Fig. 1. European moth species named Biston betularia adapted to pollution, in the Birmingham industrial district of England, by changing from dominantly light colored to dominantly dark in less than 100 years. Dark form, known appropriately as carbonaria, is safer from preying birds when it rests on soot-blackened trees; thus the gene for dark coloring has become more abundant in the population. Man's problems with polluted air are not so simple, nor so easily solved.

Fig. 2. Limestone outside facing of the National Gallery in London (below) shows typical results of accelerated weathering caused by high concentration of SO, in urban air. SO, is oxidized and hydrated to sulfuric acid-H,SO,-which attacks limestone (CaCO,) and converts it into the hydrous mineral gypsum (CaSO2H,O). This reaction helps break up original surface and gypsum's greater solubility in rain water finishes the destructive job. The same difficulties, except for the last one, are encountered indoors, as in the Arena Chapel in Padua, Italy, where "Mission of Gabriel" fresco by Giotto shows tan powdery efflorescence of gypsum crystals formed by H,SO, attack on lime in the underlying plaster.

markable tonnages in the air over cities large and small. Add to these, uncomfortably high amounts of carbon monoxide from auto exhausts, and chlorides, fluorides, and ammonia from diverse sources and you begin to get the picture. Finally toss into the pot formaldehyde and other aldehydes, acrolein, and an array of intensely oxidizing substances including deadly ozone-which are just a few of the possible products of photochemical and other reactions among various of the contaminants and the ordinary components of air-and it becomes evident that the two problem areas that I described briefly above do only limited justice to the magnitude of the problems presented by polluted urban air.

But it is the third specific dictum of the Clean Air Act-which directs the Department of Health, Education, and Welfare, through the Public Health Service, to set up nationwide air quality criteria-that may really open the door to research opportunity.

More to quality control than meets the eye

At first glance the problem of setting such concentration criteria for various contaminants in air seems simple. And so does using them for control purposes. Once the limitstated as a concentration that ought not to be exceeded when either averaged or integrated over a certain length of time-is set, a series of control steps seem to follow logically.

First, inventory the volume of air available-again either averaged or integrated over a meaningful time interval-for diluting pollutants over an urban region or throughout a regional airshed. Then inventory the sources of pollutants in the region, as well as the sources of those pollutants coming from upwind regions in the airshed, in terms of kind and amount. At this stage don't overlook the possibility a likely one-that photochemical and ordinary reactions in both local and transported pollution clouds may make more harmful products of initially innocuous substances, and vice versa. Next determine limits for the emission of each pollutant from each source, under the worst possible dispersal conditions, so that the time-averaged or integrated total of their individual contributions to the ambient air remains below allowable limits. And of course develop and enforce the use of whatever changes in processes or equipment may be necessary to keep emissions from each source within indicated limits. Finally work out techniques for monitoring air in the region for conformance to the ambient air standardsand maybe even also techniques for spotting sources of trouble when monitoring shows that ambient standards are being violated.

Getting to the moon may be easier. Little more than a moment's reflection is required to appreciate some of the difficulties that develop at each stage of this more or less "ideal" solution to modern regional air pollution prob

lems. Political and economic factors are not the least of them.

Each step also presents major problems of meteorological understanding. Not so much on a scale as small as what happens to the pollution plume from a single exhaust stack, where much is known, but on a scale that permits fuller evaluation and prediction of the wind-stagnation and thermal-inversion conditions that can inhibit the ventilation of any region. Each step in control also poses largely unsolved problems and unprobed opportunities in chemical and meteorological sensing and monitoring, in atmospheric chemistry, in telemetry, and in data handling.

Controlling air quality is, in short, a systems problem of challenging magnitude in which social, political, economic, and technical factors mingle inextricably.

Controls need not-should not-wait

In many ways the situation with regard to engineering of devices and hardware, and to improving process variables, is in or can easily be put in much the best shape. While there's always room for technical improvements and lowered costs, there has long existed a formidable arsenal of scrubbers, filters, electrostatic precipitators, centrifuges, and more recently sonic agglomerators that can take most particulate matter out of industrial exhaustgas streams. And burning waste dumps and faulty incinerators are largely political problems, not technical ones. Even the more recalcitrant problems of gaseous pollutants like SO, nitrogen oxides, and hydrocarbons promise to yield to research efforts that are now being prodded into higher gear.

If these efforts fail to produce results, there are always alternatives available-such as

Fig. 4. The few substances shown in this table as being overly abundant in city air do not nearly exhaust the list. Absent are hydrocarbons by the hundreds--some of them potent carcinogens in laboratory animals-and the highly toxic gas nitrogen dioxide. Since these two react with each other under the stimulus of solar ultraviolet radiation to produce photochemical smog with its characteristic haze, eye irritation, and damage to sensitive plants-it is perhaps fortunate that pollutants help cut down on solar ultraviolet reaching street levels in the city. Of course some air bacteria that might be killed by such uv are not, but one must be grateful for small favors when living in the urbs.

atomic powered instead of fossil-fuel-powered generating stations, or, farther in the future perhaps, cars operated by batteries or fuelcells instead of internal combustion engines, or electrified mass transportation. The art of pollution-control is not so much primitive in technical means as deficient in social ones. Whether anybody has to do anything about the pollutants that their products or processes emit-and how well they must do it-depends in part on progress made farther back along the "ideal" pollution control chain. Here the first step-setting air-quality criteria-is the most crucial. It is also the most complex.

Though all the needed data are not in, most people would agree, I think, with Vernon MacKenzie, head of PHS' Division of Air Pollution, when he says that "we must... get on with the job of developing air-quality criteria and standards against a background of technical and scientific knowledge which is not now and probably never will be perfect." Engineers can recognize the validity of this approach; as professionals they live with it.

There's a worm or two in the apple

Nevertheless, in view of large gaps in existing knowledge about the chemistry of normal and polluted atmospheres, there are under

standable differences in outlook between industrial and public health people over which substances'-and which industrys'-emissions should be controlled. Many industrial people, including the most enlightened ones, remain somewhat loathe to pursue very much controlorientated research in the absence of definitive legal standards based on equally definitive criteria. Indeed, the automobile industry claims it wants such criteria precisely so it will know where it stands over the longer haul. It points to its experiences of recent years in California, the State that has led the country-perhaps the world as well-in pollution-control activities. California has more cars and a more poorly ventilated climate, in the Los Angeles area anyway, than any State in the Union. It also is richly endowed with sunshine. So it's not surprising that it leads the country in its concern about the auto component of its air pollution problem.

Auto exhausts emit two main contaminants: hydrocarbons and nitrogen oxides. The hydrocarbons come mostly from fuel that is not completely oxidized to carbon dioxide and water during engine operating cycles. The hydrocarbons are diverse in kind-compounds originally in the fuel as well as new compounds that are formed during the hightemperature combustion process. Nobody has identified all of the compounds, but there were about 200 of them at last count. Nitrogen oxides, particularly nitric oxide (NO) with lesser amounts of NO2, result from the hightemperature dissociation of molecular nitrogen and oxygen from the intake air used to burn the fuel. The liberated atomic nitrogen and oxygen then combine to yield the oxides. These reactions are reversible at high temperatures, but they are prevented from reversing the two oxides are literally "frozen in"-as combustion temperatures drop rapidly from peak values during expansion of the gases in both auto engine cylinders and powerstation exhaust stacks.

Of these two groups of contaminants California has so far established emission standards only for the hydrocarbons, chiefly because it was originally thought that the nitrogen oxides in auto (and other) exhausts would be far more difficult, and maybe unwise to eliminate. It was also felt originally that reducing the concentration of either one of the starting reactants would help reduce the buildup of troublesome final photochemical reaction products in the air.

Over the years, as California re-inventoried the dimensions of its pollution problems in finer detail, changes were made in automobile and other emission limits. The average uncontrolled automobile exhaust emits about 700 ppm hydrocarbons. Current California standards, set by the State Department of Health and approved by the legislature, limit this to 275 ppm and include an additional restriction

on carbon monoxide of 1.5% (the average exhaust emits perhaps 3), for reason of its own toxicity and its rising concentration in Los Angeles air, not because it contributes to photochemical smog. By 1970 the state will reduce these limits still further, to 180 ppm and 1 respectively. But even now, for reasons that we will discuss shortly, the Los Angeles County Air Pollution Control District-the trail-blazing agency in photochemical pollution control-would like to see the hydrocarbon limit reduced to below 100 ppm.

Such differences in opinion on the part of different control agencies is of course not frivolous in intent. It reflects uncertainty about precisely what should be controlled, and by how much. But at each such turn in the road to inevitable controls, a good deal of prior research on control techniques gets bypassed. And at the spectre of a patchwork of differing, locally determined control limits a national industry, like the auto industry in particular, gets understandably upset. To more clearly understand the basis for some of these difficulties we need a closer look at the nature of the smog which now afflicts all urban regions that are highly populated with automobile traffic.

Sun-stewed auto exhaust

smog

Photochemical smog's chief immediate and obvious symptoms are eye irritation, damage to sensitive plants (as far as 100 miles from urban centers), accelerated cracking of rubber, and decreased visibility due to the formation of a haze of solid and liquid particles which are collectively described as aerosols. The first three symptoms are caused by the products of a complex of reactions that start, as we said, with hydrocarbons and oxides of nitrogen. In the course of the reactions some highly reactive intermediates and ozone are formed. These help polymerize the organic compounds in the mixture, according to Prof. A. J. Haagen-Smit of CalTech who pioneered work in this field, and lead to the formation of additional, non-volatile, oxidation and polymerization products which add to haze and eye irritants from other sources of pollution.

This kind of pollution is characterized by the same extreme dilution of reactive constituents and intermediate and final reaction products that characterize all air pollution. An ozone concentration of 0.3 ppm for instance is 10 times its normal background concentrations at ground levels. Haagen-Smit points out that this extreme dilution is one reason why it takes many years to unravel even relatively simple atmospheric reactions such as the photodecomposition of acetone. In ordinary laboratory work, "slow" and "fast" reactions are characterized against a backdrop of reactant concentrations that average about 10%. But in atmospheric reactions concentrations are only on the order of one-millionth as much. Under these conditions reactions which

Fig. 5. High wintertime concentrations of SO, in air over (British) cities and surrounding regions come from burning of sulfur-bearing coals and add to ordinary cardio-respiratory hazards of winter weather, especially for the elderly. The map makes clear that pollution is not just a downtown problem; nor is the SO, problem exclusively a British one. Abatement efforts center on switching to low-sulfur fuels and eliminating SO, from flue gases.

would proceed in thousandths of a second at ordinary laboratory concentrations take an hour in the atmosphere.

This is important: It is essentially what permits the highly potent ozone to survive attack by reducing agents such as SO, in the atmospheric mixture. Ozone thus can go on to participate in reactions at several stages in the photochemical chain. The immense slowdown in reaction rates also gives ozone and various, highly reactive, free radicals that are formed a far better chance to survive long enough to play significant roles, not only in the over-all reactions but in some of the observed symptoms as well.

The still far from completely elucidated chain of daily reactions leading to smog is set in motion by the photochemical dissociation of nitrogen dioxide. It is a yellowbrown gas that most effectively absorbs pho

tons in the blue and near ultraviolet. This dissociation produces nitric oxide and atomic oxygen NO2+ hv→ NO+O. The atomic oxygen thus formed combines with atmospheric oxygen molecules to form ozone (Og).

Other photochemical reactions that contribute to the stew form excited oxygen molecules. These and ozone, perhaps aided by atomic O present in some dynamic equilibrium concentration, attack organic materials, probably by removing hydrogen atoms from the hydrocarbons. This oxidizing assault forms reactive intermediate substances such as alkyl and acyl radicals. These radicals can unite with oxygen to form still more reactive peroxyl

radicals. And these, in turn, can react with oxygen to form more ozone, with NO to form larger quantities of chain-initiating NO, and with NO to produce short-lived but plant damaging peracylnitrates (usually abbreviated to PAN) and an abundance of more stable oxidation products such as eye-irritating formaldehyde.

Bewildered? Let's go over it again with reference to Fig. 8. The pivotal group of oxidizing substances consists of atomic oxygen (from dissociation of NO), excited molecular oxygen (from solar irradiation of the atmosphere's abundant molecular oxygen), peroxyl radicals (from the action of other oxidants on hydrocarbons), and ozone (formed as a byproduct in several of the photochemical reactions). During the daylight hours all of the oxidants contrive to react with the original starting materials-hydrocarbons (particularly unsaturated, olefinic ones and some aromatics) and nitrogen oxides-as well as with their reaction products. Thus at any particular time, the air is filled (relatively speaking-remember the troublesome concentrations are down at or below the part per million level with a very complex mixture of intermediate oxidation and reaction products. Unquiet controversy in California

A vast amount of experimental and theoretical photochemistry remains to be done at the low concentrations and low temperatures which characterize polluted atmospheres before we completely understand these processes. In the laboratory (Fig. 9, 10) symptoms of photochemical smog can be produced by irradiating with mock sunlight suitably low concentrations of hydrocarbons in the presence of oxides of nitrogen. But considerable controversy exists among both atmospheric chemists and simulators of smog about the precise course and time constants of each of the innumerable reactions occurring in polluted atmospheres. Not all such disputes are academic; indeed one such controversy is particularly instructive. It illustrates in a simple way the difficulty of understanding what our pollution problems really are and casts a long shadow over approaches to controlling photochemical pollution.

Although nitrogen oxides and hydrocarbons are the essential starting ingredients, only hydrocarbons are now being controlled in California, as was mentioned. This also is the control approach called for in a recent bill proposed to Congress and aimed specifically at the motor vehicle pollution problem. It should work; according to the tenets of the familiar chemical law of mass action this should inhibit formation of even the minute concentrations of final photochemical reaction products. Indeed, the consensus among pollution experts in Washington, Detroit and California is that by reducing hydrocarbons enough now,