Stratospheric Ozone Destruction by Man-Made Chlorofluoromethanes

Calculations indicate that chlorofluoromethanes produced by man can greatly affect the concentrations of stratospheric ozone in future decades. This effect follows the release of chlorine from these compounds in the stratosphere. Present usage levels of chlorofluoromethanes can lead to chlorine-catalyzed ozone destruction rates that will exceed natural sinks of ozone by 1985 or 1990.

cipally CF2Cl2 and CFC13, are being produced as aerosol propellants and refrigerants in large and growing amounts, and their atmospheric concentrations are increasing (1)(2)(3). These compounds have been considered valuable as tracers of atmospheric motions because they are relatively inert chemically with atmospheric lifetimes exceeding 10 years (4). Unlike CCl4 (1), they seem to have no natural sources or sinks in the troposphere; their lifetimes are controlled by diffusion into the stratosphere where they can be photodissociated by ultraviolet light (5). Molina and Rowland (5) have noted that this stratospheric sink for CFClI, also represents a potential sink for stratospheric 03. This is so because the photodissociation of CF.CIY releases chlorine atoms which can catalytically destroy 03 through reactions like those of the nitrogen oxides (NO) with 03 (5)(6)(7)(8). Because of the great importance of stratospheric 03 (9), we have reexamined this potential effect and its likely time evolution. We find that current CFXCIY usage levels and trends can lead to chlorine-catalyzed 03 destruction rates exceeding all natural sinks of stratospheric 03 by the early 1980's. Stratospheric changes will continue long after ground-level emissions cease. For example, if emissions were curtailed now, the resultant 03 destruction would maximize around 1990 and would remain significant for several decades. Our calculations also indicate that the ClX concentrations (the sum of the concentrations of Cl, ClO, and HCl) will increase significantly in the stratosphere but not in the troposphere. 27 SEPTEMBER 1974 Langmore, Science 168, 1338 (1970); F P.
Ottensmeyer, E. E. Schmidt, A. J. Olbrecht, ibid. 179, 175 (1973 Hence, critical monitoring of this problem will require measurements in the stratosphere; tropospheric observations alone will not suffice. After release, CFrCl, molecules diffuse upward to be photolyzed by solar radiation (chiefly 175to 220-nm ultraviolet wavelengths) in the stratosphere. On the basis of several atmospheric measurements (1-3) and known chemical properties of the chlorofluoromethanes, stratospheric photolysis ap- pears to be the major sink for CF,ClI (5). Several other possible sinks have been suggested but appear to be negligible (10). After release of the first chlorine atom by a solar photon, chemical reactions will probably remove the remaining chlorine and fluorine atoms from the CFXC1,,_1 radical, temporarily forming phosgene-type molecules (5). Once free in the stratosphere, chlorine atoms can catalyze the recombination of O3 and atomic oxygen. The key reaction are (5)(6)(7)(8) Cl + 03 -* ClO + 02 ClO + O Cl + 02 03+O-02 + 2 (net)

Stratospheric Ozone Destruction by Man-Made Chlorofluoromethanes
Abstract. Calculations indicate that chlorofluoromethanes produced by man can greatly affect the concentrations of stratospheric ozone in future decades. This effect follows the release of chlorine fromn these compounds in the stratosphere. Present usage levels of chlorofluoroinethanes can lead to chlorine-catalyzed ozone destruction rates that will exceed natural sinks of ozone by 1985 or 1990.
We have quantitatively estimated globally averaged rates of 03 destruction due to the chlorine atoms released in the initial photodissociation from man-made CFXCIU. Figure Figure 1 also presents globally integrated 0 destruction rates due to the chemical reactions of oxygen alone (labeled "Chapman") and for the NOx catalytic cycles. These two O3 sinks are currently believed to control stratospheric O 1 concentrations (8, 12). Models 1 and 2 yield rapidly increasing 03 destruction rates that will equal the natural sinks by about 1982 and 1986, respectively. Model 3 shows that, if CF",C1, emissions were curtailed now, the ensuing O3 destruction rates would maximize around 1990 at a rate comparable to major natural cycles and would persist for several decades. Even larger effects may be possible becauLse in our calculations we consider only the first chlorine atom released from each CFLCI molecule; in reality, all four halogen atoms may be freed (13). Important parameters in calculations leading to Fig. 1 appear in Table 1. To account for vertical transport in our time-dependent, one-dimensional (altitude) model we adopted and smoothed the relatively high eddy diffusion coefficient (K) profile of Wofsy and McElroy (14). To compute photodissociation coefficients, J, for CF,Cl, and CFCI3 we used available laboratory photoabsorption data (15). We evaluated the 03-destroying effe^t of the chlorine oxides (ClO,.) by using the photochemical reaction scheme of Stolarski and Cicerone (6) after adding the reactions ClO + NO -4 Cl + NO., and H., + Cl --HC1+ H (16). In the calculations we assumed an atmospheric background of 0, 03, CH4, OH, NO, and Ho that was constant in time. A constant O3 profile representing present conditions is probably not consistent with the large 03-destroying rates of CIO., of Fig. 1, but through its use we can compare the effect of the CF.,.Cl1 to the present natural 03 sinks. Other procedures and input data are discussed in (17).
The accumulation (mixing ratio) of CFX.Cl,, molecules is shown as a function of altitude and time in Fig. 2, predicted from emission history model 2.
Corresponding profiles are also shown for CIX. Because the relevant photochemical processes are fast relative to the transport processes, one may calculate the amounts of Cl, CIO, and HCI present at each altitude from the CIX concentration and the equations of photochemical equilibrium (6)(7)(8). The time evolution of the CIX profiles show large increases above 20 km and virtually no change below 15 km. The precise CIX concentration predicted for a given altitude and time depends on several computational parameters, especially K, but the profile shapes clearly indicate the effects of a mid-stratospheric source (see the Q(, values in emissions are assumed. This man-made ClX source is comparable to the projected NO.r input due to large supersonic transport (SST) fleets, and it appears that CIO,, destroys 03 more efficiently than NO,. (6). McElroy et al.
(12) find that 03 depletion from SST NO, is sensitive to the injection altitude. The 03 effect due to added CIX will be difficult to evaluate, but current models (6)(7)(8) indicate that CIX added above 25 km is much more important than that added below.
We have not attempted to predict 03 concentrations or their changes due to increasing amounts of CIX from CFCl,, usage. Predictions of 03 concentrations and trends due to perturbations are still the subject of considerable debate (18). However, Fig. 1 indicates that, regardless of the precise magnitude of the effect of increasing destruction rates on 03 content, chlorine atom produc-    by reactions in the atmosphere. Although the fluorine atoms released by similar reactions can also destroy 03 catalytically, the reaction CH4 + F e* CHa + HF will terminate the fluorine chains more quickly than CH4 + Cl -) CH, + HCI does for chlorine. Further, reinitiation of fluorine chains will not proceed via HF + OH -F + H20, whereas HCI +OH-e Cl + H20 does reinitiate the chlorine chains. 14. S. C. Wofsy  then served as chlorine production rates, Q, in the photochemical, vertical transport model for CIX. Table 1 shows an example of Q along with selected input data. The OH profile was taken from (12); the reaction rate for HCI + OH -Cl + H20 was 2 X 10-12 exp Amounts of methane in the marine environment in excess of that resulting from equilibration with the atmosphere most frequently occur in anoxic waters such as those found in fjords (1, 2) and in the interstitial water of anoxic sediments (3)(4)(5)(6). In the absence of sources associated with shipping and industrial activity or natural seeps from oil and gas reservoirs (7), most of this methane results from anaerobic bacterial decomposition of organic matter. Methane bacteria are readily found in anoxic environments, where they are terminal organisms in the microbial food chain (8); moreover, there is some evidence that methane is not produced until dissolved sulfate has been previously removed by sulfate-reducing bacteria (5,6).
The interstitial waters of Recent organic-rich marine sediments are ideal for studying the relation between methane and dissolved sulfate distributions 27 SEPTEMBER 1974 (-335/T) (where T is the absolute temperature) (J. G. Anderson, personal communication). The remainder of the photochemical scheme and input data were adopted from (6). For the CIX calculations the boundary conditions were: zero flux at the upper boundary (80 km), ground upward flux of 2 X 1010 cm-2 sec-I [see (7)], and diffusion to the ground at a velocity of 0.4 cm sec-1. These conditions yield ground-level concentrations of 1 part per billion (ppb), consistent with data from C. Junge [Tellus 9, 528 (1957)], an initial stratospheric concentration of 0.55 ppb, below the 0. 8  because of the large concentration changes over short depth intervals that result from high bacterial activity. In this report we present the results of a study of methane and dissolved sulfate in the interstitial waters of Long Island Sound sediments which suggest that significant production of methane does not begin until dissolved sulfate concentrations approach zero. Results of laboratory studies of time-dependent changes in the chemistry of anoxic marine sediments indicate that sulfate reduction and methane production are mutually exclusive metabolic processes.
Gravity cores were collected at three stations in Long Island Sound. Station TH is located approximately 2 km south of the coastal town of Guilford, Connecticut, at a water depth of approximately 7 m. Stations BS and SC are located in two shallow harbors near Guilford. The water depths at these two stations ranged from 1.5 to 4 m be-cause of tidal fluctuations. Interstitial waters were sampled wi.thout coming into contact with air by transferring sediments from sealed core liners to a filter-press type squeezer through an interlock flushed with CO2 or He (9). Dissolved methane was measured by liquid stripping techniques developed for measuring dissolved gases in seawater by Swinnerton et al. (10) and applied to interstitial water measurements by Reeburgh (11). Dissolved sulfate was measured by gravimetric analysis as BaSO4. Blank corrections for precipitation of nonsulfate material led to uncertainties of + 0.5 mmole liter-1 in dissolved sulfate concentrations.
Concentrations of methane and dissolved sulfate in the interstitial waters are plotted as a function of depth in Fig. 1. Differences in the depth of complete sulfate reduction at the three stations are probably a result of variations in sedimentation rates and the content of organic matter in the sediments. The leveling off of methane concentrations at station SC at about 1.0 mmole liter-1 is consistent with reaching saturation with respect to methane. The solubility of methane calculated from the Setchenow relation by using solubility data from Winkler (12) and Atkinson and Richards (2) ranges from 1.1 to 2.3 mmole liter-1 (25 to 51 ml liter-') for the temperature range (40 to 28°C) and salinity range (27 to 31 per mil) encountered in these sediments.
Reeburgh (4) has presented evidence of methane saturation in Chesapeake Bay sediments, which results in the formation of trapped methane bubbles that strip other dissolved gases such as N2 and Ar from interstitial waters in the sediments. We infer from the low concentrations of dissolved N2 and Ar at station SC that this process may be taking place there (13).
The data in Fig. 1 show that in the interstitial waters of Long Island Sound sediments high methane concentrations do not occur unless sulfate concentrations have been appreciably lowered.
Only where dissolved sulfate concentrations approach zero do concentrations of methane attain saturation. Four alternative hypotheses can be used to explain these results. The first is that methane is produced at roughly the same rate throughout the sediment column by methane bacteria but is consumed by sulfate-reducing bacteria through the reaction CH4 + S042-+ 2HW H2S + C02 + 2H20 1167 Methane Production in the Interstitial Waters of

Sulfate-Depleted Marine Sediments
Abstract. Methane in the interstitial waters of anoxic Long Island Sound sediments does not reach appreciable concentrations until about 90 percent of seawater sulfate is removed by sulfate-reducing bacteria. This is in agreement with laboratory studies of anoxic marine sediments sealed in jars, which indicate that methane production does not occur until dissolved sulfate is totally exhausted. Upward diflusion of methane or its production in sulfate-free microenvironments, or both, can explain the observed coexistence of measurable concentrations of methane and sulfate in the upper portions of anoxic sediments.