Insights into the Photochemical Transformation of Iodine in Aqueous

Oct 8, 2012 - Marine aerosol is highly enriched in iodine, mostly in the form of iodate (IO3–) ions, compared to its relative abundance in seawater...
2 downloads 16 Views 1MB Size
Article pubs.acs.org/est

Insights into the Photochemical Transformation of Iodine in Aqueous Systems: Humic Acid Photosensitized Reduction of Iodate Russell W. Saunders, Ravi Kumar,† Samantha M. MacDonald, and John M. C. Plane* School of Chemistry, University of Leeds, Leeds LS2 9JT U.K. S Supporting Information *

ABSTRACT: Marine aerosol is highly enriched in iodine, mostly in the form of iodate (IO3−) ions, compared to its relative abundance in seawater. This paper describes a laboratory study of the photochemical reduction of IO3− in the presence of humic acid. Spectroscopic analysis showed that ∼20% of IO3− was converted to “free” iodide (I−) ions and this fraction remained constant as a function of time. Direct detection of an organically fixed fraction (i.e., ∼ 80%) was not possible, but a number of test reactions with surrogate organic compounds containing functional groups identified in humic acid structures indicate that efficient substitution of iodine occurs at aromatic 1,2 diol sites. These iodinated humic acids are stable with respect to photolysis at near-UV/ visible wavelengths and are likely to account for a significant proportion of the soluble iodine-containing organic material occurring within aerosols. In the lower atmosphere, oxidation of I− to I2 in marine aerosol occurs mostly through the uptake of O3, with H2O2 playing a very minor role. A model of iodine chemistry in the open ocean tropical boundary layer, which incorporates these experimental results, is able to account for the observed enrichment of iodine in marine aerosol.

1. INTRODUCTION

therefore likely to vary considerably in importance within different aerosol size modes. Chemical speciation studies in waters and soils show that an important class of compound is likely to be humic acid (HA), for which iodine has been shown to readily be taken up and fixed into a nonvolatile form.13,14 Humic acids represent a diverse group of multifunctional organic compounds which are soluble in water at pH values greater than 2. Such compounds are known to act as chromophores which participate in photosensitized chemical reactions in natural waters.15 Humiclike substances (HULIS) have also been identified as a significant component in atmospheric aerosol.16 While a number of techniques have been used to attempt to identify the mechanism by which iodine is assimilated into such compounds, no firm picture has yet emerged to fully characterize the chemical pathways leading to the formation of iodinated humic substances. Another unresolved issue is the subsequent fate of iodide in aged aerosol. The uptake at aerosol surfaces of soluble, reactive atmospheric species such as O3 and H2O2 from the gas phase will potentially lead to the oxidation of I− (directly aerosolized from seawater17 or formed via iodate reduction) to I2 and HOI. Evasion of these relatively insoluble species to the gas phase would thus recycle iodine into actively O3-depleting species such as I and IO and therefore provide a potentially important

The speciation of iodine within aquatic and terrestrial ecosystems is crucial in the cycling and bioavailability of the element to a host of diverse organisms including humans.1 In natural waters, the inorganic forms of iodate (IO3−) and iodide (I−) have been extensively studied in a wide variety of locations,2,3 with the latter having highest concentrations (typically 0.1−0.2 μM) at the surface. The direct photoreduction of IO3− to I− only occurs at wavelengths below the actinic cutoff of 290 nm. Hence this conversion is biologically mediated,4 and organic matter also plays an important role.5 Dissolved organic forms have been studied in detail with both volatile (i.e., organo-iodides such as CH3I or CH2I2), and nonvolatile fractions being identified.6−8 While the volatile forms (organo-iodides and I2) are present at very low levels (10 h) irradiation of the HA−iodate solution to which chloride ions had been added at a concentration of 0.5 M, i.e. similar to that found in seawater. The first-order loss rate of iodate during the periods of irradiation shown in the top panel of Figure 2 is (1.2 ± 0.3) × 10−5 s−1. The dashed black line between the data points at t = 10 and 12 h indicates that, even though the solution was stored under dark conditions overnight (15 h), a slow dark reaction takes place between the two reactants at a rate which is about 20% of the photochemical rate. This dark reaction was further investigated by monitoring spectral changes from a HA solution mixed with a 10−3 M solution of NaIO3, kept in the dark, and

Figure 2. Top panel: variation of the fitted absorbances of iodate (IO3−), iodide (I−), and humic acid (HA) as a function of photolysis time. The solid black lines indicate linear fits to I− data points during periods of continuous irradiation. The dashed black line indicates the increase resulting from storing the solution overnight in the dark for 15 h, followed by a 2 h photolysis period. Bottom panel: calculated fraction of I− in solution resulting from the photosensitized reduction of IO3− (the dashed line indicates the average value of 0.19 (±0.02)).

aliquots taken for spectroscopic analysis at the same time each day over a week. A continual change in HA absorbance from 400 to 600 nm was observed, with an ∼12% decrease after the first day and a total decrease of ∼62% after one week. However, even after one week, there was no evidence in these spectra for any additional absorbance due to products formed by the dark reaction. Next, we conducted spectroscopic studies of the dark- and light-initiated reactions between iodate solutions and a number of surrogate compounds with potentially reactive functional groups found within humic acids (see chapter 9 of Stevenson28), to try to identify the key functional groups which are involved in the HA−iodate reactions. These compounds included catechol (1,2-benzenediol), 1,4 (para)-benzoquinone, salicylic (2-hydroxybenzoic) acid, and phthalic (1,2-benzenedicarboxylic) acid. The molecular structures of these compounds are shown in Figure S2 in the Supporting Information. The last two compounds showed no detectable reaction with iodate under either dark or irradiated conditions, i.e. these species do not undergo photolysis at near UV−visible wavelengths and therefore, the reactive species (which would then initiate iodate reduction) are not generated, in contrast to photolysis of HA. Para-benzoquinone similarly showed no reaction with iodate in the absence of visible light. Under irradiation, this compound is converted to the 2,5 dihydroxy form29 which has a small visible absorption feature peaking at 11856

dx.doi.org/10.1021/es3030935 | Environ. Sci. Technol. 2012, 46, 11854−11861

Environmental Science & Technology

Article

480−490 nm, giving rise to a distinct pink coloration to the initially colorless solution. However, there was no evidence for any subsequent reaction of this species with iodate. In contrast, catechol clearly reacts with iodate. Figure 3 shows UV−vis spectra taken from high-concentration (0.1 M)

Figure 4. Comparison of the spectra of the iodate−catechol reaction intermediate (0.1 M catechol and iodate solutions) formed under dark conditions, with the 4-iodocatechol sample (see Figure S3 in the Supporting Information).

absorption feature (at around 400 nm) will be at a slightly shifted position for each species. Therefore, the absorption spectrum of the catechol−iodate intermediate shown in Figure 4 will be a composite of these 3 species. Consequently, the slight “blue shift” (5−10 nm) compared to the spectrum of the only pure sample at our disposal is likely to result from such a “mixture” of the different iodinated forms. The relatively small molar extinction coefficient of the iodocatechol species in the visible region would explain why the corresponding transition was not observed in the HA−iodate solutions, where HA absorption would be dominant. In addition, the disappearance of the iodocatechol peak when photolyzed under visible light (as seen in Figure 3) is consistent with a relatively small aromatic C−I bond energy of ∼230 kJ mol−1,32 equivalent to a threshold photon wavelength of ∼520 nm. 3.2. Photoreduction Mechanism. These experiments show that aerosol IO3− should be reduced photochemically in the lower atmosphere in the presence of HA on a time scale of about 2 days. We infer that the major product is soluble organic iodine, with about 80% yield. Although this product was not observed directly in our experiments, our conclusion is consistent with a number of previous studies using other techniques which indicate that the majority of iodine is in a bound organic form. For example, a radioactive iodine study reported bound organic fractions of 70−80% from experiments under similar solution conditions, i.e. 25 °C and pH ∼7.33 The following sequence of steps is likely to occur in the reduction of iodate in marine aerosol. First, the irradiation at near-UV/visible wavelengths of organic chromophores such as humic acids in water results in the generation of solvated electrons.25,34 The iodate ion is known to be an effective scavenger of solvated electrons, with a rate coefficient of 7.7 × 109 M−1 s−1 at pH = 7.35 Reduction of IO3− in this manner will produce HOI (IO3− + 4e− + 5H+ → HOI + 2H2O) and then I2. At typical seawater/fresh sea salt aerosol pH (7−8), these pathways are likely to be slow, but in more acidic forms such as aged sea salt and non sea salt sulfate aerosol, significantly faster. Both HOI and I2 have been proposed to undergo substitution reactions within the HA structure (e.g., 15, 36−38). Substitution most likely takes place at diphenolic functional groups, as observed with the catechol−iodate reaction discussed above. Aromatic 1,2 diol groups have also been identified in the structure of fulvic acid,39 another important organic component of natural waters which has been shown to

Figure 3. Comparison of the solution spectra of catechol (0.1 M; molecular structure shown), a mixture of catechol and iodate kept in the dark, and the same mixture exposed to sunlight for 2 h.

solutions of (i) catechol (black line), (ii) catechol added to iodate under dark conditions and mixed for 2 h (blue line), and (iii) the solution obtained from (ii) which was then irradiated by natural sunlight for a further 2 h (red line). The high initial catechol concentration was chosen merely to ensure that any spectroscopic changes were clearly above the detection limit of the spectrometer. Catechol itself does not undergo photolysis at visible wavelengths, and so a dark reaction occurs with a broad absorption feature beyond ∼350 nm being evident. It should be noted that we observed the same absorption feature produced by the reaction of catechol solutions with both I− and I2 solutions, thereby indicating a general propensity of iodine to be substituted at this functional group, regardless of the initial iodine oxidation state (we refer the reader to the recent review article of Stavber et al.30 for details of the mechanisms of these reactions). Under irradiation, this absorption is seen to disappear, with the release of I− back into solution. There was no evidence for the formation of I2 in aqueous solution which is characterized by a broad absorption peaking at ∼460 nm, and an additional absorption peak at ∼350 nm due to the triiodide (I3−) species.23 To corroborate whether the observed intermediate compound formed by the reaction of iodate and catechol solutions under dark conditions was indeed an iodinated catechol species, we obtained a sample of laboratory synthesized 4-iodocatechol.31 The solid sample was first dissolved in ultrapure water to give a solution concentration of 6.8 × 10−3 M. This stock solution was then used to prepare a number of more dilute solutions for UV−visible spectroscopic analysis (see Figure S3 in the Supporting Information). Figure 4 shows a comparison of the spectra of the 4iodocatechol sample and that of the catechol−iodate dark reaction product. Iodination of catechol is likely to produce not only 4-iodocatechol, but also 3-iodocatechol and 3,4diiodocatechol, i.e. there is no specificity to the iodination process. These iodinated species are not commercially available for spectroscopic analysis and require specialized synthesis and purification procedures which lie beyond the capability of our laboratory (we were fortunate to obtain the purified 4iodocatechol; see Acknowledgment). The aromatic C−I 11857

dx.doi.org/10.1021/es3030935 | Environ. Sci. Technol. 2012, 46, 11854−11861

Environmental Science & Technology

Article

participate in aqueous iodine chemistry. 40 While the iodocatechol species formed by the catechol−iodate reaction was observed to decompose under visible irradiation, the proportion of free iodide in the HA−iodate reactions was found to remain constant (∼20%) throughout the course of the irradiation experiments. This indicates that while the “isolated” iodinated catechol species readily undergoes photolysis, within the complex structure of the HA, photolysis of this moiety does not occur. 3.3. H2O2 Oxidation of Iodide-Evasion of I2. Figure 5 is an example of the total detected particle mass (and

the role of H+ ions in both the formation of HOI (which acts as an autocatalytic species) and the subsequent formation of I2 (see Schmitz42 for the detailed mechanism). The rate of reaction of I− with H2O2 to form HOI includes a pH dependent term in the rate coefficient (k = 1.2 × 10−2 + 0.17[H+] M−1 s−1 42) which is very slow at pH 7−8 but becomes important when the pH is below 2. The particle masses measured in the present experiments can be used to estimate the time scale for converting IO3− to I2 (via I−) as a result of the HA photosensitized reduction of the initial iodate solution. The flux of I2 from solution was calculated using the average particle mass during each experimental run, the efficiency for converting I2 to IOPs in this system (determined at 0.042%), and the flow rate through the solution cell. Knowing the number of molecules of IO3− in solution, the e-folding lifetime for conversion of IO3− to I2 was calculated to be 50 days without added H2O2, decreasing to 30 days in the presence of 6.6 × 10−5 M H2O2. These time scales are clearly much longer than the time constant for IO3− reduction to I−. The I− + H2O2 reaction should therefore be rate-determining; using the rate coefficient above from Schmitz,42 the time constant would be 15 days, which is in reasonable accord with the estimate of 30 days from the measured IOP mass. In contrast, the characteristic time scale for I− oxidation in the presence of ambient lower tropospheric O3 should be only a few seconds as discussed below. 3.4. Modeling and Atmospheric Implications. To explore the implications of these laboratory results, we now employ a 1-D chemistry transport model of the marine boundary layer. The model THAMO was used previously43 to model the chemistry of iodine at Cape Verde (16.9° N, 24.9° W), which is an open ocean tropical location where the IO radical has been observed along with a large number of ancillary chemical and physical measurements.43,44 We therefore use THAMO for the present study under Cape Verde conditions (for April, though there is relatively little seasonal variation). The model contains 200 stacked boxes at a vertical resolution of 5 m (total height 1 km). Wind speed measurements collected at three heights (4, 10, and 30 m) were used to construct an eddy diffusion coefficient (Kz) profile.43,45 Aircraft measurements of temperature in the boundary layer (BL) at Cape Verde show a strong temperature inversion about 1 km from the surface. Hence, the Kz profile is assumed to increase up to a height of 30 m (where it peaks at 3 × 104 cm2 s−1), after which it decreases at a constant rate to a value of 2 cm2 s−1 at the top of the BL. The model treats iodine, bromine, O3, NOx, and HOx chemistry using over 210 reactions. The chemical scheme is from Saiz-Lopez et al.46 and Mahajan et al.43 The model is constrained with typical measured values of the following species: [NOx] = 25 ppt; [CO] = 110 ppb; [DMS] = 30 ppt; [CH4] = 1820 ppb; [ethane] = 925 ppt; [CH3CHO] = 970 ppt; [HCHO] = 500 ppt; [isoprene] = 10 ppt; [propane] = 60 ppt; and [propene] = 20 ppt.44,47−50 The average background volumetric aerosol surface area used is 1.0 × 10−6 cm2 cm−3, corresponding to a volumetric volume of 1.3 × 10−11 cm3 cm−3, in agreement with average values measured at Cape Verde.51 The modeled HOx concentrations are in sensible accord with measured values at Cape Verde.52 As discussed in Mahajan et al.,43 an additional sea-to-air flux of I2 was included in the model to supplement the organoiodide flux, in order to match the observed levels of IO in the lower marine boundary layer which peak at around 1.4 ppt at

Figure 5. Variation in total detected particle mass (left-hand side axis) resulting from the irradiation of an iodate−humic acid solution (10−4 M IO3−) without (red) and with 6.6 × 10−5 M H2O2 added (green). Scan-by-scan variability associated with the plotted data points was