Tobacco Documents Online

307 IV. DETERMINATION OF NITRIC OXIDE AND NITROGEN ~ DI®XIDE IN CIGARETTE SMOKE BY ~ CHEMISLUMINESCENT ANALYSIS R. A. Jenkins and B. E. Gill Introduction. As a result of S&HP management's interest in cofactor inhalation bioassays, in which animals are exposed to cigarette smoke which has been enriched with some specific constituent, we were asked to develop an analytical method for the determination of small amounts of nitrogen dioxide in inhalation exposure atmospheres. Two major difficulties had to be addressed if the analytical method were to be successful. The rate of formation of N02 from NO and 02 in cigarette smoke is fairly rapid. Because of the differences in biolgoical impact between NO and NO2, it is important that the analytical method be capable of distinguishing between NO, present in high concentrations, and N02, present in low concentrations, and respond rapidly. Secondly, because of the complex nature of tobacco smoke, there are many potential interferences with the procedure. The chemiluminescent method of analysis was chosen because of its speed, sensitivity, and discrimination capability. Well after the method development was underway, NCI decided not to pursue the N02 addition bioassays at this time. Thus, most of the effort has been limited to the development of direct analysis of oxides of nitrogen in fresh cigarette smoke, which is more limited in scope than the original task of monitoring enriched exposure atmospheres. However, we believe that methods developed within this more limited scope will be directly applicable to bioassay monitoring, should the need arise. Methods. Cherniluminescept approaches to measurement of oxides of nitro- gen have been spurred by the Environmental Protection Agency's interest in measuring atmospheric pollution. Briefly, in a controlled reaction charnber, ozone is caused to react with nitric oxide (NO), a product of that reaction

308 being N02*, which is in an excited electronic state (about 10% yield). As the N02* relaxes to a ground state,-it emits a photon, which is registered on a photomultiplier tube. Total oxides of nitrogen (NOx) are measured by first flowing the gas sample through a reductive chamber, which reduces N0, to NO. Thus, the amount of NOL is taken as the difference in system response between the NO and NOx "modes" of the analyzer. There are many commercial chemiluminescent analyzers available today. Most differ in the nature of the reductive converter which they employ in the NOx mode. One of the major problems with this analytical method is its suscepti- bility to quenching. Quenching occurs when the N02* collides with another molecule and relaxes to a ground or near ground state without emitting a photon. Several species, including C01, HzO, CO, H2, etc., have been re- ported as having some quenching effect. This was of special concern to us, because of the high concentrations of CO (5%0) and COz (10%) in most fresh cigarette smokes. We have conducted thorough studies on these potential quenchers. Using a dynamic dilution procedure, in which certified standard mixtures of NO and N02 in nitrogen are mixed with appropriate concentrations of potential quenching agents, we found quenching of chemiluminescent response to be about 3.7% for a concentration of COz at 10 volume percent. However, because the system which we have developed rapidly dilutes the smoke vapor phase before it reaches the analyzer, effective quenching by C02 in smoke is less than 1%. Hydrogen, methane, and carbon monoxide have not been found to be important quenching agents under the present system configuration. First attempts to collect the smoke vapor phase in a gas sampling bag f for subsequent analysis proved unsatisfactory. Apparently, NO and NOz undergo a rapid reaction with some of the gas phase constituents with cigarette smoke. 89737864

309 It was not possible to obtain a constant response of the analyzer to a single smoke sample for even a few minutes. Response to NO/NO2 standard mixtures (ti 700 ppm in nitrogen) in the same sampling bags were constant over several tens of minutes. Because of the need to perform a puff-by-puff analysis of the oxides of nitrogen, the following system has evolved. The gaseous effluent from a single port smoking machine is introduced as a bolus into a large bore stand tube, one end of which is enclosed with a one-way (in) check valve. The analyzer, with its built-in pump, continually sweeps the contents of the stand tube. Thus, because it is rapidly mixed with in- coming air, the smoke is analyzed as a dilute puff. Dilution of the smoke in the stand tube acts to reduce artifactual conversion of NO to NO2. The integrated voltage output of the analyzer is related to the absolute amount of NO and/or N0z introduced into the stand tube. The analyzer is designed to actually process a fixed gas volume per unit time (about 300 ml/minute), with the remainder of the flow being bypassed around the reaction chamber. Thus, as flow into the analyzer increases, response to a given mass of NOx decreases because the fixed volume represents a smaller proportion of the total process stream. Since the analyzer possesses sufficient sensitivity, the input stream is kept at a relatively high flow rate (about 2.5-3.0 1/min). This reduces aging of the smoke and acts to dilute it more rapidly. Under present conditions, the smoke puff clears the analyzer completely in less than 10 seconds. Because the system response changes with analyzer flow rate, it is essential to standardize the response in a manner identical to the smoke analysis. Thus, "puffs" of gas standards are introduced into the stand tube I from a continually replenished atmospheric pressure source.