STRATOSPHERIC OZONE 1In the 21st Century The Chlorofluorocarbon Probkem Every September in the past few years has brought a precipitous loss of ozone in the Antarctic stratosphere, amounting to nearly 100% loss at altitudes between 15 and 20 km, and 50% or more of the total ozone integrated vertically over all altitudes. The magnitude of this loss in 1990 paralleled the losses in 1987 and 1989 and corresponded to the disappearance of about 3% of the entire global supply of stratospheric ozone in a period of 4 to 6 weeks. Ozone (0,) exists in a dynamic equilibrium in the stratosphere, balanced between formation by solar ultraviolet photolysis ( h< 242 nm) of molecular 0, (0 + 0, + 0,) and destruction by various chemical processes including several chain reaction sequences triggered by HO,, NO,, and ClO, radicals. The ozone dissipates over Antarctica by November through northward mixing, only to begin reappearing in late August of the following year. Substantial ozone losses have also appeared, although not as spectacularly as over Antarctica, in the Northern Hemisphere’s temperate and polar regions. The primary cause for the Antarctic ozone loss, and the probable cause for the northern losses, is the increasing concentration in the stratosphere of anthropogenic chlorine, especially chlorine released by solar W photolvsis from chlorofluorocarbon (CFC) compounds such as CCl,F, (CFC-121. ~-~- ._,and -..,. CC1.F ICFC-111 CCl,FCClF, ( C k - 1 1 3 ) . Because these molecules have average atmospheric lifetimes of many decades, excess anthropogenic chlorine will persist in the stratosphere for comparable time periods, and the Ant~~~
622 Envimn. Sci. Technol., Vol. 25. No. 4. 1991
University of California Irvjne, CA 92717
The C1 atom released in Equation 1 reacts with 0, within a fraction of a second as shown in Equation 2, with the formation of the free radical chlorine oxide C10.
arctic ozone hole will be an important atmospheric phenomenon throughout the 21st century.
c1+0, -+ c10 + 0,
F. Sherwood Rowland
Stratospheric photolysis of CFCs The relationship between increasing concentrations of the chlorofluorocarbon gases and stratospheric ozone depletion was first outlined in a short paper in Nature (11, followed immediately by a comprehensive review 121. The property of chemical inertness that has been central to many technological uses of CFCs is also observed in the atmosphere, where they can survive unchanged for very long periods of time, as much as 100 years or more. The physicochemical processes that normally remove trace impurities from the atmospheretropospheric sinks such as solar photolysis, rainout, or oxidation-don’t affect the CFCs because they are transparent, insoluble, and chemically unreactive. Eventually, however, the individual CFC molecules diffuse randomly upward into the middle stratosphere and are decomposed by short-wavelength solar UV radiation, as shown in Equation l for CCl,F,.
(2)
The C10 radical can then react within minutes with atomic oxygen as shown in Equation 3, once again releasing C1.
c10 + 0 + c1+0,
(3)
The sequence of Equation 2 plus Equation 3 constitutes the C10, free radical catalytic chain reaction which repeats itself over and over again in the stratosphere before finally being stopped. 0 + 0, + 0,
+ 0,
(4)
The net effect of the two reactions as shown in Equation 4 converts back to molecular oxygen one ozone molecule and one atom of oxygen w h i c h would otherwise have formed ozone, making this C10, chain an exceptionally efficient method for ozone destruction. The release of CCl,F, (CFC-12)to the atmosphere has averaged about 400 kilotons per year for the past 15 years, while CC1,F (CFC-11)yearly emissions have been in the 250300-kiloton range over that period. Manufacture and atmospheric reCCl,F, + W (< 220 nm) + lease of the third major chlorofluoc1+CClF, (1) rocarbon, CCl,FCClF, (CFC-113), have increased sharply over the These wavelengths of UV radia- past decade and are now also -300 tion do not penetrate down to the kilotons per year. All of these CFC surface of the Earth because they are emission rates are expected to deabsorbed in the lower stratosphere crease significantly during the either by molecular 0, or by ozone. 1990s because the recognition of 0013-936W91/0925-622$02.50/0@I 1991 American Chemical Society
the cause-and-effect CFC relationship with stratospheric ozone has led to international agreement on a decade-long phaseout of CFC production. The C10, chain can be interrupted by reactions such as Equation 5 with CH, or Equation 6 with NO,, which form temporary reservoir molecules such as HC1 and chlorine nitrate (CIONO,). C1+ CH,
+ HC1+ CH,
(5)
C10 + NO, + M +ClONO, + M (6) Stratospheric chlorine can remain chemically inactive in these reservoir molecules for a few hours or days, only to be released once more as C1 by photolysis of ClONO, or HO reaction with HC1. The average
length of the C10, chain in the stratosphere before drifting back into the troposphere (usually as HC1) is about 100,000. This chain multiplication factor of io5 combines with atmospheric release every year of about one million tons of CFCs to make stratospheric ozone depletion by CFCs a major global environmental problem. Tropospheric sinks for CFCs? Substantial effort has been expended in a search for other tropospheric sinks for the CFCs. If such sinks were found, these investigations would substantially diminish the CFC threat to stratospheric ozone, but an unsuccessful search verifies that solar UV decomposition is indeed the only significant atmospheric removal process for
the CFC molecules. Can rainout, oxidation, tropospheric photolysis, some other as yet unidentified process, or a combination of all of these processes remove a significant fraction of the CFCs, and thereby reduce the eventual amounts of chlorine released in the stratosphere? Although specific tests can be devised for known processes and all such have shown the postulated sink to be negligibly important, the most complete test of the inertness of the CFCs rests upon the determination in the atmosphere itself of the lifetime of the CFC molecules. If some important tropospheric removal process were to exist for CFCs (e.g., photolysis by near-UV light while absorbed on Sahara sand, or freezing out on Antarctic snow have both been suggested), Environ. ki.Technol., Vol. 25, No. 4, 1991 623
then substantial fractions of the seYera1 CFCs would already have disappeared from the atmosphere. However, in situ experimental measurements have shown that the atmospheric concentrations of CCl,F,, CCI,F, and CCl,FCCIF, have increased very rapidly over the past 15 years. Comparison of the amounts of CFCs now in the atmosphere with the amounts already released confirm that most of the CFC molecules are still there, leading to estimated lifetimes of ahout 75 years for CFC-11,100 years for CFC113, and 120 years for CFC-12. The total organochlorine concentration of the atmosphere at the end of 1990 (Figure 1) is now approaching 4.0 parts per billion by volume (pphv), and can he compared with only 1.8 pphv C1 in 1974 and 0.8 pphv in 1950increase by a factor of 5 in only 40 years. Methyl chloride is the only natural organochlorine compound found in the atmosphere in significant quantities (ahout 0.6 ppbv, unchanged during the past 15 years), and the total atmospheric chlorine concentration in 1900 was probably about 0.6 pphv. The major contributions to C1 in the atmosphere now come from CCl,F,, 1.0 pphv C1 (0.5 ppbv CCl,F, x 2 C1 atoms per molecule); CCl,F, 0.8 pphv; CCl,FCClF,, 0.2 ppbv; CH,CCl, (methylchloroform), 0.5 ppbv; CC1, (carbon tetrachloride), 0.6 ppbv; and CH,C1 (methyl chloride), 0.6 ppbv. Ozone loss over Antarctica An intensive search for evidence of stratospheric ozone loss was initiated soon after the publication of the Rowland-Molina theory of ozone depletion by CFCs. By 1984, ozone loss in the upper stratosphere-the most sensitive altitude for loss by the C10, chain of Equation 2 plus Equation 3-had been established, but statistical searches for total ozone loss summed over all altitudes had not disclosed any significant decreases. This situation was decisively changed by the discovery with ground-based Dobson UV spectrometers of the massive losses of ozone over Antarctica (3, 41, followed quickly by confirmation with TOMS (Total Ozone Mapping Spectrometer) and SBUV (Solar Band UltraViolet) instruments on the Nimbus 7 satellite that the ozone losses had occurred above a geographic area roughly the size of Antarctica itself-1.5 times the area of the United States. An expedition to the southern po624 Envimn. Sci. Technol., VoI. 25, No. 4, 1991
Others 7
CCI,FCCIF,
7
CCI,F,
7
CCI,F
CH,CCI,
7
CCI,
I
lar region was carried out during the late austral winter and early spring of 1986, followed by two expeditions in the same seasons of 1987. These expeditions have successfully elucidated the chemical and physical processes involved in the formation of the Antarctic ozone hole each September. First, the south polar winter meteorology is dominated by the polar vortex which keeps stratospheric air trapped in the darkness over Antarctica throughout the winter and well into the following spring. The temperatures in this air fall to -85 "C and -90 OC, cold enough for precipitation of clouds even at the 4 to 6 ppmv levels of H,O in very dry stratospheric air. These polar stratospheric clouds, or PSCs, occur in two forms: initially, crystals of nitric acid trihydrate (HN03.3H,0) are formed, and later, when the air is still cooler, water ice condenses on the existing surfaces. The abundant particles in these PSCs furnish surfaces on which important heterogeneous nitrogen and chlorine chemistry occurs very rapidly in the darkness. First, N,O, formed from the simpler nitrogen oxides (NO, NO,, NO,) is converted into nitric acid and mpped in the cloud particles, leaving a gaseous air mass in which nitrogen oxides are essentially absent. Then chlorine reservoir compounds such as chlorine nitrate and HCl react with each other or with H,O to form reactive C1, or HOC1, as shown in Equations 7 and 8. These chlorine compounds escape into
CH,CI
the gas phase while the nitric acid remains in the clouds. HCl + ClONO,
+ C1, + " 0 ,
(7)
H,O t ClONO, + HOCl + "03
(81
When the first sunlight of approaching spring breaks into the polar darkness, molecules such as C1, and HOCl are photodissociated to release atomic C1, which then attacks 0, by Equation 2 to form C10. In the absence of NO,, chlorine nitrate cannot form; the concentration of C10 increases to the pphv level. Then, two C10 radicals can react to form the chlorine oxide dimer, ClOOC1, as shown in Equation 9. This molecule is broken apart by sunlight as shown in Equation 10, releasing both C1 atoms again. 2C10 + M + ClOOCl + M
(9)
ClOOCl + uv + c1+c10, --f 2c1 t 0, (10) This sequence of Equation 2 + Equation 9 +Equation 10 represents the polar C10, chain reaction mechanism, and differs from the temperate C10, chain because it sums to 2 0 , + 30,. Unlike the temperate C10, chain, which operates most efficiently in the upper stratosphere, the polar C10, chain is most effective in the altitude range coincident with the PSCs, ahout 15-20 km. The polar C10, chain does not require 0 atoms, which are scarce in the lower stratosphere, especially with the
temperatures with PSCs present during the winter of 1988-89. These C10 abundances in the polar regions show more chlorine as the active C10 radical than was present in all chemical forms in 1970, as shown in Figure 1. Detailed study of the experimental data conclusively shows that the C10, chains are the cause of the ozone depletion over Antarctica and that most of the chlorine in the C10, chains was put into the atmosphere by humans. The most convincing single experimental data set from the Antarctic missions is illustrated in Figure 2, which displays the relationship between C10 and 0, found on the August 23 and September 21,1987, flights of the high-altitude ER-2 aircraft operating from a base in Punta Arenas, Chile (5). As the aircraft flew south over Antarctica at an altitude of 18 km, very high C10 concentrations were observed within the polar vortex on both days (and on all other flights of this series). The ozone levels inside and outside the polar vortex were approximately equal on August 23, but the inside concentrations had been reduced on September 21 to only about 20% of the levels exhibited outside the vortex. August 23 is late winter in the Southern Hemisphere, and at these latitudes, sunlight has only ended the winter darkness a few days earlier. Eighty percent of the ozone initially present inside the vortex at this altitude disappeared in the one month that elapsed between flights.
IIIII" ' l " ' l ' " l ' " l
September 21,1987
93
I
long atmospheric paths for solar radiation with the sun barely above the horizon. The polar ClO, chain continues to operate effectively until the ozone at that altitude is completely depleted, or until the brighter sunlight of mid-spring causes the PSCs to evaporate. As the clouds disappear, "0, is set free and soon reacts to
release NO,. This, in turn, can combine with C10 to form the reservoir molecule chlorine nitrate and put an end to the rnnaway chain reaction of September and early October. Concentrations of C10 as high as 1.3 pphv have been measured over Antarctica, and comparably high C10 concentrations were also found over the Arctic in regions of low
Global ozone losses During 1986-1988, the historical ground-based records of total ozone were reexamined by the NASA Ozone Trends Panel and were found to show distinct seasonal differences not allowed for in the previous statistical analyses (6-8).The computational modeling of ozone depletion developed and applied prior to observation of the Antarctic ozone hole had not made any allowance for heterogeneous reactions or other special polar effects, and had not provided any reason for expecting seasonal variations in stratospheric ozone losses. The revised statistical procedures applied by the Ozone Trends Panel emphasized a search for possible ozone depletion on a month-by-month basis (e.g., losses might occur in January but not in July]. When analyzed in this manner, the data showed that ozone losses had already occurred in the broad Environ. Sci. Technol.. VoI. 25, No. 4, 1991 625
latitude bands from 30 ON to 64 ON, with losses of about 2% since 1970 averaged over the entire year (Figure 3). The losses were shown to have been heaviest in the fivemonth period from November through March, and progressively greater as one moved poleward from 30 ON. The probability is high that much of this ozone loss originated through chemical mecbanisms similar to those operating over Antarctica. These statistical procedures also evaluated and allowed for the effects on ozone concentrations at each measuring station attributable to the ll-year cycle of the solar magnetic field, the periodic oscillations in stratospheric wind directions (a variable cycle averaging about 27 months a n d known as the quasibiennial oscillation or QBO), and atmospheric testing of nuclear weapons three decades ago. An insufficient number of ground stations exists with 25 years or more of high-quality ozone measurements to allow estimates of the ozone changes in other latitude bands in the tropics or in the Southern Hemisphere. However, the satellite instruments on Nimbus 7 can be calibrated for the long-term drift in experimental sensitivity as they pass over the Dobson instruments at ground stations in the north temper-
628 Emiron. Sci. Technol., Vol. 25,No. 4, 199i
ate regions, and then used to extrapolate changes since 1978 in the other latitude regions. These calibrated satellite data indicate losses of ozone at essentially all latitudes over the past decade. Trace gases, greenhouse effect The chlorofluorocarbons are not the only trace gases whose concentrations are increasing in the atmosphere. Carbon dioxide has been regularly measured for more than 30 years, and the concentration pattern exhibits a yearly cycle superimposed on a steady growth, as shown in Figure 4. The amount of carbon dioxide decreases every spring and summer in the Northern Hemisphere as it is photosynthesized into growing green plants, only to retun to a higher concenbation the following autumn and winter. This 10% increase in 30 years in the total amount of carbon dioxide in the atmosphere is chiefly the result of the burning of carbon-based fossil fuels, that is, coal, oil, and natural gas. Trace impurities such as SO, are also released during the burning of fossil fuels, but even clean fuels always give carbon dioxide. Additional carbon dioxide is released during the burning of tropical forests. The atmospheric concentration of methane is also growing. Our mea-
surements since 1978 show (Figure 5)that the amount of methane in the atmosphere has risen steadily from 1.52 to 1.71partspermillionbyvolume, 13% in 12 years. The most abundant sources for atmospheric methane involve anaerobic biology-microbial action in the absence of oxygen or air. For example, methane is emitted from swamps and flooded rice paddies, and the stomachs of cattle give it off in large quantities. However, some CH, is also given off from fossil fuel applications. “Natural gas” is itself primarily methane. Furthermore, nitrous oxide (N,O) is increasing 0.2% per year, and the quantity of ozone near the Earth’s surface is rising as well. These increases in carbon dioxide, methane, the CFCs, and other gases are the basis for the atmospheric concerns described as the greenhouse effect. Incoming solar energy arrives primarily as visible r a d i a t i o n , a n d a n equivalent amount of energy must escape from the atmosphere in order to keep the Earth in overall heat balance. The peak intensities for visible radiation occur at wavelengths around 500 nanometers, in response to the solar surface temperature of about 5600 K. With an average Earth surface temperature near 288 K, the peak intensity of escaping radiation will
be emitted at wavelengths about 20 times (Le., 5600/268)longer than 500 nm, or 10,000 nm. These wavelengths are in the infrared (IR) region and are usually ex ressed as 10 microns rather than 10? nm. The internal vibrational. motions of triatomic molecules have frequencies falling in this wavelength band with the consequence that naturally occurring molecules such as H,O,O,, and CO, are capable of absorbing this outgoing IR radiation, reemitting it in all directions. With such IR escape to space partially hindered, Earth's atmosphere warms up and increases total IR emission to the extent needed to force enough additional IR out through the transparent wavelengths to maintain the needed energy balance with incoming solar radiation. Estimates have been made that Earth's natural atmosphere is warmer by about 35 O C than a hypothetical atmosphere free of such IR-absorbing molecules. The growing atmospheric concentrations of molecules such as carbon dioxide, methane, and the CFCs are also able to absorb this infrared radiation, providing an incremental hindrance to IR escape directly into space. The atmosphere will respond to this increased IR forcing by the accumulating trace gases, but the quantitative response is difficult to
I 6 ppbvlyear 1978-1990
I estimate. Most climatologists expect an increase in the average global temperature as one consequence of this trace gas perturbation, with estimated increases of 3.0 f 1.5 "C by the year 2050. A major uncertainty in these estimates is connected with possible changes in cloudiness. An increase in
horizontal area obscured by clouds would increase the visible albedo, reflecting more incoming solar energy back into space, and reducing the amount of sunlight available for photosynthesis. An increase in the vertical extent of clouds would, however, reduce IR escape (cloud tops are cooler at higher altitudes) and enhance the Environ. Sci. Technol.. Vol. 25, No. 4, 1991 627
need for higher surface temperatures to maintain an energy balance. Actual measurements have shown that the average temperature of the Earth‘s atmosphere over the past 100 years has risen by more than 0.5 “C. The current yearly CO, concentration increase of 1.5 ppmv is approximately 100times larger thanfor C”,at 16 ppbv, and 75,000times greater than CCI,F, at -20 pptv. However, these incremental CO, molecules absorb precisely the same wavelengths absorbed by all of the GO, molecules already existing in the atmosphere, greatly reducing the incremental GO, IR absorption efficiency. Because the vibrations corresponding to the stretching of 1and C F bonds fall in transparent wavelength bands, additions of molecules such as CCl,F, and CC1,F are about 15,000 times as efficient in IR absorption as incremental additions of CO,. Incremental methane is about 30 times more efficient than CO,. For this reason, the accumulations of the other trace gases had an incremental greenhouse forcing effect in the 1980s approximately equivalent to that from CO,.
International regulation of CFCs An international protocol to control the emissions of CFCs to the atmosphere was agreed upon in Montreal in September 1987. Under its terms, the total global emissions of CFCs were to be reduced 20% by 1994 (below 1986 emission levels) and an additional 30% by 1999. This agreement was strengthened at the London meeting in June 1990, and now calls for an essentially complete phaseout of CFC production and release by the year 2000. Many of the major CFC-producing counhies are independently committed to an even faster phaseout schedule for CFCs. Nevertheless, the amounts of CFCs emitted to the atmosphere during the fiveyears fmm 1985 through 1989 exceeded the emissions in any preceding five-year period, and measurements of the CFCs in the atmosphere do not yet show any slackening in the rate of increase in their atmospheric concentrations.
Consequences of ozone depletion The primary consequences of ozone depletion can be described under the categories of climatic and biological. The actual existence of the stratosphere, with temperatures increasing with altitude, is caused by the absorption of solar ultraviolet radiation in the ozone layer. With less ozone in the upper stratosphere, as already observed, the temperature in the 40-50-km altitude range can be expected to decrease and, indeed, satellite measurements have shown a globally averaged decline near 50 km of 1.7 f 1.0 OC in the past decade. With altered stratospheric temperatures and gradients, wind patterns could also eventually be affected. Ultraviolet wavelengths between 293 nm and 320 nm are only partially absorbed by stratospheric ozone, and some of it penetrates to the surface of the Earth where it can be absorbed by the various biological species. With humans, the major consequences of this W - B radiation are increased incidence of skin cancer, increased numbers of eye cataracts,and probably partial suppression of the immune system. Many other biological species have been demonstrated to be susceptible to UV-B radiation, but the effects are well-studied and well-documented in relatively few nonhuman species. 628 Envimn. Sci. Technol., Vol. 25, No. 4. 1991
CFC substitutes Replacements for CFCs in most of their uses are being developed very rapidly. The major uses in 1988 included aerosol propellant (CFC-11 plus CFC-12), plastic foam blowing (either CFC-11 or CFC-121, cleaning of electronics (CFC-113),refrigerant (CFC-12),and the like. Some of the replacements can be described as “not-in-kind,” as with hydrocarbons for CFCs or a manual air pump for an aerosol. Conservation is an-
other important opportunity for saving on CFC emissions. Finally, several new compounds from the HCFC and HFC classes are being produced. These molecules are different from CFCs because they have C-H bonds that can react with HO in the troposphere, as shown with CHClF, (HCFC-22) in Equation 11. This compound is already in wide usage as the refrigerant in all home air conditioners. HO + CHCIF, + H,O CCIF,
+ (11)
Other new compounds that will soon be available are HCFC-123 (CHCI,CF,) for CFC-11 and HFC134A (CH,FCF,) for CFC-12 in some applications. Long life for CFCs and ozone depletion Because about 10% of the CFCs are employed in usages that involve slow emission to the atmosphere, the maximum concentration of CFCs in the troposphere can be expected about a decade after the completion of the phaseout process (Le., about the year 2010 under the terms of the London revision of the Montreal Protocol). The maximum stratospheric effect on ozone can be anticipated to arrive an additional 5 or 10 years into the future because of the delay from release at the surface to decomposition in the upper stratosphere, or about the years 2015 to 2020. With the very long observed atmospheric lifetimes for CFCs, stratospheric ozone depletion and the effects of this ozone loss will be felt on a global scale throughout the 21st century. References (1) (2)
(3)
F. Shenvood Rowlond, Donald Bren Professor of Chemistry of the University of Californio, Imine, was first chair of that institution’s Department of Chemistry. He earned his B.S degree from Ohio Wesleyan University and his M.S. degree and Ph.D. from the University of Chicago. He became President-elect of the American Association for the Advancement of Science (AAAS) in February 1991. Rowland is a specialist in atm o s p h e r i c c h e m i s t r y a n d , with colleague Mario Molina. was the first to worn that CFCs released into the atmosphere were depleting the ozone layer. He has also been investigoting the impact of methane gas on the atmosphere.
(4) (5)
(6)
(7)
(8)
Molina, M. I.; Rowland, F. S. Nature 1974,249.81&12. Rowland, F. S.; Molina, M. I. Rev. Geophys. Space Phys. 1975.13.1-35, Chubachi, S. In Atmospheric Ozone; Reidel. D., Ed.; Dordrecht. Holland, 1984. pp. 28549. Farman. J. C.; Gardiner, B. G.; Shanklin, J. D. Nature 1985.315.207-10. Anderson,J. G.; Brune, W. H.:Proffitt. M. H. I. Geophys. Res. 1989, 94, 11.465-79. Rowland. F. S. et al. In Ozone in the Atm0sphere;Bojkov. R. D.: Fabian. P., Eds.; Deepak Publishing: Hampton, VA, 1989. NASA Ozone Trend Panel, 2 vols.; US. Government Printing Office: Washington, DC, 1990. World Meteorological Organization. ”Scientific Assessment of Stratospheric Ozone: 1 9 8 9 Global Ozone Research and Monitoring Project; Report No. 20: 1990. Vol. 1.
