Here we

The recommendations of the workshop are given in detail following each session. Summarize the major themes of the workshop and list the most important areas requiring additional experimental or theoretical work.

Major advances in the chemistry of the troposphere depend on understanding the chemistry of large molecules and free radicals. Real atmos

pheres contain large amounts of hydrocarbons C1 and greater, aromatic compounds, natural products such as terpenes, and large aldehydes, ketones, phenols, etc, which are their photooxidation products, as well as a large class of oxygen containing free radicals which are the intermediates in these photooxidation reactions. These complex molecules are not only involved in the NO-NO conversion process, but almost certainly are important precursors to atmospheric aerosols. We will not understand either oxidant or aerosol formation until we attack the problem of large molecules.

Clearly, moving away from fairly simple Surrogate reactants such as propylene, and considering the whole range of atmospheric pollutants creates a problem in scale. There are far too many molecules and potential reactions to measure everything. A proper attack on this problem involves a judicious mix of experiment and theory. We may illustrate this by considering one of the most pressing problems - the chemistry of alkoxy radicals. Large alkoxy radicals can isomerize, decompose, or react with oxygen (as well as NO and SO, see below). The relative rates of these processes must be known. Even though this problem was extensively considered at this workshop it is doubtful if a convincing solution was given. Measurements are needed to provide a base set to allow for the development of theoretical estimation schemes.

Another class of reactions of great importance both in the atmosphere and in laboratory investigations are alkyl peroxy radical reactions. In the absence of NO, these radicals react with themselves to produce aldehydes and alcohols. There is considerable uncertainty as to the mechanism of these reactions. The suggestion made at the meeting that one of their products might be the Criegee intermediate emphasizes the need for much more work in this area. At the same time the peroxy radicals formed in the reactions of OH with olefins in the presence of O2 need identifying.

In addition we need data on acetyl and acetylperoxy type radicals. There are questions as to formation of acids, particularly from formyl radical reactions, which cannot be explained on the basis of existing data.

Another major deficiency is in the area of

reactions of aromatic compounds. Recognition of their importance is fairly recent. In particular we need to know about rates and mechanisms of reaction of aromatics with OH radicals. This will involve extension of existing experimental approaches and development of new ones. The branching ratios for different products need to be measured and the subsequent chemistry of these products needs to be considered.

The possibility of making the Criegee intermediate from alkyl peroxy radicals was noted above. The Criegee intermediate is presumably a primary product of an ozone-olefin reaction. Its subsequent fate is of great importance. A crucial question is whether it decomposes or is stabilized, and if stabilized what chemical reactions it can undergo. There is some evidence that small Criegee intermediates decompose. For the large ones however, there is very little quantitative data. This question needs resolution since the Criegee intermediate has been postulated to be a potential oxidizer for NO, SO2, olefins, etc.

In addition to treating free radical reactions in terms of isomerization, scission, self-reaction, and reaction with 02, we must consider reactions with NO and SO. Here we are faced with problems of rates and mechanisms and in particular the problem of the role of association reactions. For peroxy radicals the starting point is the HO2-NO reaction. The new value for the rate constant has had a dramatic effect on the models. It needs to be studied over a wide range of conditions (temperature, pressure) to confirm this value under atmospheric conditions. The use of the rate constant data for HO2 + NO for RO2 + NO reactions may be invalid. Direct measurements are needed. In addition, for large alkylperoxy radicals we need to know if alkyl nitrates are products since this is a chain terminating reaction.

Similar considerations apply in the case of the reactions of alkylperoxy radicals with NO2, although it is not likely that the peroxynitrates formed in simple association reactions would have a significant lifetime in the atmosphere. However, if the reaction can lead to an aldehyde and nitric acid it could be of considerable importance.

Reactions of alkoxy radicals with NO and NO2 can also proceed via channels leading to adducts or to HNO or HONO respectively. The overall rate constants and branching ratios need to be determined. Similar considerations apply to OH reactions with NO and NO2.

A somewhat different approach to the question of the importance of adduct formation is to consider the thermal stability of peroxy radicals. Work of this kind has been done for some PAN

Although we now have a qualitative idea of the mechanisms of reactions of olefins in the atmosphere there are still many areas in which quantitative detail is needed. There is a need for rate data on reactions of hydroxyl radicals and ozone with cyclo-olefins and natural products (isoprene, terpenes, etc). The rate of reaction of OH with ethylene should be studied at higher pressure where the possibility exists of the hot C2H4OH* adduct reacting with 02. Above all the questions of ozone reaction mechanisms remains unresolved, particularly for large olefins, cyclo-olefins, terpenes, etc. Studies under atmospheric conditions are needed. Also we need to know under what conditions OH abstraction vs. addition become important.

The problems of heterogeneous chemical kinetics and aerosol formation came up repeatedly during the meeting. Surface effects in smog chambers include free radical initiation of chamber reactions, and absorption and desorption of reactive species. The role of wall effects in chamber studies will have to be resolved before chamber studies can be used to validate complex chemical kinetic models.

Aerosol formation initiated by reactions of large organic molecules involves aspects of both homogeneous and heterogeneous chemical kinetics. There is a whole range of problems which need study. The role of hydration of free radicals was touched on many times during the meeting. We do not know which radicals (if any) are truly hydrated, and what are the kinetic consequences of hydration. Are condensation nuclei formed in free radical reactions? Thest problems are not only of great interest in modeling atmospheric chemistry, but are

of great importance in the design and execution of laboratory studies of the elementary chemical kinetics to be used in modeling studies.

Certainly a principal objective of homogeneous chemical kinetics should be a fundamental understanding of the initial reactions leading to the formation of atmospheric aerosols.

Finally, since this workshop was directed toward kinetics data needs for modeling the troposphere, it is appropriate to include a genera comment on the meeting by R. J. Cvetanovic:

"Adequate understanding of the chemistry of photochemical smog will be made possible only through comprehensive modeling of the chemical processes which occur in the polluted troposphere. The success of such modeling will depend very critically on the availability of as complete a list as possible of the elementary chemical reactions likely to be involved and of reliable values of their rate constants under tropospheric conditions. Incomplete or unreliable information could lead to erroneous conclusions and result in ultimately very costly misinterpretation of the pollution. Accumulation and continuous updating nature of the problems posed by tropospheric of the necessary information will require the following steps: 1) establishment of a comprehensi list of chemical reactions potentially involved in tropospheric chemistry, preferably in the form of a reaction grid of the type first used in the Climatic Impact Assessment Program (CIAP) for modeling the stratosphere; 2) critical selection of the most reliable values of the rate constants of these reactions with estimates of their limits of uncertainty; 3) initiation of measurements of the rate constants which are at present not available; 4) updating the rate data through continuous monitoring of the new values which become available; 5) initiating any studies necessary for improved understanding of reaction mechanisms when this information is not available or is uncertain".

The major recommendations of the workshop, in approximate order of priority are as follows:

Recommendations:

1. The rate constants for isomerization, scission and reaction with oxygen, of a base set of alkoxy radical reactions should be measured. This initial set should include methoxy, ethoxy, propoxy, n-, s-, t-butoxy, and hydroxy-butoxy. Both absolute and relative rate measurements shoul be considered.

2. Continued efforts should be devoted to improv ing existing theoretical approaches, based on RRK to calculate isomerization, and scission rates of alkoxy radicals.

3. The mechanisms and rates of self-reaction of alkylperoxy and hydroxy-alkyl peroxy radicals shou be studied. The question as to the formation of Criegee intermediates should be resolved.

4. The radicals formed in the reaction of OH wit olefins in the presence of 02 should be identifie