concentration levels. The values for bacteria kill shown in Table 7 represent de- gree of bacteria kill (rated on a scale of 1 through 9). They were determined by visually estimating the area and distribution of prepared culture strip affected by exposure to the irradiated exhaust. Since the bacteria were plated on the test strip in con- centrations that increased logarithmically, the degree of kill given in Table 7 represents an exponential rather than a linear kill scale. Degree of bacteria kill was influenced only by exhaust con- centration. No effect was observed for changes in fuel composi- tion or average irradiation time. No difference in bacteria kill was noted between exposure to gases irradiated at atmospheric ------- Summary of Results 29 levels and gases irradiated at concentrated levels and diluted to atmospheric levels. SUMMARY OF RESULTS Within the limits of experimental values selected for this series of tests, the three major variables examined, (1) initial exhaust composition, (2) fuel composition and (3) average irradi- ation time can be ranked according to their influence on the chemistry of the irradiation process, and biological effects of reaction products. Concentration of exhaust at the start of ir- radiation produced the greatest effect on the greatest number of test parameters, including NC>2 formation rate, oxidant and for- maldehyde production, plant damage, and bacterial kill. Chang- ing fuel from number 3 (14% olefins) to number 5 (23% olefins) also affected plant damage, NO2 formation rate, and possibly formaldehyde production. Changing average irradiation time from 85 to 120 minutes had essentially no effect on any of the test parameters. More detailed conclusions are appropriate for each of the test parameters and are given below: 1. The NC>2 formation rate was increased significantly by increases in both exhaust concentration and olefin con- tent of the fuel. Average NO2 formation rate increased from 1.8 to 2.9 pphm/min as the average initial exhaust concentration increased from atmospheric levels (14 ppm C) to concentrated levels (36 ppm C). An average increase from 2. 0 to 2. 6 pphm/min was observed in going from fuel number 3 to fuel number 5. Increasing average irradiation time from 85 to 120 minutes had no effect on NC>2 formation rate. A good relationship was observed between NO2 formation rate and initial olefin concentration in the chamber. 2. The percentage of disappearance of NC>2 was decreased significantly by increasing initial exhaust concentration. Fuel composition and average irradiation time had no effect. 3. Increasing exhaust concentration appeared to cause a delay in initial oxidant production of approximately 30 minutes. The amount of oxidant (mostly ozone) pro- duced was influenced by exhaust concentration and fuel type, but no consistent pattern was apparent. Average irradiation time had no apparent effect on oxidant pro- duction. ------- 30 IRRADIATION CHAMBER TESTS 4. Reduction in organic concentration as measured by the flame-ionization analyzer ranged from approximately 15 to 25%. The percentage decrease in the terminal olefin varied between 35% and 88%, depending on the specific hydrocarbon, and averaged about 65% decrease in concentration with irradiation. All internally bonded olefins reacted almost completely to form products. Paraffins and acetylene were essentially unreactive. The aromatic hydrocarbons disappeared through reac- tion by amounts ranging from about 10% for toluene to about 50% for styrene and appeared to depend on type of substitution on the benzene ring. No consistent patterns of influence of exhaust concentration, fuel type, or av- erage irradiation time were observed for any group of hydrocarbons. 5. Increasing initial exhaust concentration from 14 ppm carbon to 36 ppm carbon resulted in proportional in- creases in average formaldehyde concentrations of from 50 pphm to 120 pphm at kinetic equilibrium. 6. In a few test runs analyses were made for both total aldehydes and formaldehydes. Approximately 60% of the total aldehydes appeared to be formaldehydes. 7. Most of the aldehyde present came from chamber reac- tions. (About 20% originated in the auto exhaust. ) Gen- erally aldehydes varied with exhaust concentration and fuel type similarly to the NO£ formation rate — indica- ting a dependency on initial olefin concentration. This was as expected, since the attack of ozone and atomic oxygen on olefins produces most of the aldehyde present. 8. Differences in extent and distribution of type of plant damage were observed with changes in exhaust concen- tration at irradiation and with fuel type. No effects were attributable to changes in average irradiation time. The high-olefin fuel generally gave a higher in- jury index than did the low-olefin fuel. Plants exposed to exhaust irradiated at 36 ppm carbon and diluted to the 14-ppm equivalent prior to plant exposure generally gave a lower injury index than plants exposed to exhaust irradiated at 14 ppm carbon. 9. The average degree of bacterial kill increased from 3 to 8 on an exponential scale as irradiated exhaust con- centration increased from 14 to 36 ppm carbon. When the concentrated irradiated mixture was diluted to at- mospheric levels, the degree of kill decreased to a value of about 2. Fuel composition and average irradi- ation time had no apparent effect on degree of bacterial kill. ------- Summary of Results 31 In addition to the results enumerated above, this first series of dynamic irradiation chamber investigations has demon- strated the usefulness and versatility of the facility, permitted development of reliable experimental and analytical techniques, suggested possibilities for future investigations, and provided the basis for their design. A second series of tests with the dynamic irradiation chamber has been completed, and analysis of data is now under way. These runs were designed to explore in detail the effects of varying ratios of hydrocarbon to oxides of nitrogen. Tests were run with a single fuel (similar to fuel 3), 120-minute average irradiation time, and exhaust con- centrations at atmospheric levels or below. A report is being prepared on this second series. Additional test series -will follow. Investigations of the effect of raw and irradiated exhaust on animals have been greatly expanded and transferred to a separate facility since this first series of tests. The present study in- volves nearly 3000 small animals and continuous exposures over the life period of the animals to varying levels of raw and irradiated exhaust. ------- REFERENCES 1. Teague, D.M., Bishop, W. , Nagler, L.H., Onishi, G. , Sink, M. V. , Stonex, K.A., and VanDerVeer, R.T., "Los Angeles Traffic Pattern Survey, " Presented at the SAE National West Coast Meeting, Seattle, Wash- ington, August 1957. 2. Luckish, Matthew, D. Sc. , D.E., "Applications of Germici- dal, Erythemal, and Infrared Energy, " D. VanNostrand Co., Inc., 1946. 3. Altshuller, A. P. , Wartburg, A.F., "The Interaction of Ozone with Plastic and Metallic Materials in a Dyna- mic Flow System, " Int. 3_. Air and Water Pollution. 4, 70 (1961). ~ 4. Bellar, TV, Sigsby, J. E. , Jr., demons, C.A., and Alt- shuller, A. P. , "Direct Application of Gas Chromato- graphy to Atmospheric Pollutants, " Anal. Chem. , 34, 763 (1962). 5. Cohen, I. R. and Altshuller, A. P. , "A New Spectrophoto- metric Method for the Determination of Acrolein in Combustion Gases and in the Atmosphere, " Anal. Chem. , 33, 736 (1961). 6. Saltzman, B. E. , "Colorimetric Microdetermination of Ni- trogen Dioxide in the Atmosphere, " Anal. Chem. , 26, 1949 (1954). 7. Hamming, W. J. , Mader, P.P., Nicksic, S. W. , Romanov- sky, J.C., and Wayne, L. G. , "Gasoline Composition and the Control of Smog, " Air Pollution Control Dis- trict, County of Los Angeles and Western Oil and Gas Association, September 1961. 8. Stephens, E. R. , "The Reactions of Auto Exhaust in Sun- light, " Scott Research Laboratories, Inc. , and Uni- versity of California, Presented at the Air Pollution Research Conferences on "Atmospheric Reactions of Constituents of Motor Vehicle Exhaust, " Los Angeles, California, December 5, 1961. 9. Schuck, E.A. , Ford, H.W., and Stephens, E. R. , Report No. 26 Air Pollution Foundation, San Marino, Californ- ia, October 1958. 10. Altshuller, A. P. and Sleva, S. F. , "Spectrophotometric De- termination of Olefins, " Anal. Chem. , 33, 1413(1961). 33 ------- 34 IRRADIATION CHAMBER TESTS 11. Altshuller, A. P. and demons, C.A., "Gas Chromato- graphic Analysis of Aromatic Hydrocarbons at Atmos- pheric Concentrations Using Flame lonization Detec- tion, " Anal. Chem. , 34, 466(1962). 12. Neligan, R. E. , "Hydrocarbons in the Los Angeles Atmos- phere, " A_rch. _Environ. Health, 6, 581 (1962). 13. Sigsby, J. E. , Jr. and Eisele, M. L. , "Use of Mass Spec- troscopy in Air Pollution Studies, " Presented at Meeting of the American Chemical Society, Chicago, 111. , Sep- tember 1961. 14. Sawicki, E. , Hauser, T. R. , Stanley, T.W., and Elbert, W.C., "The 3-Methyl-2Benzothiazolone Hydrazone Test, " Anal. Chem. , 33, 93(1961). 15. Altshuller, A. P. , Miller, D.L., and Sleva, S.F., "De- termination of Formaldehyde in Gas Mixtures by the Chromotropic Acid Method, " Anal. Chem. 33, (1961). 16. Murphy, S. J. , Leng, J.K., Ulrich, C.E., and Davis, H. V. , "Effects on Experimental Animals of Brief Ex- posure to Diluted Automobile Exhaust, " Submitted to A.M. A. Arch, of Environmental Health, 1962. ------- APPENDIX A RAW DATA FOR IRRADIATION CHAMBER REACTION PRODUCTS ------- Table Al. CONCENTRATION VERSUS TIME AFTER IRRADIATION STARTED (Mean values from two or more tests) NO (pphrn) NO2 (pphm) HC (ppm) Oxidant (pphm) Olefin (pphm) HCHO (pphm) 3 3 5 Average irrad. 85 1ZO 120 85 1EO 85 120 85 85 120 B5 120 85 120 85 31 85 120 85 120 Atmospheric levels (11-15 ppm carbon) 62.3 79. 0 35. 0 2. 5 2. 0 14. 3 12.2 14. 4 0. 0 0. 0 0. 0 0. 0 16. 1 46. 0 13. 7 15. 0 0. 0 0. 0 27.3 21. 5 4.0 61.5 7 .0 6 .5 1 . 3 1 . 5 0. 5 2.0 2. 0 2. 0 12. 9a 35763- 15. 3a 21. 5 23. 0 32. 0 1.7 0.5 0. 5 52. 3 36. 5 36. 0 13. 2a 12. 1 24. 1 25.0 12. 5 IB. 8 5. 6 14. 9a 31.0a 31. 5a 54. Oa 40. 5a 90 1. 7 0. 5 0. 0 44. 5 38.0 31.5 31.0 12. 3 10.6 11.5* 33.5 31. 5 19. 0 19.5 3.3 10, 8a 39.7 36. 5a 58.5 43. 0 150 1. 3 2.0 0.0 43.0 31.0 29. 0 12.0 10. 2a 1 1. 1 33. 1 33. 5 19. 0 23.3 2.8* 7.3a 43. 3a 44. 5a 46. 5 240 Min 1. 0 0. 5 2.0 0.0 42. 3 28. 5 29. 0 12. 0 10.4 1 1. 3 36. 5 37. 5 26. 0 28.0 2. 7 6.2 46. 7 53. 0 54. Oa Concentrated levels (32-40 ppm carbon) 0 Min 214.5 195.8 216. 1 159- 4 14. 9 14.9 19. 7 18.2 33.0 35.5 0. 0 0. 0 0. 0 0. 0 45. 69. 39. 29. 0 33. 5 40. 0 30 Min 150. 0 159. 2 137. 1 102.7 83. 1 114. 9 101. 6 33. 0 36.0 35. 5 0. 0 0. 0 0.0 0. 0 41. 2* 61. Oa 54. 0 44. 0 50, 0 46. 5 60 Min 50. 8 73. 3 17. 3 30. 6 176. 1 182. 8 200. 1 31.5* 33.5 33.0 0. 0 2. 0 6. 5 3. 5 2S.5 25.5 89. Oa 60. 5 77. 0 68. 5 90 Min 22.3 38.9 16. 1 20. 3 173. 6 210.7 156.1 179,3 29.5 32.9 30.5 9.0 lb. 0 16. 3 20. 6* 20. 2a 97. Oa 68. 5 90. Oa 93.0 150 Min 16.9 16. 1 1 18.8 "151. 3 ' 203.8 135.8 156, 2 27.8 31.1 29. i 21. 3 20.2 29. 8 27. 5 it. 4» — mp- 103. 0 92. 55 112.5 107. 5 240 Min 16.9 20. 6 14.8 17. 1 148. 8 129.1 145. 1 28.0 50.8 29.0 23.4 32.5 32. B 32.8 15.0 12.2 113.5 116.5 137. 0 128. 0 Table A2. HYDROCARBON REACTIONS; CONCENTRATED SERIES - FUEL No. 3 Compound Diolefin Butadiene Terminally bonded olefins Butene-1, isobutene Pentene-1 2-Methylbutene-l Average Internally bonded olefins c-Butene-Z t-Butene-2 c-Pentene-2 t-Pentene-2 2-Methylbutene,-2 Average C4 + olefinsd Paraffins' Isobutane n-Butane leopentane n-Pentane Average 85-Min avg irradiation time Concentration, ppma Beforeb U.047 0.094 0.031 0.049 0. 013 0. 015 0. 049 0.086 0. 101 3.78 0.027 0.222 0.313 0. 107 After0 0.018 0.038 0. 014 0. 016 0. 005>Ne 0. 005>N ~ 0.007 0. 016 0.013 1. 50 0. 026 0.212 0.294 0.098 % Decrease 62 60 55 67 61 > 62 > 67 ~86 81 87 85 60 4 5 6 a 6 120-Min avg irradiation time Concentration, ppma Beforeb 0.026 0.052 0. 014 0.031 0.006 0.008 0. 021 0.042 0.068 3. 64 0.008 0. 087 After1 0. 021 0.045 0. 015 0. 017 0.005>N« 0, 005>N 0.005>N 0. 005>N 0. 009 0.8 0.022 0. 188 0.289 0. 106 % Decrease 19 13 0 45 >13 >37 >76 >88 87 -90* 78 ^Chamber concentration before irradiation. cChamber concentration after two average residence times. ^Olefins determined by wet chemical analysis (6); results in micrograms per liter. ^Values are less than stated value, which is limit of detection. ^Average value is estimated; actual value is not determinable because of limit of detection. ------- Table A3. HYDROCARBON REACTIONS; CONCENTRATED SERIES - FUEL No. 5 Compound Diolefin Butadiene Terminally bonded olefins Pentene- 1 2-Methylbutene- 1 Average Internally bonded olefins C-Butene-2 t-Butene-2 C-Pentene-2 t-Pentene-2 2-Melhylbutene-2 Average C4 * Olefmd Paraffins Isobutane n-Butane Isopentane n-Pentane Average 85 Mm C Before15 0. 072 0. 044 0. 110 0.029 0. 030 0.068 0. 127 0. 145 6.57 ,,g 0. 028 0. 189 0.409 0. 120 Afterc 0.024 0. 019 0.013 0. 005>Ne 0. 005>N ~ 0. 003 ~0. 006 0. 010 2. 37 0. 027 0. 213 0.428 0. 123 on time ppma % Decrease 67 57 88 71 >83 >84 -96 ~95 93 95 64 4 -13 5 3 4 120 Ml C Beforeb 0. 046 0. 045 0.076 0.024 0.024 0.054 0. 101 0. 093 3.82 0. 022 0. 165 0.316 0. 098 n avg irrac oncentration, Afterc 0. 016 0. 018 0. 024 0. 005>Ne 0. 005>N 0. 005>N 0. 005>N 0. 005 >N 0. 68 0.023 0. 169 0. 317 0. 092 la on clme ppma % Decrease 65 68 60 68 65 79 79 91 95 95 ~65f 82 5 2 0 6 0 aAverage values for two or rnoi ^Chamber concentration before dQlefins determined by wet che ^Average value is estimated, actual value adiation. al analysis (6); re t dete ults in micrograms per liter. minable because of limit of detectn Table A4. HYDROCARBON REACTIONS; ATMOSPHERIC SERIES - FUEL No. 5 Compound Diolefin Butadiene Terminally bonded olefins Butene-1, iaobutene Pentene- 1 2-Methylbutene- 1 Average Internally bonded olefins c-Butene-2 t-Butene-2 c-Pentene-2 t-Pentene-2 2-Methylbutene-Z Average C4 t Olefind Paraffins Isobutane n-Butane leopentane n-Pentane Average 85-Min avg irradiation time Concentration, ppma Beforeb 0.030 0.040 0.022 0.073 0.016 0.016 0.045 0.085 0. 151 2.07 |ig 0.023 0. 154 0.232 0.072 Afterc 0.010 0.019 0.014 0. 018 0.001>Ne 0.001>N 0.001>N 0.001>N 0.001>N 0.40 0.024 0. 150 0.240 0.074 % Decrease 67 52 36 75 58 >94 >94 >98. >99 >99.4 ~99f 81 - 4 3 3 3 2 120-Min avg irradiation time Concentration, ppma Beforeb 0.024 0.036 0.019 0.052 0.014 0.014 0.035 0.068 0. 129 2.0ug 0.015 0. 110 0. 194 0.058 Afterc 0.006 0.012 0.007 0.006 0.001>N 0.001>N 0. OOJ>N 0.001>N 0. 001>N 0.26 0.015 0. 108 0. 184 0.055 % Decrease 75 67 63 88 73 > 93 > 93 > 97 > 99 > 99 ~99£ 87 0 2 5 5 3 aAverage values for two or more runs. ^Chamber concentration before irradiation. ^Chamber concentration after two average residence times. ^OlefinB determined by wet chemical analysis (6); results in micros eValuee are less than stated value, which is limit of detection. ^Average value is estimated; actual value is not determinable becau rams per liter. ie of limit of detection. ------- Table A5. HYDROCARBON REACTIONS; ATMOSPHERIC SEMES - FUEL No. 3 Compound Diolefin Butadiene Terminally bonded olefms Butene- 1 , isobutylene Pentene-1 2-Methylbutene-l Average Internally bonded olefins c-Butene-Z t-Butene-2 c-Pentene-2 t-Pentene-2 2-Methylbutene-2 Average C4 + Olefinsd Paraffins Isobutane n- Butane Isopentane n-Pentane Average Co Beforeb 0.025 0.042 0.017 0. 042 0. 006 0. 007 0.030 0. 050 0. 107 1.81 ^g 0. 020 0. 154 0. 191 0.066 avg irradiat ncentration, p Afterc 0. 008 0. 016 0.009 0. 012 0. 001>Ne 0. 001>N 0. 003 0. 005 0.002 0.40 0. 18 0. 132 0. 167 0.058 pma % Decrease 68 62 47 71 62 >83 >86 90 90 98 96 78 10 14 13 12 12 Co Before13 0.022 0. 040 0. 012 0. Olb 0.006 0. 007 0.019 0. 039 0. 078 1. 86 fig 0.015 0. 115 0. 143 0.051 ncentration, Afterc 0.006 0.010 0.004 0.006 0. 001>Ne 0. 001>N 0. 001>N 0.001>N 0.001>N 0.41 0. 013 0. Ill 0. 128 0. 046 ppma % Decrease 73 75 67 77 73 >83 >86 J95 >97 >99 >97f 78 13 3 10 10 9 stated mated; •mical analysis (6); results in micrograr lue, which ie limit of detection. :tual value is not deterrninable because < if limit of detection. ------- AVERAGE DATA FROM TWO OR MORE RUNS 0 30 60 90 120 150 180 210 240 270 TIME, minutes Figure Al. Test conditions: atmospheric level, 85-minute AIT, fuel no. 3 ------- a. a. Z o t— at. O Z o o a. o. O t— of O O o AVERAGE DATA FROM TWO OR MORE RUNS 10 0 30 60 90 120 150 180 210 240 270 TIME, minutes Figure A2. Test conditions: atmospheric level, 120-minute AIT, fuel no 3 GPO 806—304-4 ------- AVERAGE DATA FROM TWO OR MORE RUNS 0 30 60 90 120 150 180 210 240 270 TIME, minutes Figure A3. Test conditions: atmospheric level, 85-minute AIT, fuel no. 5 ------- 20 AVERAGE DATA FROM TWO OR MORE RUNS 0 30 60 90 120 150 180 210 240 270 TIME, minutes Figure A4. Test conditions: atmospheric level, 120-minute AIT, fuel no. 5 ------- AVERAGE DATA FROM TWO OR MORE RUNS 0 30 60 90 120 150 180 210 240 270 TIME, minutes Figure A5. Test conditions: concentrated level, 85-minute AIT, fuel no. 3 ------- AVERAGE DATA FROM TWO OR MORE RUNS 0 30 60 90 120 150 180 210 240 270 TIME, minutes Figure A6. Test conditions: concentratedlevel, 120-minute AIT, fuel no. 3 ------- AVERAGE DATA FROM TWO OR MORE RUNS 0 30 60 90 120 150 180 210 240 270 TIME, minutes Figure A7. Test conditions:<;oncentr«ted level. 85-minute AIT, fuel no. 5 ------- AVERAGE DATA FROM TWO OR MORE RUNS 30 60 90 120 150 ISO 210 240 270 TIME, minutes Figure A8. Test conditionsrconcentratedlevel, 120-minute AIT, fuel no 5 ------- APPENDIX B COMPUTER PROGRAM FOR REDUCTION OF OXIDES OF NITROGEN DATA ------- COMPUTER PROGRAM FOR REDUCTION OF OXIDES OF NITROGEN DATA The Borman colorimetric instrument used to measure nitric oxide and nitrogen dioxide has a very slow time response. The concentration indicated at any particular time is a value averaged over a considerable period of time, since the deadtime is 4 minutes and the time constant ranges from 9 to 22 minutes over the concentration range of interest. This large time constant produces extreme "tailing" or "lagging" of the indicated concen- tration behind the actual concentration of the gas being sampled. For adjusting the observed concentration to true or instantaneous concentrations at the designated times, a computational pro- cedure was developed for the IBM 650 computer. In general the operation of instruments and other dynamic systems may be described as: (Output Function) (Input Function) (System Function) If the system function for a particular instrument can be defined, it is possible to compute the input function from the output function. The system function for the Borman instrument was deter- -kt mined experimentally as being (1-e ) where "t" - time and "k" is a constant for any given concentration. Therefore: Output Input' (l-e~kt). This equation was evaluated experimentally by observing the response of the instrument as the concentration of NO or NO£ diluted in nitrogen was switched abruptly from zero to various levels ranging from 0.25 to 2.0 ppm and from these levels back to zero. Samples of NO and NO2 in nitrogen were prepared in Mylar bags. This procedure introduces a step input to the in- strument and produces an output from •which the exponent can be determined and the equation validated. On the basis of LaPlace transforms, the exponential (l-e~kt) is equivalent to the differential expression: \Ts + I/ Where T is the time constant function, f ( — ), and s is the dif- ferential operator (d/dt). In this form the system function is usually called the transfer function. If the output concentration is termed Y and the input concentration is termed X, the equa- tion becomes: 49 ------- 50 IRRADIATION CHAMBER TESTS Y X \Ts + 1 j which reduces to dy _ Y + T dT This differential equation is solved by the computer to ad- just the oxides of nitrogen data at 3-minute intervals. The time constant function, T, is a variable that is dependent upon the output concentration Y. This relationship, determined experi- mentally as previously described, is stored in the computer. The coefficients of this function must, of course, be determined each time the instrument is modified or a different instrument is used. ------- COMPUTER PROGRAM FOR DATA REDUCTION INPUT CONSTANTS 1) to 2) At 3) dtNO 4) °l vmax 6) tits INPUT FUNCTIONS 1) YNO f (t). f (t), 2> YNO2 3) TNO+ 4) TNO- f (YNO) f (YNO) 5) TNQ2+ f (YNOz) 7) PPM NQ = f (XNO) 8) PPM NQ = f (XNo2) Univariate table, up to 200 points (Linear Interpolation) Univariate table, up to 200 points (Linear Interpolation Table or equation, whichever is more convenient for computer programmer. Equation con- stants must be adjustable. OUTPUT DtL 4) PPMNO 5) X NO Additional printout of constants involved in curve fits, etc. should be printed once for each case. 7) XNO2 PROCEDURE 1) Read Program 2) Read Input 3) t =t0 4) YNQ = f (t), Tabular input function 5) Y-KJQ f (t), Tabular input function 51 ------- 52 IRRADIATION CHAMBER TESTS 6) YNO+1 - YNO YNO - YNO-1 (^) _ tYNO+l tYNO tYNO tYNO-l dt NO 2 7) YN02 +1 YN02 YN02 YN02-1 (^) tYN02+l tYN02 tYN02 tYN02-l dt NO — L 2 8) TNO+ = f (YNO) 9) TNO- f (YNO) 10) TN02 + = f (YN02) 11) TN02- f (YN02) 12) (Is dy negative?) Yes -14 (dt~ NO2 No -13 r-r, rp *- 1 K 13) TN02 - TN02+ 14) TNO2 - TNO2- 15) (Is dy negative?) Yes -17 (dF NO No -16 ^r *r * 18 16) 1NO ^-NO4" 17) TNO TNO- 18) XNQ YNQ + TNO 19)XNOz YNOz + TNQ2 |^dt.N02 20) PPMNO - f (XNO) 21)PPMN02 f (XNOz) 21A) tj^ t t^ts 22) tNO 'L dtNO 23) tNOz *L dtNO2 24) t - t + At 25) Is t >tmax. ? Yes-^26 No 4 26) Print Output 27) Go to i, read next case. NOMENCLATURE t0 Time that instrument was put on stream (minutes). At Time increment at which output is to be printed (minutes). Deadtime for NO function. Deadtime for NO2 function. Last time value for input functions YNO and YNO (minutes). Chart indication for NO. YNO+1 Chart indication for NO at next time point. ------- Appendix B 53 ^NO- 1 Chart indication for NO at preceding time point. YNO2 Chart indication for NO2- (^) dt NO Average slope of YNQ function at this time point. dy (— ) -,„ Average slope of YJ^Q,, function at this time point. T"NO+ Time constant for NO function for positive slopes. TJSJQ- Time constant for NO function for negative slopes. TJ^Q . Time constant for NO2 function for positive slopes. TJ^Q Time constant for NO2 function for negative slopes. TNO Time constant used for NO computation. - Time constant used for NO2 computation. XNO Adjusted NO chart indication. XNO-, Adjusted NO2 chart indication. PPM^O Parts per million NO. PPMNo2 Parts per million NO2- tj^jQ Time for NO function after deadtime correction. " Time for NO2 function after deadtime correction. - Time after t0 that lights were turned on. ------- BLOCK DIAGRAM START 1 . READ PROGRAM , READ INPUT ' t - t0 {First time only) YNO = i W YN02 = f (t) INPUT < l^ft * 'YNO*I - '*NO 'VNO-'YNO.I ^0+ = f (YNO) TNO- = f (YNO) TN02+ = f TN02- = f GPO 806-304-5 ------- BIBLIOGRAPHIC: Korth, Merrill W. DYNAMIC IRRADIATION CHAMBER TESTS OF AUTOMOTIVE EXHAUSTS. PHS Publ. No. 999-AP-5. 1963.54pp. ABSTRACT: As part of an intensive study by government and private agencies the U. S. Public Health Service has built an irradiation chamber facility for investigation of irradiated auto exhaust under mixing conditions similar to those in the atmosphere. The facility consists of a programmed continuous-cycling chassis dynamometer, an exhaust dilution system, a dilution-air purification system, two irradiation chambers, and exposure facili- ties for evaluation of bacteria kill, plant damage, and various effects on small animals.. Of the three variables studied during the first test series, the exhaust concentration at the start of irradiation appeared to produce the most significant effects. Fuel composition had a lesser influence. Very little difference was noted in the effects produced at two different average irradiation times. ACCESSION NO. KEY WORDS: Air Pollution Auto Exhaust Irradiation Chamber Photochemistry Vegetation Bacteria Kill BIBLIOGRAPHIC: Korth, Merrill W. DYNAMIC IRRADIATION CHAMBER TESTS OF AUTOMOTIVE EXHAUSTS. PHS Publ. No. 999-AP-5. 1963. 54 pp. ABSTRACT: As part of an intensive study by government and private agencies the U. S. Public Health Service has built an irradiation chamber facility for investigation of irradiated auto exhaust under mixing conditions similar to those in the atmosphere. The facility consists of a programmed continuous-eye ling chassis dynamometer, an exhaust dilution system, a dilution-air purification system, two irradiation chambers, and exposure facili- ties for evaluation of bacteria kill, plant damage, and various effects on small animals. Of the three variables studied during the first test series, the exhaust concentration at the start of irradiation appeared to produce the most significant effects. Fuel composition had a lesser influence. Very little difference was noted in the effects produced at two different average irradiation times. ACCESSION NO. KEY WORDS: Air Pollution Auto Exhaust Irradiation Chamber Photochemistry Vegetation Bacteria Kill ------- |