Astm stp 598 1976

CALIBRATION IN AIR MONITORING A symposium presented at the University of Colorado at Boulder, Colo 5-7 Aug 1975 AMERICAN SOCIETY FOR TESTING AND MATERIALS ASTM SPECIAL TECHNICAL PUBLICATION 598 R L Chapman, symposium chairman D C Sheesley, symposium cochairman List price $33.00 04-598000-17 AMERICAN SOCIETY FOR TESTING AND MATERIALS 1916 Race Street, Philadelphia, Pa 19103 (~) by AMERICAN SOCIETY FOR TESTING AND MATERIALS 1976 Library of Congress Catalog Card Number: 75-39445 NOTE The Society is not responsible, as a body, for the statements and opinions advanced in this publication Printed in Luthervillo-Timonium, Md May 1976 Foreword The Symposium on Calibration in Air Monitoring was presented at the University of Colorado at Boulder, Colo., 5-7 Aug., 1975 The symposium was sponsored by The American Society for Testing and Materials through its Committee D-22 on Methods of Sampling and Analysis of Atmospheres R L Chapman, Beckman Instruments, Inc., presided as symposium chairman D C Sheesley, Ambient Analysis, Inc., presided as symposium eochairman Related ASTM Publications Instrumentation for Monitoring Air Quality, STP 555 (1974), $15.25 (04-555000-17) Measurement of Lead in the Atmosphere; Sampling Stacks for Particulates; and Determination of Oxides of Nitrogen in Combustion Products, DS 55-$5, $6, $8 (1975), $18.00 (05-055099-17) A Note of Appreciation to Reviewers This publication is made possible by the authors and, also, the unheralded efforts of the reviewers This body of technical experts whose dedication, sacrifice of time and effort, and collective wisdom in reviewing the papers must be acknowledged The quality level of ASTM publications is a direct function of their respected opinions On behalf of ASTM we acknowledge with appreciation their contribution A S T M Committee on Publications Editorial Staff Jane B Wheeler, Managing Editor Helen M Hoersch, Associate Editor Charlotte E DeFranco, Senior Assistant Editor Ellen J McGlinchey, Assistant Editor Contents Introduction PHILOSOPHY/PoLICY OF CALIBRATION Calibration of Stack Gas Instrumentation R L CHAPMAN Monitoring Systems Sites and Installation Notes Test Procedures Results Discussion 13 14 Gas Turbine Emission Measurement Instrument Calibration R U DIECK Statistical Framework Calibration Standards Pratt and Whitney Aircraft Emission Instrumentation Documentation of Instrumentation Performance Establishing Instrument Calibration Curves Instrumentation Precision Variability Uses of Accuracy Data Conclusions 16 17 22 23 24 27 30 34 37 Preparation of Stable Pollution Gas Standards Using Treated Aluminum Cylinders s G WECHTER Effects of Preconditioning as a Means of Deactivating Cylinder Walls Experimentation with Cylinder Wall Materials Evaluation Program for Treated Aluminum Cylinders Conclusions 40 43 45 46 53 Application and Description of a Portable Calibration System F J DEBBRECHT AND E M NEEL Methods of Standard Preparation Description of a Portable Calibration System Performance Discussion Conclusion 55 56 58 64 65 QUALIFICATIONOF REFERENCEMETHODSAS FIDUCIARYSTANDARDS Qualification of Ambient Methods as Reference Methods J B CLEMENTS Reference Methods Equivalent Methods 69 70 71 Method Evaluation and Standardization Nitrogen Dioxide Methodology Original Reference Method Selection of Candidate Methods to Replace Original Reference Method Single Laboratory Evaluation of Candidate Methods Interlaboratory Collaborative Tests of Candidate Methods Special Study Other Reference Methods 71 72 72 73 74 75 76 76 Qualification of Source Test Methods as Reference Methods J E HOWES,JR., AND R N PESUT Role of Collaborative Testing in Establishing Reference Methods Elements of the Collaborative Test Application of Collaborative Test Data 80 82 83 92 Certification Experience with Extractive Emission Monitoring Systems-W L BONAMAND W F FULLER Method Certification Procedure Results Effect of Procedures on Certification Results Effect of Reference Method Accuracy and Sampling Errors Summary and Conclusions 96 97 100 102 104 105 Verification of I n Situ Source Emission Analyzer Data H c LORD System Description Operation Calibration Comparison Data Summary 107 107 108 109 111 117 Statistical Implications of the Environmental Protection Agency Procedure for Evaluating the Accuracy of Sulfur Dioxide and Nitrogen Oxide Monitors of Stationary Sources J B REEVES Relationship Between Errors in the Reference Measurements and Errors in the Monitoring System Measurements Probability of Accepting or Rejecting a Monitoring System Probability of Rejecting NOx and SO2 Monitoring Systems Influence of Errors in the Reference Method on the Probability of Rejection Results of the Evaluation of SOs and NOx Monitoring Systems at an Electric Power Station 118 119 121 123 125 126 A R E REFERENCE METHODS ABSOLUTE Interagency Comparison of Iodometric Methods for Ozone Determination-n DeMORE, J C ROMANOVSKY, MILTON FELDSTE1N, W J HAMMING, AND P K MUELLER W Method Results Discussion 131 132 139 140 Evaluation and Collaborative Testing of a Continuous Colorimetric Method for Measurement of Nitrogen Dioxide in Ambient Air j H MARGESON, R G FUERST, P C CONSTANT, M C SHARP, AND G W SCHEIL Experimental Procedure Discussion Conclusions 144 145 146 153 Evaluation of Data Obtained by Reference Methods J K TAYLOR Nomenclature Chemical Measurement Process Minimization of Analytical Error Intercalibration Conclusion 156 156 158 160 161 162 Calibration of Permeation and Diffusion Devices by an Absolute Pressure Method R N DIETZAND J D SMITH Experimental Results Discussion Conclusions 164 165 169 174 178 DISCUSSION OF SPECIFIC TECHNIQUES Permeation Tube Equilibration Times and Long-Term Stability D L WILLIAMS Procedure Experimental Work Conclusions Application of Gas Phase Titration in the Calibration of Nitric Oxide, Nitrogen Dioxide, and Oxygen Analyzers K A REHME GPT Dynamic Calibration System NO Cylinder Calibration Calibration of Ambient Air Monitors by GPT Verification of GPT Technique and Comparison with Other Calibration Techniques Conclusions Stability of Nitric Oxide Calibration Gas Mixtures in Compressed Gas Cylinders w G LEE AND J A PAINE Experimental Section Results and Discussion Conclusion | 83 184 184 194 198 199 200 201 205 209 210 211 214 219 ROLE OF THE NATIONAL BUREAU OF STANDARDS 1N CALIBRATION Role of the National Bureau of Standards in Calibration Problems Associated with Air Pollution Measurements -E E HUGHES Production of Gaseous SRM's Evaluation of Some Gaseous SRM's Conclusion 223 225 227 230 330 CALIBRATIONIN AIR MONITORING dispenser was calibrated before deposition of each type of solution by weighing aliquots delivered into a preweighed dish with a cover to prevent evaporation Ten aliquots were weighed for each calibration The relative standard deviations of these multiple weighings were in the range 0.1 to 0.2 percent The densities of the calibrated solutions were measured with a pycnometer which was calibrated with deionized water The same procedure was followed for preparation of standard reference strips for lead Reagent grade lead nitrate, recrystallized and dried in air at 105~ was used for preparation of calibrated solutions In Tables to 5, results are shown of chemical analyses of a batch of standard reference strips for sulfates and nitrates and another batch of strips for lead Nitrates were determined by the hydrazine reduction and diazotation method [25], sulfates by the Sulfaver 5turbidimetric method [26], and lead by atomic absorption The results for nitrates show good agreement at higher concentrations (except for series 4000), while for the lower concentrations the method is apparently not accurate enough The data for sulfates show good agreement with the standard values after the blank value is subtracted, except for the three lowest concentrations The results for lead show excellent agreement for all concentrations Particulate Standards for Sulfates, Nitrates, and Trace Metals The requirement for these standard samples is to incorporate in a particulate matrix sulfates, nitrates, and the following trace metals: arsenic, cadmium, chromium, copper, mercury, manganese, nickel, lead, palladium, platinum, selenium, vanadium, and zinc in concentration ranges similar to national annual averages for air particulates [27] These requirements can be met by spiking a particulate matrix with appropriate amounts of trace metals Preparation of Spiked Particulate Matrix Pure, fused quartz (General Electric) was reduced to particle sizes below 10 um by grinding in an agate mortar followed by an air-jet pulverizer The quartz powder was then washed with nitric and hydrochloric acids and with deionized water until all contamination, mainly, iron, chromium, and nickel abraded from the grinder, was removed The removal of contaminating metals was checked by energy dispersive X-ray spectrometry Particle size classification was done by sedimentation from a water suspension The particle sizes obtained were confirmed by microscopic examination This 4The analyses have been performed by the EnvironmentalProtection Agency,Quality Assurance Branch, Environmental Monitoring and Support Laboratory Permission for publication of the data is gratefullyacknowledged 5Trademark of the Hach ChemicalCo Copyright by ASTM Int'l (all rights reserved); Fri Jan 23:31:18 EST 2016 Downloaded/printed by University of Washington (University of Washington) pursuant to License Agreement No further reproductions authorized PRADZYNSKI AND RHODES ON AIR PARTICULATES 331 TABLE Standard and analyzed values for nitrate in standard reference strips Nitrate per Strip, ~g Series Standard Values Analyzed Values Difference After Blank Subtraction 1000 2000 3000 4000 5000 6000 7000 8000 9000 638.3 1636.5 157.9 962.4 0.0 94.3 1443.0 1277.7 299.6 626.3 • 20.6 1668.2 • 29.2 173.8 • 2.8 868.8 -4- 26.3 14.2 • 2.3 124.6 4- 6.7 1383.8 4-34.4 1247.2 + 25.3 327.6• 5.5 26_0 +17.5 +1.7 79.4 +14.2 +16.1 73.4 -44.7 +13.8 Difference, 7o -4.1 +1.1 +1.1 -8.2 +17.0 -5.1 -3.5 +4.6 TABLE -Standard and analyzed values of sulfate in standard reference strips Sulfate per Strip, ~g Series Standard Values 1000 2000 3000 4000 5000 6OOO 7000 8000 900O 6382.1 1537.4 3948.2 498.0 0.0 2482.2 846.0 5110.9 159.8 Analyzed Values 6831.4+ 1894.6• 4326.0 -4936.0• 345.3 • 2898.5 • 1098.0• 5557.6 • 469.3 -4- 89.9 34.3 128.6 18.6 25.0 65.1 22.4 111.8 13.2 Difference After Blank Substration 104.0 11.9 32.5 92.7 0.0 71.0 -93.3 101.4 -35.8 Difference, 1.6 0.8 0.8 18.6 2.9 -11.0 2.0 22.0 TABLE Standard and analyzed values of lead in standard reference strips Lead per Strip, ug Series Standard Values Analyzed Values Difference Difference, 7o 100 200 300 400 500 600 1262 121 1811 804 350 1242.4 121.5 1810.1 821.6 BDL a 347.8 19.6 +0.5 0.9 +17.6 1.6 0.4 0.05 2.2 -212 016 BDL = below detection limit Copyright by ASTM Int'l (all rights reserved); Fri Jan 23:31:18 EST 2016 Downloaded/printed by University of Washington (University of Washington) pursuant to License Agreement No further reproductions authorized 332 CALIBRATIONIN AIR MONITORING method was regarded as sufficiently accurate since particle size is not a critical characteristic of this standard reference sample The trace metals were divided into two groups, nonvolatile and volatile The nonvolatile ones (cadmium, chromium, copper, manganese, nickel, palladium, platinum, vanadium, and zinc) were used for spiking in the form of standard solutions of nitrates or chlorides The standard solutions were prepared by dissolving pure metals in nitric or hydrochloric acid or, in the case of platinum and palladium, in aqua regia All chemicals were reagent grade or spectral purity The matrix was first ignited at 500~ for h and cooled in a desiccator Two 45-g aliquots were weighed using an analytical balance Each of the aliquots was placed in a Teflon beaker, and a solution containing the appropriate quantities of the spiking elements was added in small portions For each of the two aliquots, different amounts of trace metals were used After each addition, the matrix was mixed thoroughly and dried in an oven After all the spiking solution was added, the material was dried and ignited in a muffle oven at 500~ for h At this temperature, all nitric and hydrochloric acids are evaporated, all nitrates are decomposed, and the nitrogen oxides volatilized Chromium may remain in the form of chloride The rest of the trace elements remain in the form of oxides The spiked matrices were homogenized by tumbling in glass jars with several ceramic balls for 24 h, and finally deposited on glass fiber filter strips as described next The elements arsenic, selenium, mercury, and lead would volatilize in the course of the treatment just described Therefore, they were deposited on the glass fiber filter strips from solutions in the same way as the sulfates and nitrates The representativity of the trace elements on a single strip was assessed by calculation of the "sampling error," C , for a given trace element, using the following approximate formula [28] for the relative standard deviation / C, = 100 x] w~ percent (4) r l W8 where = weight fraction of the trace element, Wp = weight of a single particle, and W8 = weight of the sample i'1 Assuming that the average particle diameter is #m, the average specific gravity of spiked particulate is g/cm a, the weight of particulate on a single strip is 45 mg, and the element concentration is #g/strip (the lowest value), a value of 0.75 percent for C, is obtained Selenium was eventually omitted because it was found to interfere with subsequent nitrate determinations Copyright by ASTM Int'l (all rights reserved); Fri Jan 23:31:18 EST 2016 Downloaded/printed by University of Washington (University of Washington) pursuant to License Agreement No further reproductions author PRADZYNSKI AND RHODES ON AIR PARTICULATES 333 The following tests were performed in order to check the uniformity of distribution of trace elements in the spiked matrix Five samples of each spiked matrix weighing 2.50000 g each were taken and mixed with an equal amount of cellulose powder (Whatman, chromatography quality) Pellets were pressed and subsequently analyzed using wavelength dispersive X-ray spectrometry Each sample was measured on both sides, which was equivalent to measuring ten samples, since the samples were "infinitely thick" for the characteristic X-rays of the respective elements The results of measurements on the eight elements tested are given in Table The sampling errors calculated using Eq were of the order of 0.1 percent Deposition M e t h o d The following deposition method was developed Seven glass fiber filter strips, 1.9 by 20.3 cm (0.75 by in.), are placed in a specially designed holder on top of a high volume sampler The sampler is placed on top of a cylindrical deposition device (Fig 4) facing down, and the spiked particulate is placed in a small open container on the bottom of the device, in the center A particle cloud is stirred up using a compressed air jet fixed in front of the particulate container The sampler is switched on for a predetermined time (45 s), and the particulates are deposited on the glass fiber filter strips Each strip is weighed before and after deposition to 4-50 #g using an analytical balance The deposited weight is controlled by the amount of particulate used to make the cloud The objective is to deposit 45 4- mg on each strip In production runs, the recovery was found to be in the range 46 to 58 percent, and the weight of deposits ranged from 39 to 50 mg Single batches of seven strips showed the deviations from their mean weights of 1.1 to 4.7 percent TABLE Results of homogeneity tests of spiked particulate Relative Standard Deviation, % Element and X-Ray VK~ CrK MnK~ NiK~ CuK~ ZnK~ PtL~ PdKa Total (T)~ 1.29 2.20 1.39 1.19 1.08 0.87 2.45 5.27 Statistical(S) b Instrumental (I) c Homogeneity(/-/)d 1.20 2.03 0.77 1.07 0.96 0.58 2.16 2.18 0.22 0.22 0.22 0.22 0.22 0.22 0.22 1.29 0.42 0.82 1.14 0.47 0.44 0.61 1.14 4.62 a Experimental value for 10 sample sides bCalculated from counting statistics c Determined in separate experiment a Residual, determined as follows H = ~V/ T~ S2 - 1~ Copyright by ASTM Int'l (all rights reserved); Fri Jan 23:31:18 EST 2016 Downloaded/printed by University of Washington (University of Washington) pursuant to License Agreement No further reproductions authorized 334 CALIBRATIONIN AIR MONITORING FIG -Powder deposition device for glass fiber strips: (a) general view and (b) manifold with mask for seven strip~' Use of Synthetic Standards in Intercomparison Stndies The dried solution standards have been used in at least three recent intercomparison studies Some 3000 of the standard reference strips containing sulfates and nitrates, or lead, have been employed with satisfactory results in the Environmental Protection Agency quality assurance programs Multielement dried solution deposits have been employed in two intercomparison studies designed primarily to test new X-ray spectrometric techniques for trace dement analysis of air particulates [16,29] The synthetic particulate samples have not yet been used in interlaboratory comparisons but are intended for that purpose The multielement dried solution deposits used in the preceding two intercomparisons contained and 10 elements, selected (within the constraints of chemical compatibility) to produce characteristic X-ray lines with various types and degrees of spectral interference In the first study [29], eleven participants used X-ray spectrometry and one, neutron activation In the second [16], eighteen participants used various X-ray spectrometric techniques, two used atomic absorption, one, emission spectrometry and one, neutron activation All investigators used calibration and analysis procedures of their own choice In both studies it was concluded that the multielement dried solution deposits had adequate uniformity, accuracy, and stability for such studies The estimated uniformity of both sets of standards was to percent (one relative standard deviation of the mean of the Copyright by ASTM Int'l (all rights reserved); Fri Jan 23:31:18 EST 2016 Downloaded/printed by University of Washington (University of Washington) pursuant to License Agreement No further reproductions authorized PRADZYNSKI AND RHODES ON AIR PARTICULATES 335 monitored characteristic X-ray line intensities) The long-term stability was found to be at least as good as that of the X-ray spectrometer used for monitoring samples sent out and returned, that is, about percent No significant systematic bias occurred between the amounts of each element deposited and the reported determinations, with the exception of bromine in the first set of samples Excluding a small proportion of individual results whose large errors could be explained otherwise, the average ratio between amount deposited and amount reported, for 66 element determinations in the first intercomparison, was 0.98 • 0.09 for deposits on Whatman 41 filters and 1.02 -4- 0.10 for deposits on Millipore In the second intercomparison the best data were obtained by energy dispersive X-ray spectrometry where 120 element determinations (spread between 15 laboratories and 10 elements, aluminum to gold) gave an average ratio of 1.00 • 0.11 Acknowledgments Different parts of this work were performed under the auspices of the U S Atomic Energy Commission, now the U S Energy Research and Development Administration (Contract AT-(40-1)-4205), and the Environmental Protection Agency (Contracts 68-02-1739 and 68-02-1754) We would also like to acknowledge the untiring efforts of Dr R A Susott (Columbia Scientific Industries) who performed a great deal of supporting experimental work References [1] McNesby, J R., "Standard Reference Materials for Air Pollution," Special Environmental Report No 3, WMO No 368, World Meteorological Organization, Geneva, 1974, pp 595-603 [2] Kirchhoff, W., private communication [3] Rhodes, J R., Pradzynski, A H., Hunter, C B., Payne, J S., and Lindgren, J L., Environmental Science and Technology, Vol 6, No 10, Oct 1972, pp 922-927 [4] Rhodes, J R American Laboratory, Vol 5, No 7, July 1973, pp 57-73 [5] Winslow, E H and Liebhafsky, H A., Analytical Chemistry, Vol 21, 1959, p 1338 [6] Hunter, C B and Rhodes, J R., X-Ray Spectrometry, Vol 1, No 3, July 1972, pp 107-111 [7] Rhodes, J R and Hunter, C B., X-ray Spectrometry, Vol 1, No 3, July 1972, pp 113-117 [8] Rhodes, J R., Pradzynski, A H., Sieberg, R D., and Furuta, T in Low Energy Xand Gamma-Ray Sources and Applications, C A Ziegler, Ed., Gordon and Breach, London and New York, 1971, p 317 [9] Rhodes, J R., "Recommended Procedures for the Use of Columbia Scientific Industries Thin Standards for X-Ray Fluorescence Spectrometry," Applied Research Division Internal Report No 206, Columbia Scientific Industries, Austin, Tex., May 1975 [10] Dzubay, T G and Nelson, R O in Advances in X-Ray Analysis, Vol 18, W L Pickles, C S Barrett, J B Newkirk, and C O Ruud, Eds., Plenum Press, New York, 1975, pp 619-631 [11] Pfeiffer, H G and Zemany, P D., Nature, Vol 174, 1954, p 397 [12] Felten, E J., Fankuchen, I., and Stei~mnan, J., Analytical Chemistry, Vol 31, No 11, Nov 1959, pp 1771-1775 Copyright by ASTM Int'l (all rights reserved); Fri Jan 23:31:18 EST 2016 Downloaded/printed by University of Washington (University of Washington) pursuant to License Agreement No further reproductions authorized 336 CALIBRATIONIN AIR MONITORING [13] Gunn, E L., Analytical Chemistry, Vol 33, No 7, June 1961, pp 921-926 [14] Bumsted, H E., American Industrial Hygiene Association Journal, Vol 25, 1964, p 392 [15] Faye, G H., "Standard Reference Ores and Rocks Available from the Mines Branch as of October 1973," Mines Branch Information Circular 1C 309, Department of Energy, Mines, and Resources, Mines Branch, Ottawa, Canada [16] Camp, D C., Van Lehn, A L., Rhodes, J R., and Pradzynski, A H in X-Ray Spectrometry, Vol 4, No 3, July 1975, pp 123-137 [17] Verdingh, V., Nuclear Instruments and Methods, Vol 102, 1972, pp 431-434 [18] Luke, C L., Analytica Chimica Acta, Vol 41, 1968, pp 237-250 [19] Gilfrich, J V., Burkhalter, P G., and Birks, L S., Analytical Chemistry, Vol 45, No 12, Oct 1973, pp 2002-2009 [20] Leroux, J., Staub-Reinhaltung der Luft, Vol 29, No 4, April 1969, pp 33-36 [21] Giauque, R D., Goulding, F S., Jaklevic, J M., and Pehl, R H., Analytical Chemistry, Vol 45, No 4, April 1973, p 671 [22] Rhodes, J R and Taylor, M C., "Rapid Major Element Mine Dust Analyzer," Final Report of U S Bureau of Mines Contract H0110985, Bureau of Mines File No 49-73, Columbia Scientific Industries, 1972 [23] "Quality Control Practices in Processing Air Pollution Samples," Report APTD-1132, U S Environmental Protection Agency, Washington, D C [24] lntersociety Committee, "Methods of Air Sampling and Analysis," No 501, Tentative Method of Analysis for Suspended Particulate Matter in the Atmosphere: (HighVolume Method), American Public Health Association, Washington, D C., 1972, pp 365-372 [25] Kemphoke, L et al, Water Research, Vol I, 1967, p 205 [26] Standard Methods for Examination of Water and Wastewater, 13th Ed., Method 156C, 1971 [27] "Air Quality Data for 1968," APTD-0978, U S Environmental Protection Agency, Washington, D C., Aug 1972, p 15 [28] Grant, C L and Pelton, P A in Advances in X-Ray Analysis, Vol 17, C L Grant, C S Barrett, J B Newkirk, and C O Ruud, Eds., Plenum Press, New York, 1974, p 44 [29] Camp, D C., Cooper, J A., and Rhodes, J R., X-ray Spectrometry, Vol 3, 1974, pp 47-50 D H Stedman, G Kok, R Delumyea, ~ and H H A l v o r d I Redundant Calibration of Nitric Oxide, Carbon Monoxide, Nitrogen Dioxide, and Ozone Air Pollution Monitors by Chemical and Gravimetric Techniques REFERENCE: Stedman, G E., Kok, G., Delumyea, R., and Alvord, H H., "Redundant Calibration of Nitric Oxide, Carbon Monoxide, Nitrogen Dioxide, and Ozone Air Pollution Monitors by Chemical and Gravimetric Techniques," Calibration in Air Monitoring, A S T M STP 598, American Society for Testing and Materials, 1976, pp 337-344 ABSTRACT: Calibrated nitric oxide (NO) sources are described using ammonia and nitromethane permeation followed by pyrolysis A field carbon monoxide calibrator based on nickel carbonyl as the liquid fill is discussed Two techniques for intercalibrating chemiluminescent NO and ozone (03) meters using the photolysis of nitrogen dioxide (NO2) are demonstrated, together with the feasability of a continuous check on NO, NO~, and 03 meters in ambient air, based on a photostationary state measurement KEY WORDS: calibration, air pollution, analyzing, chemiluminescence, nitric oxide, carbon monoxide, nitrogen dioxide, ozone, photochemistry The difficulty of calibrating air pollution monitors at sub-parts-permillion levels is well documented Thus, we feel that a number of techniques should be applied to one monitor to ensure reliability by redundant calibration To this end, we have developed gravimetric sources of nitric oxide (NO) and carbon monoxide (CO), and demonstrated several intercalibrations for NO, nitrogen dioxide (NO~), and ozone (03) monitors The aim of the gravimetric study was to use permeation tube technology [1-3] and quantitative chemical conversion to provide sources of such Assistant professor and research assistant, respectively, Chemistry Department, University of Michigan, Ann Arbor, Mich 48104 Assistant professor, Harvey Mudd College, Claremont, Calif 91711 Research scientist, Great Lakes Research Division, University of Michigan, Ann Arbor, Mich 48104 The italic numbers in brackets refer to the list of references appended to this paper 337 Copyright | 1976 by ASTM International www.astm.org 338 CALIBRATION IN AIR MONITORING species as the permanent gases NO and CO For NO, two suitable permeands were found which could be converted by pyrolisis to NO These were ammonia (NH3) [4] and nitromethane (CH3NO2) For CO, no suitable pyrolysable compound has yet been found Therefore, we investigated the use of nickel carbonyl as the liquid fill for a permeation tube The equilibrium is such that, at 30 to 40~ a pressure of CO of several atmospheres is present Thus, permeation of CO is expected; stable CO permeation was observed However, a nonstoichiometric process makes the weight loss less than the observed CO output; thus, this device may not function as an absolute but only as a transfer standard Intercalibration of NO, NO2, and 03 monitors by the NO + 03 titration reaction has been demonstrated previously [5] We demonstrate two alternate approaches to NO/O3 intercalibration using the photolysis of NO2 The three fastest reactions in this system are N O ~ + h ~ N O + O O + O + M ~O3+M O + N O ~ N O ~ + 02 If low concentrations of NO2 (,-~1 ppm) in air are photolyzed, then equal concentrations of NO and O2 are produced [6,7] This equality provides a direct NO/O3 intercalibration The photolysis of any concentration of NO2 makes steady-state NO and 03 concentrations If the photolysis lamp is turned off, then the stoichiometry of reaction gives equal initial decreases of NO and 03, again an intercalibration The pl{otolysis of NO~ also leads to a method for continuous field testing of NO/NO2/O3 monitors via the photostationary state equation kl[NO2] = k3[NO] [03] (1) where kl = instantaneous rate of solar photolysis of NO2 and k3 = known rate of the reaction NO + 03 * NO2 + O5 We have shown recently [8] that the photostationary state, Eq l, holds under a wide range of conditions, and that measurement of k~ can be made with available equipment [9] If Eq is rewritten k~[NO2]/k3[NO] [03] = l (2) then using a k~ detector and ambient air monitoring data for NO, NO2, and 03, continuous calculation of the left hand side should lead to a steady value of 1.0 This test can give extra reliability to NO, NO~, and 03 measurements, and deviations will indicate calibration errors in one or more detectors STEDMAN ET AL ON AIR POLLUTION MONITORS 339 Procedure NO was determined by commercial or home built chemiluminescent NO/O3 detectors [5] 03 was detected with the same detectors using excess NO and with a commercial O3/ethylene detector NO~ was measured with NO as total nitrogen oxides (NOx) using various chemical converters [10] tested to ensure 100 percent efficiency CO was determined by methanation gas chromatography [11] Air flows were measured using calibrated soap film flowmeters Commercial standard gas cylinders ~50 ppm were used as an original basis for comparison, together with several flow dilution systems based on calibrated syringe pumps for pure gas sources and large flows of air measured with a wet test meter, calibrated using a bell prover Weight losses were measured using a single analytical balance Photolysis of NO2 in the laboratory was carried out in a 2-cm inside diameter Pyrex tube, centrally located between four F40 black light (BL) tubes (General Electric) The sampling point was typically 70 cm down the tube, giving s of photolysis of 1-ppm NO2 (100-ppm NO2 diluted in air) This system and the kl detector are described elsewhere [9] Permeation tubes were made by tightening Swagelok ferrules over fluorinated ethylene propylene (FEP) Teflon tubing with glass reservoirs Best results were obtained with nylon back ferrules and brass or aluminum front ferrules Results Permeation Sources for NO Many organic nitrites decompose readily to NO We tested several compounds including ethyl and i-amyl nitrites; however, they were too unstable to be reliable permeands, apparently due to the instability of the parent compounds to thermal decomposition All tubes were kept in the dark to minimize photodecomposition For NH3, reliable permeation has already been demonstrated [2] We observed reliable NH3 permeation in the 30 to 40~ range Quantitative 100 percent conversion of NH3 at parts-per-million levels to NO could be achieved over a platinum gauze packing in a stainless steel tube, at temperatures greater than 800~ Figure shows this conversion in air at 7-ppm NH3 As an alternate, and possibly more easily pyrolysable NO source, CH3NO2 was tested In the range 30 to 40~ permeation rates were slow and variable At >100~ the permeation rate was adequate and independent of humidity It seems advantageous to work at or near the boiling point of these organic permeands We have kept CH3NO2 permeation tubes at 100~ with good weight loss rate response for up to 45 days Typical permeation rates are 3500 to 340 CALIBRATIONIN AIR MONITORING PLATINUM PACKED VYCOR TUBE X GOLD PACKED TUBE ,",, PLATINUM PACKED TUBE >- 100 o z o g0 u,l z 80 w > z 70 6O 50 I I 600 I I 800 I TEMPERATURE I 1000 I I 1200 ~ FIG Conversion of 7-ppm NH3 in air to NO as a function of temperature on various substrates 7000 ng/min (5 to 10 mg/day) with 0.5 cm of 1~ in inside diameter by ~/16-in wall FEP Teflon tubing At these permeation rates, our tubes will last a year or more The pyrolysis of CH3NOz was studied in detail Typical conditions are shown in Table Three pryolysis furnaces were tested for conversion of CH3NO~ A plain stainless steel tube reached complete conversion at 600~ a quartz tube with platinum gauze, at 720~ and the stainless steel tube with platinum gauze at 550~ Figure shows the behavior of the stainless steel tube with platinum gauze, which was the best converter In studying the pyrolysis of CH3NO~ to produce NO, the question as to whether NOz was also formed in this process came up According to Crawforth and Waddington [12], the first step in the pyrolysis of CH3NO2 is CH3NOz ~ CH3 + NO~ Since a quantitative conversion to NO was desired, we measured the temperature conversion relationship for NO2 in our apparatus The experiTABLE Typical conditions for CHsNO~ pyrolysis Air flow CHsNO2 Pressure Heated tube I/4 ID 30 cm long Residencetime in hot zone .- o 100 , - o0 n" %"% 80 uJ > O C) 60 I 3O0 400 50O I 600 I 700 TEMPERATURE ~ FIG Conversion o f 5-ppm CHsN02 in air to N O in a stainless steel furnace with platinum gauze catalyst mental conditions were the same as the ones for the CH3NO2 pyrolysis except that the NO~ came from a gas cylinder In the stainless steel converter, 100 percent conversion of ppm NO2 to NO had occurred when the temperature reached 380~ The results indicate that, in a stainless steel converter, NO~ formation would not be a problem for our system Since the converter operates well above 380~ any NO~ formed in the first stage of decomposition is converted quantitatively to NO The pyrolysis of CH3NO2 was also studied using a gas chromatograph (Varian 1520) with a flame ionization detector Since this detector is sensitive only to carbon compounds, the disappearance of the parent species could be observed as the temperature of the converter tube increased The same flow system that was used with the NO detector was also used for this study; however, the flow rate was reduced to 50 cmS/min to concentrate the sample As a result, sample residence time in the heated converter tube was about 11 s CH3NO~ had disappeared completely by 360~ No organic products were observed when CH3NO2 was pyrolyzed Thus, NO production by CHaNO2 pyrolysis was expected to be quantitative and interference free Measured NO formation (calibrated using a number of standard tanks) was plotted versus calculated NO based on weight loss data between and 25 ppm Data were taken over a 30 day period with no evidence of drift, the observed slope of 1.00 -4- 0.03 indicates the success of this method NH3 as a permeant is entirely safe CH~NO~ is listed as a shock sensitive explosive at 100~ however, in our glass/Teflon systems we have had no problems with our 0.5-ml liquid samples 342 CALIBRATIONIN AIR MONITORING Permeation Sources for CO The use of nickel carbonyl (Ni(COh) as a CO source was investigated, together with the question of whether Ni(COh itself permeates Several permeation tubes containing Ni(COh were constructed similar to the CH3NO2 tubes Over a period of eight months, their weight loss was linear, yielding to ~g/min at 44.4~ Calibration of the CO output revealed, however, that it was a constant factor of 1.95 q- 0.06 larger than the weight loss This is probably clue to the inward permeation of some species, as yet not identified A tentative mechanism for this process may be Ni(CO)4 ~ Ni + 4CO Ni + 1~O2 ~ NiO NiO q- CO2 ~ NiCO3 CO ~ permeate where the oxygen and CO2 are supplied by the air passed over the permeation tube The foregoing mechanism results in a ratio of permeation-toweight loss of 1.87 To determine whether any Ni(CO)4 permeates under the conditions employed, two independent checks were made The permeation tube was purged with air for 67 h, and the air passed through a solution of 1M nitric acid (HNOD Under these conditions, any Ni(CO)4 would be converted to divalent nickel The resulting solution was tested for nickel, using flameless atomic absorption No nickel was observed at the limit of detection Another tube was similarly purged, and the purge-air passed through a glass capillary heated to 350~ for 240 h Since Ni(CO)4 decomposes at > 160~ a nickel mirror should form in the capillary None was observed This permeation tube, therefore, provides a source for a constant amount of CO which is reproducible and safe, making it suitable as a field calibration tool Until the chemistry is better understood, however, it cannot be used for an absolute calibration source due to the lack of stoichiometry Intercalibration of NO and 03 If air containing initially only NO2 is photolyzed, then by stoichiometry during the first minute of photolysis (at ppm), equal quantities of NO and 03 are formed [6,7] The flow system used for kl studies [8,9] gave an opportunity to study these products during the first to 15 s of photolysis The apparatus is shown as Fig of Ref 9, and was used with two slight modifications A potassium dichromate (K~Cr2OT) impregnated filter was used at the input to ensure low NO concentrations in the incoming NO~, and two blackened probes were inserted together into the stream, one for NO, the other for 03 Well calibrated NO and 03 detectors gave the same response within their ~l-ppb noise limitations The residence times from the probes to the NO and O3 detectors were both < s With NO = 03 = STEDMAN ET AL ON AIR POLLUTION MONITORS 343 0.1 ppm, the largest value we used, the decay of NO and 03 is 1/NO-I/NO0 = 25.5/60 = 0.43 where 25.5 ppm -1 min-1 is the rate of reaction This shows that a percent decay of NO and O3 can be expected, but, provided the residence times are equal, the stoichiometry of reaction again ensures the equal response, as observed Overall the foregoing method uses the stoichiometry of reactions and to produce equal NO and 03 The stoichiometry of reaction is the basis of the gas phase titration (GPT) method Using the apparatus described here, it can also be used as a rapid intercalibration The system previously used produced 0.1-ppm NO and 03 each in 0.9-ppm NO2 If the NO input was not oxidized, some excess NO was always present Further, if the input NO2 is increased, then an excess of NO over 03 can occur readily [6,7,9] This arbitrary mixture of NO and 03 is then sampled with a dual probe First, the sampling is direct as before; then NO and 03 readings are taken Next, the lights are turned off, and the differences in NO and 03 are measured Again, with well-calibrated instruments, these reductions in NO and 03 signals are equal Since the calibrations are performed by GPT, this is not surprising, but it is a simple check with simpler apparatus, since only tank or permeated NO2 and air are required, together with 110 V power for the lamps, and no flows need be measured The final intercalibration technique is based on the photostationary state, Eq The ratio kl[NO~]/k3[03] [NO] can be continuously calculated, provided simultaneous data for NO, NO~, 03, k3, and kl are available The first three are available because of the air monitoring application According to Garvin and Hampson [13], ks = X 10-13 exp (-1200/T) cm3mol-ls-~; thus, continuous temperature data can be used to obtain k3 We have shown that kl can be determined using an NO2 actinometer, and also that kz may be adequately determined by an ultraviolet (UV) pyranometer measurement (E) with multiplication by the appropriate calibration factor kl (min-~) = 0.019 E(UV W M-S) Thus if kl is measured by either method, calculation of the photostationary equation can be performed as a continuous check on equipment calibration The check is most useful during the day, when all parameters are measurable, and the ratio should stay at 1.0 At night, the equation degenerates to the solutions if [03] > if [NO] > [NO] = [03] = The validity of these terms is a check on the zeroing of NO and 03 instruments, provided the monitors are not so close to sources (< 30 s downwind) that steady state is not obtained The validity of Eq was demonstrated in downtown Detroit [8] Continuous values of the ratio were derived by hand from strip charts and shown to stay close to 1.0 We are currently extending these measurements to other areas and removing some of the problems associated with the 344 CALIBRATION IN AIR MONITORING previous study [8] So far, all data indicate the validity of the equation and its usefulness as a continuous calibration check Notice that the photostationary state will not prove valid if there is a long residence time between the air sampling and NO/NO2/O3 detection equipment Times greater than s can introduce significant concentration errors Conclusions We have demonstrated two techniques for permeation/pyrolysis NO sources These can be absolute standards They can also be intercalibrations if an NH3 or CH3NO2 detector should be available Ni(CO)4 proved to be a stable but nonquantitative CO source with no Ni(CO)4 impurity This may be a useful transfer standard but is not an absolute calibrator Two NO/O3 intercalibration techniques based on NO2 photolysis provide simple intercalibrations with portable equipment The photostationary state equation has been demonstrated to be a feasible method for a continuous intercalibration check on NO, NO2, and O~ monitoring data Acknowledgments For various parts of this study, we acknowledge the efforts of J O Jackson, and the support of the Environmental Protection Agency under grant No R802373 References [1] O'Keeffe, A E and Ortman, G C., Analytical Chemistry, Vol 38, 1966, p 760 [2] Scaringelli, F P., O'Keeffe, A E.,Rosenberg, E., and Bell, J P., Analytical Chemistry, Vol 42, 1970, p 871 [3] Saltzman, B E., Burg, W R., and Ramaswamy, G., Environmental Science and Technology, Vol 5, 1971, p 1121 [4] Hodgeson, J A., Bell, J P., Rehme, K A., Krost, K J., and Stevens, R K., American Institute of Aeronautics and Astronautics Paper, 71-1067, Palo Alto, Calif., 8-10 Nov 1971 [5] Stedman, D H., Daby, E E., Stuhl, F., and Niki, H., Journal of the Air Pollution Control Association, Vol 22, 1972, p 260 16] Stedman, D H and Niki, H., Environmental Science and Technology, Vol 7, 1973, p 735 [7] Wu, C H and Niki, H., Environmental Science and Technology, Vol 9, No 46, 1975 [8] Stedman, D H and Jackson, J O., International Journal of Chemical Kinetics Symposium, Vol 1, 1975, p 493 [9] Jackson, J O., Stedman, D H., Smith, R G., Hecker, L H., and Warner, P O., Review of Scientific Instruments, Vol 46, 1975, p 376 [10] Breitenbach, L P and Shelef, M., Journal of the Air Pollution Control Association, Vol 23, 1973, p 128 [11] Stevens, R K., O'Keeffe, A E., and Ortman, G C., Proceedings, 8th Conference on Air Pollution Control, Purdue University, Oct 1969 [12] Crawforth, C G and Waddington D.J Transactions of the Faraday Society, Vol 65, 1969, p 1334 [13] Garvin, D and Hampson, R F., NBSIR 74-730, National Bureau of Standards, Washington, D C., 1974