(i) Recent studies have demonstrated that natural waters can promote the indirect (or sensitized) photoreaction of dissolved organic chemicals. This reactivity is imparted by dissolved organic material (DOM) in the form of humic substances. These materials absorb sunlight and produce reactive intermediates that include singlet oxygen (¹0₂) (Zepp et al. (1977) under paragraph (f)(20) of this section, Zepp et al. (1981) under paragraph (f)(17) of this section, Zepp et al. (1981) under paragraph (f)(18) of this section, Wolff et al. (1981) under paragraph (f)(16) of this section, Haag et al. (1984) under paragraph (f)(6) of this section, Haag et al. (1984) under paragraph (f)(7) of this section); peroxy radicals (RO₂−) (Mill et al. (1981) under paragraph (f)(9) of this section; Mill et al. (1983) under paragraph (f)(8) of this section); hydroxyl radicals (HO−) (Mill et al. (1981) under paragraph (f)(9) of this section, Draper and Crosby (1981, 1984) under paragraphs (f)(3) and (4) of this section); superoxide anion (0₂⁻−) and hydroperoxy radicals (HO−). (Cooper and Zika (1983) under paragraph (f)(1) of this section, Draper and Crosby (1983) under paragraph (f)(2) of this section); and triplet excited states of the humic substances (Zepp et al. (1981) under paragraph (f)(17) of this section, Zepp et al. (1985) under paragraph (f)(21) of this section). Synthetic humic waters, prepared by extracting commercial humic or fulvic materials with water, photoreact similarly to natural waters when optically matched (Zepp et al. (1981) under paragraphs (f)(17) and (18) of this section).

(ii) The indirect photoreactivity of a chemical in a natural water will depend on its response to these reactive intermediates, and possibly others yet unknown, as well as the ability of the water to generate such species. This latter feature will vary from water-to-water in an unpredictable way, judged by the complexity of the situation.

(iii) The approach to standardizing a test for indirect photoreactivity is to use a synthetic humic water (SHW) prepared by water-extracting commercial humic material. This material is inexpensive, and available to any laboratory, in contrast to a specific natural water. The SHW can be diluted to a dissolved organic carbon (DOC) content and uv-visible absorbance typical of most surface fresh waters.

(iv) In recent studies it has been found that the reactivity of SHW mixtures depends on pH, and also the history of sunlight exposure (Mill et al. (1983) under paragraph (f)(11) of this section). The SHW solutions initially photobleach with a time-dependent rate constant. As such, an SHW test system has been designed that is buffered to maintain pH and is pre-aged in sunlight to produce, subsequently, a predictable bleaching behavior.

(v) The purpose of Phase 1 is to prepare, pre-age, and dilute SHW to a standard mixture under defined, reproducible conditions.

(i) Twenty grams of Aldrich humic acid are added to a clean 2-liter Pyrex Erlenmeyer flask. The flask is filled with 2 liters of 0.1 percent NaOH solution. A stir bar is added to the flask, the flask is capped, and the solution is stirred for 1 hour at room temperature. At the end of this time the dark brown supernatant is decanted off and either filtered through coarse filter paper or centrifuged and then filtered through 0.4)m microfilter. The pH is adjusted to 7.0 with dilute H₂SO₄ and filter sterilized through a 0.2)m filter into a rigorously cleaned 2-liter Erlenmeyer flask. This mixture contains roughly 60 ppm DOC and the absorbance (in a 1 cm path length cell) is approximately 1.7 at 313 nm and 0.7 at 370 nm.

(ii) Pre-aging is accomplished by exposing the concentrated solution in the 2-liter flask to direct sunlight for 4 days in early spring or late fall; 3 days in late spring, summer, or early fall. At this time the absorbance of the solution is measured at 370 nm, and a dilution factor is calculated to decrease the absorbance to 0.50 in a 1 cm path length cell. If necessary, the pH is re-adjusted to 7.0. Finally, the mixture is brought to exact dilution with a precalculated volume of reagent-grade water to give a final absorbance of 0.500 in a 1-cm path length cell at 370 nm. It is tightly capped and refrigerated.

(iii) This mixture is SHW stock solution. Before use it is diluted 10-fold with 0.010 M phosphate buffer to produce a pH 7.0 mixture with an absorbance of 5.00 × 10⁻² at 370 nm, and a dissolved organic carbon of about 5 ppm. Such values are characteristic of many surface fresh waters.

(i) Phase 2 measurements provide approximate solar photolysis rate constants and half-lives of test chemicals in PW and SHW. If the photoreaction rate in SHW is significantly larger than in PW (factor of >2X) then the test chemical is subject to indirect photoreaction and Phase 3 is necessary. Phase 2 data are needed for more accurate Phase 3 measurements, which require parallel solar irradiation of actinometer and test chemical solutions. The actinometer composition is adjusted according to the results of Phase 2 for each chemical, to equalize as much as possible photoreaction rate constants of chemical in SHW and actinometer.

(ii) In Phase 2, sunlight photoreaction rate constants are measured in round tubes containing SHW and then mathematically corrected to a flat water surface geometry. These rate constants are not corrected to clear-sky conditions.

(i) Solutions of test chemicals should be prepared using sterile, air-saturated, 0.010 M, pH 7.0 phosphate buffer and reagent-grade (or purer) chemicals.^[1] Reaction mixtures should be prepared with chemicals at concentrations at less than one-half their solubility in pure water and at concentrations such that, at any wavelengths above 290 nm, the absorbance in a standard quartz sample cell with a 1-cm path length is less than 0.05. If the chemicals are too insoluble in water to permit reasonable handling or analytical procedures, 1-volume percent acetonitrile may be added to the buffer as a cosolvent.

(ii) This solution should be mixed 9.00:1.00 by volume with PW or SHW stock solution to provide working solutions. In the case of SHW, it gives a ten-fold dilution of SHW stock solution. Six mL aliquots of each working solution should then be transferred to separate 12 × 100 mm quartz tubes with screw tops and tightly sealed with Mininert valves.^[2] Twenty four tubes are required for each chemical solution (12 samples and 12 dark controls), to give a total of 48 tubes.

(iii) The sample tubes are mounted in a photolysis rack with the tops facing geographically north and inclined 30° from the horizontal. The rack should be placed outdoors over a black background in a location free of shadows and excessive reflection.

(iv) Reaction progress should be measured with an analytical technique that provides a precision of at least ±5 percent. High pressure liquid chromatography (HPLC) or gas chromatograph (GC) have proven to be the most general and precise analytical techniques.

(v) Sample and control solution concentrations are calculated by averaging analytical measurements for each solution. Control solutions should be analyzed at least twice at zero time and at other times to determine whether any loss of chemical in controls or samples has occurred by some adventitious process during the experiment.

(vi) Whenever possible the following procedures should be completed in clear, warm, weather so that solutions will photolyze more quickly and not freeze.

(vii) Since the rate of photolysis in tubes is faster than the rate in natural water bodies, values of near-surface photolysis rate constants in natural and pure water bodies, k_pE and k_DE, respectively, can be obtained from (k_p)_SHW and (k_p)_W from Equations 5 and 6:

Equation 5

k_pE = 0.45(k_p)_SHW

Equation 6

k_DE = 0.45(k_p)_W.

The factor 0.45 is an approximate geometric correction for scattered light in tubes versus horizontal surfaces. A rough value of k_IE, the rate constant for indirect photolysis in natural waters or SHW, can be estimated from the difference between k_pE and k_DE using Equation 7:

Equation 7

k_IE = k_pE−k_DE.

(i) If no loss of chemical is found in dark control solutions compared with the analysis in tubes at zero time (within experimental error), any loss of chemical in sunlight is assumed to be due to photolysis, and the procedure provides a valid estimate of k_pE and k_DE. Any loss of chemical in the dark-control solutions may indicate the intervention of some other loss process such as hydrolysis, microbial degradation, or volatilization. In this case, more detailed experiments are needed to trace the problem and if possible eliminate or minimize the source of loss.

(ii) Rate constants determined by the Phase 2 protocol depend upon latitude, season, and weather conditions. Note that (k_p)_SHW and k_D values apply to round tubes and k_pE and k_DE values apply to a natural water body. Because both (k_p)_SHW and k_D are measured under the same conditions the ratio ((k_p)_SHW/k_D) is a valid measure of the susceptibility of a chemical to indirect photolysis. However, since SHW is subject to photobleaching, (k_p)_SHW will decrease with time because the indirect rate will diminish. Therefore, R >2 is considered to be a conservative limit because (k_p)_SHW will become systematically smaller with time.

(i) Phase 2 testing separates test chemicals into three convenient categories: “Photolabile”, “photoinert”, and those chemicals having sunlight half-lives in round tubes in the range of 1 hour to 50 days. Chemicals in the first two categories fall outside the practical limits of the test, and cannot be used in Phase 3. All other chemicals are suitable for Phase 3 testing.

(ii) The test procedure is simple and inexpensive, but does require that the chemical dissolve in water at sufficient concentrations to be measured by some analytical technique but not have appreciable absorbance in the range 290 to 825 nm. Phase 2 tests should be done during a clear-sky period to obtain the best results. Testing will be less accurate for chemicals with half-lives of less than 1 day because dramatic fluctuations in sunlight intensity can arise from transient weather conditions and the difficulty of assigning equivalent reaction times. Normal diurnal variations also affect the photolysis rate constant. Phase 3 tests should be started as soon as possible after the Phase 2 tests to ensure that the (k_p)_SHW estimate remains valid.

(i) Chemical A was dissolved in 0.010 M pH 7.0 buffer. The solution was filtered through a 0.2 )m filter, air saturated, and analyzed. It contained 1.7 × 10⁻⁵ M A, five-fold less than its water solubility of 8.5 × 10⁻⁵ M at 25 °C. A uv spectrum (1-cm path length) versus buffer blank showed no absorbance greater than 0.05 in the wavelength interval 290 to 825 nm, a condition required for the Phase 2 protocol. The 180 mL mixture was diluted by the addition of 20 mL of SHW stock solution.

(ii) The SHW solution of A was photolyzed in sealed quartz tubes (12 × 100 mm) in the fall season starting on October 1. At the end of 1 and 2 days, respectively, the concentration of A was found to be 1.13 × 10⁻⁵ M and 0.92 × 10⁻⁵ M compared to unchanged dark controls (1.53 × 10⁻⁵ M).

(iii) The tube photolysis rate constant of chemical A was calculated from Equation 2 under paragraph (c)(2)(vi)(B) of this section. The first time point at day 1 was used because the fraction of A remaining was in the range 20 to 80 percent:

(k_p)_SHW = (1/1d)Pn(1.53 × 10⁻⁵/1.13 × 10⁻⁵) (k_p)_SHW = 0.30 d⁻¹.

(iv) From this value, k_pE was found to be 0.14 d−¹ using equation 5 under paragraph (c)(2)(vii) of this section:

k_pE = 0.45(0.30 d⁻¹) = 0.14d⁻¹.

(v) From measurements in pure water, k_D for chemical A was found to be 0.085 d⁻¹. Because the ratio of (k_p)_SHW/k_D( = 3.5) is greater than 2, Phase 3 experiments were started.

(i) The purpose of Phase 3 is to measure k_Io, the indirect photolysis rate constant in tubes, and then to calculate k_pE for the test chemical in a natural water. If the approximate (k_p)_SHW determined in Phase 2 is not significantly greater than k_D measured for the experiment date of Phase 2, then Phase 3 is unnecessary because the test chemical is not subject to indirect photoreaction.

(ii) In the case (k_p)_SHW is significantly larger than k_D, Phase 3 is necessary. The rate constant (k_p)_SHW is used to choose an actinometer composition that matches the actinometer rate to the test chemical rate. Test chemical solutions in SHW and in pure water buffer are then irradiated in sunlight in parallel with actinometer solutions, all in tubes.

(iii) The actinometer used is the p-nitroacetophenone-pyridine (PNAP/PYR) system developed by Dulin and Mill (1982) under paragraph (f)(5) of this section and is used in two EPA test guidelines (USEPA (1984) under paragraphs (f) (14) and (15) of this section). By varying the pyridine concentration, the PNAP photolysis half-life can be adjusted over a range of several hours to several weeks. The starting PNAP concentration is held constant.

(iv) SHW is subject to photobleaching that decreases its ability to promote indirect photolysis based on its ability to absorb sunlight. This effect will be significant when the test period exceeds a few days. To correct for photobleaching, tubes containing SHW are irradiated in action to the other tubes above.

(v) At any time, the loss of test chemical is given by Equation 8 assuming actinometric correction to constant light flux:

Equation 8

−(d[C]/dt) = k_I[C] + k_D[C].

(vi) The indirect photolysis rate constant, k_I, is actually time dependent because SHW photobleaches; the rate constant k_I, after pre-aging, obeys the formula:

Equation 9

k_I = k_Io exp(−kt),

in which k_Io is the initial indirect photoreaction rate constant and k is the SHW photobleaching rate constant. After substituting equation 9 for k_I in Equation 8 under paragraph (d)(1)(v) of this section, and rearranging, one obtains

−(d[C]/[C] = k_Io[exp(−kt)]dt + k_D dt.

This expression is integrated to give Equation 10:

Equation 10

Pn(C_o/C)_SHW = (k_Io/k)[1−exp(−kt)] + k_D t.

The term (k_Io/k) can now be evaluated. Since in pure water, Pn(C_o/C)_W = k_D t, then subtracting this equation from Equation 10 gives

Equation 11

Pn(C_o/C)_SHW-Pn(c_o/C)_W = (k_Io/k)[1-exp(-kt)].

The photobleaching fraction, [1-exp(-kt)], is equivalent to the expression [1-(A₃₇₀/A°₃₇₀)], where A°₃₇₀ and A₃₇₀ are the absorbances at 370 nm, and are proportional to humic sensitizer content at times zero and t. Therefore, (k_Io/k) is derived from the slope of a linear regression using [Pn(C_o/C)_SHW-Pn(C_o/C)_W] as the dependent variable and [1-(A₃₇₀/A°₃₇₀)_SHW] as the independent variable.

(vii) To evaluate k_Io, the parameter k has to be evaluated under standard sunlight conditions. Therefore, the photolysis rate constant for the PNAP/PYR actinometer (k_A) is used to evaluate k by linear regression on Equation 12:

Equation 12

Pn(A°₃₇₀/A₃₇₀) = (k/k_A)Pn(C_o/C)_PNAP,

where the slope is (k/k_A) and the value of k_A is calculated from the concentration of pyridine and the absorption of light by PNAP: k_A = 2.2(0.0169)[PYR]k_a. Values of k_a are listed in the following Table 1.

The value of k_Io is then given by Equation 13:

Equation 13

k_Io = (k_Io/k)(k/k_A)k_A.

(viii) To obtain k_D, determine the ratio (k_D/k_A) from a linear regression of Pn(C_o/C)_W versus Pn(C_o/C)_PNAP according to Equation 13a:

Equation 13a

Pn(C_o/C)_W = (k_D/k_A)Pn(C_o/C)_PNAP.

The slope is (k_D/k_A), and k_D is obtained by multiplication of this slope with the known value of k_A: i.e., k_D = (k_D/k_A)k_A.

(ix) Then, (k_p)_SHW values in SHW are determined by summing k_D and K_Io as follows:

Equation 14

(k_p)_SHW = k_Io + k_D.

(x) Finally, k_pE is calculated from the precise relationship, Equation 5a:

Equation 5a

k_pE = 0.455(k_p)_SHW.

(i) Using the test chemical photoreaction rate constant in round tubes, (k_p)_SHW′ determined in Phase 2 under paragraph (c) of this section, and the absorption rate constant, kα found in Table 1, under paragraph (d)(1)(vii) of this section, calculate the molar pyridine concentration required by the PNAP/PYR actinometer using Equation 15:

Equation 15

[PYR]/M = 26.9[(k_p)_SHW/k_a].

This pyridine concentration makes the actinometer rate constant match the test chemical rate constant.

(ii) Once [PYR] is determined, an actinometer solution is prepared by adding 1.00 mL of 1.0 × 10⁻² M (0.165 gms/100 mL) PNAP stock solution (in CH₃ CN solvent) and the required volume, V, of PYR to a 1 liter volumetric flask. The flask is then filled with distilled water to give 1 liter of solution. The volume V can be calculated from Equation 16:

Equation 16

V/mL=[PYR]/0.0124.

The PNAP/PYR solutions should be wrapped with aluminum foil and kept out of bright light after preparation.

(iii) The following solutions should be prepared and individually added in 6.00 mL aliquots to 12/100 mm quartz sample tubes; 8 tubes should be filled with each solution:

(iv) The tubes are placed in the photolysis rack (Phase 2, Procedure) at 0900 hours on day zero, with the controls taped to the bottom of the rack. One tube of each composition is removed, along with their respective controls, according to a schedule found in Table 2, which categorizes sampling times on the basis of (k_p)_SHW determined in Phase 1.

(v) The tubes containing PNAP, test chemical, and their controls are analyzed for residual concentrations soon after the end of the experiment. PNAP is conveniently analyzed by HPLC, using a 30 cm C₁₈ reverse phase column and a uv detector set at 280 nm. The mobile phase is 2 percent acetic acid, 50 percent acetonitrile and 48 percent water (2 mL/min flow rate). Tubes containing only SHW (solution D) should be analyzed by absorption spectroscopy at 370 nm after storage at 4 °C in the dark. The absorbance range to be measured is 0.05 to 0.01 AU (1 cm).

(vi) If controls are well-behaved and show no significant loss of chemical or absorbance change, then k_I can be calculated. In tabular form (see Table 4 under paragraph (d)(6)(iii)(A) of this section) arrange the quantities Pn(C_o/C_t)_SHW, Pn(C_o/C_t)_SHW, [1−(A₃₇₀/A^o₃₇₀)], Pn(A^o₃₇₀/A₃₇₀), and Pn(C_o/C)_PNAP in order of increasing time. According to Equation 11 under paragraph (d)(1)(vi) of this section in the form of Equation 17,

Equation 17

Pn(C_o/C)_SHW−Pn(C_o/C)_W = (k_Io/k)[1−(A₃₇₀/A^o₃₇₀)],

plot the quantities [Pn(C_o/C_t)_SHW−Pn(C_o/C_t)_W] versus the independent variable [1−(A₃₇₀/A^o₃₇₀)]. Obtain the slope (S1) by least square linear regression. Under the assumptions of the protocol, S1 = (k_Io/k).

(vii) According to Equation 12 under paragraph (d)(1)(vii) of this section, plot the quantities Pn(A^o₃₇₀/A₃₇₀) versus the independent variable Pn(C_o/C_t)_PNAP. Obtain the slope (S2) by least squares linear regression on Equation 12 under paragraph (d)(1)(vii) of this section. Under the assumptions of the protocol, S2 = (k/k_A).

(viii) Then, using Equation 13a under paragraph (d)(1)(vii) of this section, determine the slope (S3) by least squares linear regression. Under the assumptions of the protocol, S3 is equal to (k_D/k_A).

(ix) From Equation 18

Equation 18

k_A = 0.0372[PYR]k_a,

calculate k_A using k_a values found in Table 1 under paragraph (d)(1)(vii) of this section. The value of k_a chosen must correspond to the date closest to the mid-experiment date and latitude closest to that of the experimental site.

(x) The indirect photoreaction rate constant, k_Io, is determined using Equation 19,

Equation 19

k_Io = (S1)(k_A)(S2),

by incorporating the quantities k_A, S1, and S2 determined as described in paragraphs (d)(2) (ix), (vi), and (vii) of this section, respectively.

(xi) The rate constant k_D is calculated from Equation 20,

Equation 20

k_D = (S3)(k_A),

using the quantities S3 and k_A determined as described above.

(xii) Then, (k_p)_SHW is obtained by summing k_D and k_Io, as described by Equation 14 in paragraph (d)(1)(ix) of this section:

Equation 14

(k_p)_SHW = k_Io + k_D.

(xiii) Finally, k_pE is obtained by multiplying (k_p)_SNW by the factor 0.455, as described by Equation 5a in paragraph (d)(1)(x) of this section:

Equation 5a

k_pE = 0.455 (k_p)_SHW

As determined, k_pE is the net environmental photoreaction rate constant. It applies to clear sky conditions and is valid for predicting surface photoreaction rates in an average humic containing freshwater body. It is strictly valid only for the experimental latitude and season.

(i) From Phase 2 testing, under paragraph (c)(6)(iii) of this section, chemical A was found to have a photolysis rate constant, (k_p)_SHW′ of 0.30 d⁻¹ in fall in round tubes at latitude 33° N. Using Table 1 under paragraph (d)(1)(vii) of this section for 30° N, the nearest decadic latitude, a fall value of k_a equal to 333 d⁻¹ is found for PNAP. Substitution of (k_p)_SHW and k_a into Equation 15 under paragraph (d)(2)(i) of this section gives [PYR] = 0.0242 M. This is the concentration of pyridine that gives an actinometer rate constant of 0.30 d⁻¹ in round tubes in fall at this latitude.

(ii) The actinometer solution was made up by adding a volume of pyridine (1.95 mL) calculated from equation 16 under paragraph (d)(2)(ii) of this section to a 1 liter volumetric flask containing 1.00 mL of 1.00 × 10⁻² M PNAP in acetonitrile. The flask was filled to the mark with distilled water to give final concentrations of [PYR] = 0.0242 M and [PNAP] = 1.00 × 10⁻⁵ M. Ten tubes of each of the following solutions were placed in the photolysis rack at 1,200 hours on day zero:

(iii) The test chemical had been placed in category B, Table 2 under the paragraph (d)(2)(iv) of this section, on the basis of its Phase 2 rate constant under paragraph (c) of this section. Accordingly, two tubes of each irradiated solution and two tubes of each blank solution were removed at 0, 1, 2, 4, and 8 days at 1,200 hours. The averaged analytical results obtained at the end of the experiment are shown in the following Table 3.

Data for solutions A through D are given in column 2 through 5, respectively. No significant chemical loss was found in the dark controls.

(B) Slope S1 = (k_Io/k) was calculated according to Equation 17 under paragraph (d)(2)(vi) of this section and was found to be 4.96 by a least squares regression with a correlation coefficient equal to 0.9980. The following Figure 1 shows a plot of Equation 17 under paragraph (d)(2)(vi) of this section and its best-fit line.

Figure 1 - Graphic determination of S1 = (k_Io/k) based on Equation 17 under paragraph (d)(2)(vi) of this section.

(C) Slope S2 = (k/k_a) was also derived from Table 4 under paragraph (d)(6)(iii)(A) of this section by a fit of Pn(A^o₃₇₀ /A₃₇₀)_SHW and Pn(C_o /C)_PNAP to Equation 12 under paragraph (d)(l)(vii) of this section. This plot is displayed in the following Figure 2; the slope S2 was found to be 0.295 and the correlation coefficient was equal to 0.9986.

Figure 2 - Graphic determination of S2 = (k/k_A) based on Equation 12 under paragraph (d)(1)(vii) of this section.

(i) Specific analytical and recovery procedures.

(ii) Other test conditions.

(i) Phase 2 Screening Test under paragraph (c) of this section.

(ii) Phase 3 - Indirect photoreaction with actinometer.