ENVIRONMENTAL HEALTH SERIES
                       Air Pollution
      METHODS OF MEASURING
          AND MONITORING
   ATMOSPHERIC SULFUR DIOXIDE
m      m
   U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
              Public Health Service

-------
    METHODS OF MEASURING
        AND MONITORING
ATMOSPHERIC SULFUR DIOXIDE
              Seymour Hochheiser
    U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
              Public Health Service
         National Center for Air Pollution Control
             Cincinnati, Ohio 45237

               August 1964

-------
The ENVIRONMENTAL HEALTH SERIES of reports was established
to report the results of scientific and engineering studies of man's en-
vironment: The community, whether urban, suburban, or rural, where he
lives, works, and plays; the air, water and earth he uses and reuses;
and the wastes he produces and must dispose of in a way that preserves
these natural resources. This SERIES of reports provides for professional
users a central source of information on the intramural research activities
of the Centers in the Bureau of Disease Prevention and Environmental
Control, and on their cooperative activities with State and local agencies,
research institutions, and industrial organizations. The general subject
area of each report is indicated by the letters that appear in the publica-
tion number; the indicators are

                     AP - Air Pollution

                     RH - Radiological Health

                    UIH - Urban and Industrial Health

Triplicate tear-out abstract cards are provided with reports in the
SERIES to facilitate information retrieval. Space is provided on the cards
for the user's accession number and additional key words.

Reports in the SERIES will be distributed to requesters, as supplies
permit. Requests should be directed to the Center identified on the title
page.
          Public Health Service Publication No. 999-AP-6

-------
                            FOREWORD
Use  of uniform analytical procedures  for the determination of air pollutants
would eliminate many of the difficulties presently encountered in the evaluation
of data produced by the various analysts in the field of air pollution control.
It i& very difficult to  assess and compare the nature and amount of air pollu-
tants  present in different environments when  many and various methods are
employed, each subject to different  interferences and to inherent and  systematic
errors. So that the data reported may be more meaningfully utilized by people
engaged in the various aspects of air pollution, it is desirable to have uniform
operating procedures.

A literature review of methodology relating to the measurement of atmospheric
sulfur dioxide and a description of recommended methods are presented in this
report.

Selection of the methods described in detail herein for manual and automatic
sampling and  analysis of atmospheric sulfur dioxide was based on informa-
tion, currently available. It is  recognized that further information on the meth-
odology is desirable. It is also recognized that new, more nearly ideal methods
may  be  developed. Further research and investigation to  those ends is to be
desired.

This  publication of the Division of Air Pollution,  Public Health Service, is in-
tended to serve  as a  resource document for those  involved in measurement of
pollution and  in research on new or  improved methods, and  for  those who
seek to bring about widespread agreement in matters concerning measurement
of pollution.
                           Vernon G. MacKenzie
                           Chief, Division of Air Pollution
                           Bureau of State Services
                           Public Health Service
                           U. S. Department of Health,
                           Education, and Welfare

-------
                   ACKNOWLEDGEMENTS
   Grateful appreciation is extended to those individuals who gave unselfishly
of their time and effort in reviewing this report and in contributing many help-
ful suggestions. The author wishes to acknowledge particularly the assistance
and  support provided by: Members of the  Interbranch Chemical Advisory
Committee,  Division of Air Pollution; by Dr. Aubrey P. Altshuller, Gilbert L.
Contner, Thomas R. Hauser, Charles L. Punte, Jr., Dr.  Bernard E. Saltzman,
Stanley F. Sleva, Mario Storlazzi, Elbert C. Tabor; and by Austin N.  Heller
and Jean J. Schueneman, Technical Assistance Branch, Division of Air Pollution;
and  by John S.  Nader, Laboratory  of Engineering and Physical Sciences,
Division of Air Pollution.

-------
                         CONTENTS


ABSTRACT	vii
CRITERIA FOR SELECTION OF
RECOMMENDED METHODS	  1
   Manual Sampling and Analysis	  1
     Sensitivity  	  1
     Specificity	  1
     Reproducibility	  1
     Stability of Reagents and Products  	  2
     Collection Efficiency	  2
     Simplicity and Directness  	  2
     Difficulty in Calibration	  3

   Automatic Instrumental Methods	  3
     Speed of Reaction	  3
     Temperature Coefficient	  3
     Instrument Drift	  4
     Flow or Pressure Regulation	  4
     Instrument Maintenance Requirement  	  4
RECOMMENDED METHODS	  4
   Recommended Methods for
   Manual Sampling and Analysis	  4
     West and Gaeke Method	  4
        Introduction	  4
        Reagents  	  5
        Apparatus  	  6
        Analytical Procedure	  6
        Preparation of Calibration Curve	  8
        Discussion of Procedure	  8
     Hydrogen Peroxide Method  	  9

        Introduction	  9
        Reagents  	  9
        Apparatus  	10
        Analytical Procedure	10
        Discussion of Procedure	11
   Recommended Procedure and Specifications
   for Automatic Monitoring Instrument	12
     Electroconductivity Analy/er	12

        Introduction	12
        Reagents  	12
        Apparatus  	13
        Calibration	14
        Procedure	15
        Instrument Performance Specifications  	16
        Discussion  	17

-------
COMPARISON OF METHODS CONSIDERED  	18
   Manual Methods	18
   Instrumental Methods	23
MANUAL METHODS CONSIDERED IN
ADDITION TO RECOMMENDED METHODS  	23
   Colorimetric Methods	23
     Fuchsin-Formaldehyde Method	23
     Stratman Method	26
     Barium Chloranilate Method	26
     Zinc Nitroprusside Method	26
     Astrazone Pink Method	27
     Diazo Dye Method	27
     Thorium Borate-Amaranth Dye Method	28
     Indirect Ultraviolet Determination of SO2by Means of
      Plumbous Ion  	28
     Ferrous-Phenanthroline Method	29
     Polarographic Method	29
   lodimetric Methods  	30
     Iodine Method	30
     Direct Iodine Method	30
     lodine-Thiosulfate Method	31
     Turbidimetric Method	31
   Cumulative Methods	32
     Lead Peroxide Candle Method	32
     Test Paper Method	34
   Detector Tubes   	35
INSTRUMENTAL METHODS CONSIDERED
IN ADDITION TO RECOMMENDED METHODS  	35
   Potentiometric Methods	35
   Photometric Methods  	37
     Analyzer "A"  	37
     Analyzer "B"  	38
     Analyzer "C"  	39
     Analyzer "D"  	40
   Air lonization Method	41
REFERENCES  	43

-------
                        ABSTRACT
  A literature review of methodology relating  to the measurement of
atmospheric sulfur  dioxide,  a  detailed  description of recommended
methods, and  criteria for selection of recommended methods are pre-
sented in this report. This publication is intended to serve as a resource
document for those involved in measurement of pollution and in research
on new  or improved methods, and for those who seek to  bring about
widespread agreement in matters concerning measurement of pollution.

-------
              CRITERIA  FOR  SELECTION  OF
                RECOMMENDED  METHODS
Manual Sampling and Analysis


     Evaluation of methods for the measurement  of  air pollutants was based
  upon an examination of systematic characteristics such as: Sensitivity, specificity,
  reproducibility, stability of reagents and products,  collection efficiency, sim-
  plicity and directness, and difficulty of calibration.

  SENSITIVITY

     The quantity of air needed to  be  sampled to provide a detectable amount
  of the air pollutant is a function of the sensitivity of the  analytical technique.
  To measure  fluctuation in pollution amounts  during short-term fumigations
  it is necessary to use methods capable of detecting extremely small quantities
  of the  gaseous pollutant. When integrated values over a longer time interval
  are  desired  a less  sensitive method  is applicable.  Nevertheless,  methods  ca-
  pable of sensing short-term fluctuations in gas concentrations produce more
  meaningful  data,  and  a  highly  sensitive method is, therefore,  most often
  desirable.

  SPECIFICITY

     The effect of interfering materials present in the concentration range anti-
  cipated is an important consideration in the selection of suitable air-monitoring
  methods. In many instances it is possible to eliminate interfering materials  by
  selective  absorption or by other chemical reactions and  thus avoid errors in
  analysis. At  times it may be desirable to monitor with a nonspecific method
  that measures a  group of compounds that have similar chemical properties
  and produce  the  same physiological  response.  A  knowledge of individual
  components   is,  however,  extremely valuable  in interpretation of the data.

  REPRODUCIBILITY

     Reproducibility is a  measure of the  reliability and stability of the  equip-
  ment, reagents, and technique employed in the method.  The accuracy of the
  method can be no better than its reproducibility. Relatively  good reproducibility
  does not, however, necessarily indicate relatively high  accuracy. Accuracy,
  which  is a  measure  of the deviation  from a true value, is determined by
  measuring known amounts  of a  material and observing the correspondence
  between  the  measured  and  the  true values.  In air analysis wherein many

-------
                                                METHODS OF MEASURING
methods rely upon the removal of the pollutant from the air mass, accuracy
is  a function of collection or separation efficiency, flow metering, analytical
variation, and other sources of error.
STABILITY OF REAGENTS AND PRODUCTS

   Reagents  and  absorbents  that have  comparatively  long shelf lives are
advantageous. Their use eliminates the need for frequent reagent  preparation
and  calibration.  Absorbents  that are  easily susceptible to changes in com-
position and reactivity owing to light, temperature, turbulence, or air oxidation
may produce serious errors in the analysis. If reaction products formed in air
sampling are unstable, the method becomes unsuitable for use when there is a
time lag  between  sampling and  analysis, and the method cannot, therefore,
be  employed in  certain  automatic  sampling equipment such as sequential
samplers, though  it may be satisfactory for  automatic, continuous analyzers.

COLLECTION EFFICIENCY

   The collection efficiency of methods that require scrubbers for the collection
of a gaseous impurity depends upon the chemical nature of the absorbent and
the pollutant, the  physical condition  during gas-liquid contact, and the samp-
ling rate, which in effect controls the time of contact  between the  phases. The
latter condition is particularly applicable to the physical retention techniques
used  in  adsorption processes. Methods selected far aerosol collection depend
largely  upon the  size of  the particulates in  the air or gas stream. To avoid
using multiple separators in a sampling system, a collection efficiency of greater
than  95 percent is generally  required  of one collector. It would,  however, be
permissible  to use a sampling system  having a relatively  low collection effi-
ciency provided that the desired  sensitivity, reproducibiliry,  and accuracy are
obtainable. One method of determining collection efficiency is to connect similar
collectors in series downstream and measure the ratio of pollutant occurring
in the second  and  third units  to that in the upstream unit. This is not an
absolute  method  of determining efficiency  since there  may be a systematic
threshold concentration  below which no reaction occurs. Collection efficiency
is determined by  introducing  known  amounts  of gas or  aerosol at different
dilutions. A method that has been previously calibrated and results  in relatively
high collection efficiency may be used  as a standard for comparison of another
method  provided  that the methods are  both  subject to the same interferences.

SIMPLICITY AND DIRECTNESS

   When  a  simple  technique  gives  the  same degree  of accuracy as  a more
complicated  and  time-consuming method, one would  obviously  choose  the
simpler  technique,  except, perhaps,  when equipment costs  render prohibitive
the use of the simpler method.

-------
AND MONITORING SO 2
   Indirect  methods  are  subject to greater  inaccuracies owing to calibration
 difficulties and  should be  avoided whenever possible. Factors that influence
 simplicity are the number and difficulty of unit operations and the type of
 laboratory  and field sampling controls required.

 DIFFICULTY IN CALIBRATION

   An ideal method would be one  in  which a specific compound of known
 composition is formed that is not subject to  temperature, light sensitivity,
 or other instabilities and that can be  measured  directly and without inter-
 ference from  other compounds. In this event the quantity of product formed
 is directly related to the concentration of the reactants if a quantitative reaction
 is assumed. To determine the  efficiency of reaction, a known quantity of the
 pollutant is introduced into the system at various concentrations corresponding
 to those existing in ambient air, and the amount recovered is determined.  The
 techniques  available for this type of standardization are dynamic  dilution  and
 static dilution.  The  effect of interferences can  also be determined by similar
 techniques.  After  primary  calibration  of the  system in the manner indicated
 above,  secondary calibration  methods can be employed, that  is, standard
 solutions, conductivity cells of known resistance, and so forth.
 Automatic Instrumental  Methods
    All the factors pertinent to the manual measurement  of pollutants  apply
  to  measurement in continuous,  automatic recording instruments.  Other im-
  portant  parameters are: Speed of reaction, temperature coefficient, instrument
  drift, flow or pressure regulation, and instrument maintenance requirement.

  SPEED OF REACTION

    The time required to attain reaction equilibrium is an important  considera-
  tion in the choice of reagent to be used in continuous  analyzers, particularly
  in colorimetric  recorders.  Many  colorimetric methods require a definite time
  for complete reaction, and it may become necessary, therefore, to alter reagent
  composition  or calibrate at a  definite time after sampling prior to complete
  reaction,  to attain conditions consistent with  instrument response. Instrument
  response,  which is  a function  of  flow  rate  and distance traveled  by the air
  sample  and reagent  prior  to  contact with the sensing mechanism, should
  be minimal.

 TEMPERATURE COEFFICIENT

    The instrument should be calibrated at different temperatures corresponding
 to actual operating  conditions to  determine  the effect  of temperature so that
 proper corrections can be applied. Instrumental methods employing reagent and

-------
                                               METHODS OF MEASURING
 detection systems that vary as little as possible with temperature should be
 selected.  Many  methods  of analysis and systems of detection are sensitive
 to temperature, and thermostatic regulation is required to minimize this source
 of error.

 INSTRUMENT DRIFT

   Changes  in  zero and span  or  gain settings should be minimal to avoid
 errors  in the recordings and  frequent calibration. Limitation of drift to a
 maximum  of 1 percent in  a  24-hour period  would be desirable. Excessive
 electronic noise interferes with the  interpretation of results.  Unfortunately,
 in  many instruments, noise level varies directly with sensitivity so that noise
 level will affect  the maximum sensitivity attainable. Limitation of noise level
 to a maximum of 1 percent of full scale would be desirable.

 FLOW OR PRESSURE REGULATION

   Nonuniform airflow or pressure, and reagent flow, where appropriate, may
 produce  serious errors in the  analysis since the concentration recorded is a
 function  of  the  quantity of  air sampled  (except in closed-path instruments).
 This is especially important when comparatively low flows are employed, since
 changes in flow would result in a higher percentage of error. For this reason,
 constant-flow devices should be employed, and  flow rates should be automat-
 ically recorded. Calibration curves of variation of concentration with flow
 should be prepared,  and the instrument should be operated under sampling
 conditions least affected by flow variations.

 INSTRUMENT MAINTENANCE REQUIREMENT

   The instrument  should be constructed in such a manner as to require the
 minimum amount of attention and should be designed for simple maintenance.
 Parts  that require more frequent replacement  or  servicing should be easily
 accessible. Complete instructions and diagrams for maintenance and trouble-
 shooting should be provided.   No more than 1 day per week of attended
 operation would be desirable.


                 RECOMMENDED  METHODS

Recommended  Methods  For
Manual  Sampling and Analysis
WEST AND GAEKE METHOD J

Introduction

  The West-Gaeke method is applicable to  the determination of  sulfur dioxide
(SO2) in outside ambient  air in  the concentration range from about 0 005 to

-------
AND MONITORING SO 2
  5 ppm1  SO2 in the air sample is  absorbed in 0.1  molar sodium tetrachloro-
  mercurate. Nonvolatile dichlorosulfitomercurate ion is formed in this process.
  Addition of  acid-bleached  pararosanih'ne  and formaldehyde  to the complex
  ion produces red-purple pararosaniline methylsulfonic  acid,  which is  deter-
  mined  spectrophotometrically.2  The  system obeys Beer's  law up to  about
  10 microliters of 862  per  10 milliliters of  absorbing solution. This method
  is  more sensitive than the  hydrogen  peroxide method and is not subject to
  interference from  other acidic or basic gases or  solids such as SO3, H2SO4,
  NHs,  or  CaO;  the  analysis should,  however,  be completed within 1 week
  after sample  collection, and the concentrations of  ozone and  NO2  should be
  less than that of the SO2. 3'4'5

  Reagents

     All chemicals used must be A. C. S. analytical-reagent grade.

     Absorbing Reagent, 0.1  Molar  Sodium  Tetrachloromercurate. Dissolve
  27.2 grams (0.1 mole) mercuric chloride and 11.7 grams (0.2 mole) sodium
  chloride in 1  liter of distilled water.  (CAUTION: Highly poisonous; if spilled
  on skin, flush off with water immediately.) This solution can be stored at room
  temperature for several months.

     Pararosaniline Hydrochloside (0.04%),  Acid Bleached. Dissolve 0.20 gram
  of pararosaniline hydrochloride in  100 milliliters  of distilled  water and filter
  the solution  after 48 hours. This solution is stable for at  least 3 months if
  stored  in the dark  and kept cool.  The pararosaniline used should have an
  assay  of  better  than  95  percent  and an absorbance maximum  at 543 or
  544  millimicrons. Pipette  20 milliliters  of this into a 100-milliliter volumetric
  flask. Add 6 milliliters of concentrated HC1. Allow to stand  5 minutes, then
  dilute to mark  with distilled water.  This solution should be pale yellow with
  a greenish tint. It can be stored at room temperature in an amber bottle for
  a week or for  about 2 weeks if refrigerated.

     Formaldehyde, 0.2 Percent. Dilute 5 milliliters of 40 percent formaldehyde
  to 1,000 milliliters with distilled water. Prepare weekly.

     Standard   Sulflte  Solution.  Dissolve  640  milligrams  sodium metabisulfite
  (assay  65.5% as SO2) in 1.0 liter of water. This yields a solution of approxi-
  mately  0.40 milligram per milliliter  as  SO2- The solution should be standard-
  ized by titration with  standard 0.01 normal I2 with starch as indicator, and
  should  be adjusted  to 0.0123 normal. Then 1  milliliter = 150 microliters
  S02 (25°C, 760 millimeters Hg). Prepare and standardize freshly.

    Starch  Solution  (Iodine Indicator), 0.25 Percent  Make  a thin paste of
  1.25 grams  of soluble starch in  cold water and  pour  into 500 milliliters of
  boiling water while stirring. Boil for a  few minutes. Keep in glass, stoppered
  bottle.

-------
                                              METHODS OF MEASURING
   Standard Iodine Solution, 0.01 Normal. Dissolve 12.69gramsof resublimed
iodine in  25 milliliters of a solution containing 15 grams of iodate-free KI;
dilute to the 1,000-milliliter  mark in  a volumetric flask.  Pipette exactly 100
milliliters  of this 0.1 normal solution and dilute to 1,000  milliliters in a vol-
umetric flask with 1.5 percent  KI.  This  solution  can be  used as  a primary
standard  if the  weighing is  carefully  done, or it can be checked  against a
standard  thiosulfate  solution.  This solution should  be stored in an amber
bottle, refrigerated, and then standardized on the day of use.

Apparatus

   Absorber. An  all-glass midget impinger or other  collection device capable
of removing SC>2 from an air sample using 10 milliliters of absorbing reagent
should be used.  (Among the suppliers of midget impingers are Corning Glass
Company and Gelman Instrument Company.)

   Air Pump. The air pump should be capable of drawing 2.5 liters  per minute
through the sampling assembly.

   Air-Metering and Flow Control Devices. Metering and control devices should
be  capable of  controlling and  measuring  flows with an accuracy of ± 2
percent.  The flow meter  should be calibrated for variations in reading with
temperature and pressure of the airstream so that the appropriate  corrections
can be applied.

   Thermometer (or other Temperature-Measuring Device). The thermometer
should have an  accuracy of ± 2°C.

   Mercury Manometer (or other  Vacuum-Measuring Device). The manometer
should have an  accuracy of 0.2 inch Hg.

   Spectrophotometer or Colorimeter. Color-measuring devices should be cap-
able  of  measuring  color intensity  at  560 millimicrons,  in absorbance cells
1 centimeter or larger.

Analytical Procedure

   Collection of Samples.  Set  up a sampling  train consisting of,  in order,
absorber, trap to protect flow device, flow control  and metering devices, tem-
perature and vacuum gauge, and air pump.  All probes and tubing upstream
from the bubbler should be pyrex glass, stainless steel, or teflon. Butt-to-
butt connections may be made with tygon tubing. The  downstream flow meter-
ing device can be empirically corrected to atmospheric conditions by  conducting
a dummy run with an upstream flowmeter inline that  is  open to the atmosphere.

   Pipette exactly 10 milliliters of absorbing reagent into the  absorber. Aspirate
the air sample through the absorber at a rate of 0.2 to 2.5 liters per minute

-------
AND MONITORING SO 2
  (depending upon the concentration of SOg in the atmosphere and the sampling
  time desired).  The sampling time  may vary  from a few minutes to 24 hours.
  For 24-hour sampling the absorber selected should be capable of containing
  20 milliliters or more of absorbing reagent. For best results, the sampling time
  and rate should be chosen to provide a concentration of approximately 2 to 4
  microliters  of SO2 in  10 milliliters of the absorbing reagent. The dichloro-
  sulfitomercurate formed may be stored for 3 days with only a slight decrease
  in strength (about  1 percent per day). Ifsamples are stored for longer periods,
  a correction factor should be applied. The sample may be stored in the collection
  device or transferred to a stoppered glass or polyethylene container.
     Analysis. If a mercury precipitate is present owing to the presence in the
  air sample of  inorganic sulfides, thiols, or thiosulfates, it may be removed by
  filtration or centrifugation. To the clear sample, adjusted to 10 milliliters with
  distilled  water to compensate  for  evaporative  losses, add 1.0 milliliter of
  acid-bleached  pararosaniline  solution  and  1.0 milliliter of the 0.2 percent
  formaldehyde  solution and mix.

     Treat a 10-milliliter  portion of  unexposed  sodium tetrachloromercurate
  solution  in the same manner for use  as the blank. If the collecting reagent
  remains exposed  to the atmosphere during the interval between sampling and
  analysis, the blank should be exposed in the same manner.

     Allow 20 minutes for maximum color development and read the absorbance
  at 560 millimicrons in a spectrophotometer with the blank as reference.
    Calculations. Convert the volume of air sampled to the volume at standard
  conditions of 25°C, 760 millimeters Hg:
       v    -v      (p-pm)       298.2
       Vs   ~ v x     29.97       (t + 273.2)
       Vs   =  Volume of air in liters at standard conditions
       V    =  Volume of air in liters as measured by the meter
       P    —  Barometric pressure in inches of mercury
       Pm  —  Suction at meter in inches of mercury
       T    =  Temperature of sample air in degrees centigrade.

    Ordinarily the correction for pressure is slight and may be neglected.

    Compute the microliters of 862 in the sample by multiplying the absorbance
  by the slope of the calibration plot. Then the concentration is:

         ppm  SO 2 by volume  =  ^ • •	2-

-------
                                               METHODS OF MEASURING
Preparation of Calibration Curve

   Rpette exactly 2 milliliters of standard sulfite solution into a 100-milliliter
volumetric flask and dilute to mark with absorbing reagent. This final solution
contains 3.0 microliters SO2 per milliliter.

   Add accurately 0.5-, 1.0-, 1.5-and 2.0-milliliter portions of the dilute standard
sulfite solution to  a series  of 10-milliliter volumetric flasks and  dilute to the
marks  with  absorbing  reagent. Continue with the analysis procedure given
above.

   Plot  the absorbance (optical density) as the ordinate against the microliters
of  SO2  per  10  milliliters  of absorbing solution on rectangular coordinate
paper. Compute slope of straight line best fitting the data.

Discussion of Procedure

   The  error for the combined sampling and analytical technique is ±  10
percent in the concentration range below 0.10 ppm with increasing accuracy
with concentration in the range 0.1 to 1 ppm.

   The  measurements should  be reported to the nearest 0.005 ppm  at con-
centrations  below  0.15  ppm and  to the  nearest 0.01  ppm  above 0.15 ppm.
03 and NO 2  interfere  if present  in the air  sample at concentrations  greater
than SO2.4  Interference of NO2 is eliminated by including 0.06 percent sulfamic
acid in the absorbing reagent.6  This may, however, result in a different cali-
bration curve  of lower sensitivity and in  greater losses of SO2 on  storage of
the sample for more than  48  hours after sample collection. NO2 interference
may also be eliminated by adding o-toluidine  subsequent to sample collection. 7

   Heavy metals, especially iron  salts, interfere by oxidizing dichlorosulfito-
mercurate during sample collection. This interference is eliminated by including
ethylenediaminetetracetic  acid  in  the  absorbing  reagent.    Sulfuric acid  or
sulfate  do  not interfere. There is  no  experimental evidence to indicate that
SO 3 interferes; it probably hydrolizes preferentially in the absorbing reagent
to form H2SO4   rather  than  combines with sodium tetrachloromercurate to
form  the  dichlorodisulfitomercurate complex ion. If the latter reaction should
prevail,  SO3 would interfere  positively.  If  large  amounts  of solid material
are present,  a filter may be used  advantageously  upstream; however, a loss
of SO2 may occur.8

   The color produced is independent of temperature in the range 11 to 30°C
and is stable for 3 hours.

   Much  difficulty  with  the method has been caused by the use of  impure
pararosaniline hydrochloride.9 A commercial brand is now available that is

-------
AND MONITORING SO 2
  specially  selected for this procedure (Fisher  Scientific Company, catalog No.
  P-389). The purity of the reagent may be estimated by comparing the  slope
  of  the calibration  plot  with  the value 0.15  absorbance unit per microliter
  (obtained  with 1-centimeter cells in a Gary spectrophotometer), which corres-
  ponds to a molar absorptivity of 36,700.
   HYDROGEN PEROXIDE METHOD

   Introduction

     This  method  is applicable to the determination of SO 2 in outside ambient
   air in the concentration range  from about 0.01 to 10.0 ppm.  SO2 in the  air
   sample  is absorbed in 0.03 normal hydrogen peroxide (H2O2) reagent (ad-
   justed to about pH 5).10'11  The stable and nonvolatile sulfuric acid formed
   in this process is titrated with standard alkali. The method requires only simple
   equipment and can be performed  by analysts having lesser skills; it is pre-
   ferable to  the West-Gaeke method if SO2 is the principal acid or basic atmos-
   pheric gaseous pollutant and if long storage of samples (greater than 1 week)
   prior to  analysis is required.


   Reagents

     All chemicals used must be A. C.S. analytical-reagent grade.

     Absorbing Solution, Hydrogen  Peroxide, 0.03 Normal, pH 5. Dilute 3.4
   milliliters  of  30  percent H2O2 solution to 2 liters with distilled water. Deter-
   mine  the  alkalinity of the  solution by taking a 75-milliliter portion, adding
   3 drops of mixed indicator, and  adding approximately  0.002 normal HC1
   or HNO3 from  a buret  until the indicator turns pink (pH 5). Calculate  the
   amount  of acid  necessary  to  adjust the acidity of the bulk of the absorbent
   and  add  the required  amount.  The  zero blank,  obtained  by  titrating
   75 milliliters of the adjusted reagent with 0.002 normal  NaOH  to the indicator
   equivalence point  (green),  should  be  not  more than 2 drops. The reagent
   is stable at room temperature for at least 1 month.

     Mixed  Indicator, 0.1 Percent.  Dissolve 0.06 gram bromocresol green and
   0.04  gram methyl red  in 100  milliliters of methanol.  When stored in  an
   amber bottle  at room  temperature the reagent is stable for at least 6 months.

     Standard  Sulfuric Acid  Solution, 0.002 Normal. Prepare this  solution  by
   appropriate dilution  of concentrated sulfuric acid. Standardize by the gravi-
   metric barium sulfate  method with  a 200-milliliter portion or  with a primary
   standard such as Na2B4O7-10 H2O. This reagent maybe stored indefinitely
   without change in strength.

-------
10                                            METHODS OF MEASURING
    Standard  Sodium Hydroxide Solution, 0.002 Normal. Prepare 2 liters of
 this  solution by dilution of 1 normal sodium hydroxide with freshly boiled
 (CO,,-free) distilled water. Standardize as follows: Pipette 25 milliliters of stand-
 ard  sulfuric  acid solution into an  Erlenmeyer flask, add 3 drops of mixed
 indicator  solution, and  titrate with the sodium hydroxide reagent contained
 in a buret to the green equivalence point. Store the reagent in a polyethylene
 or other alkali-resistant bottle and restandardize bimonthly.

 1 ml of 0.002  N NaOH = 64  pg SO2  = 24.47 pi  SO2 (25°C, 760 mm Hg).

 Apparatus

    Absorber. A standard all-glass impinger  or fritted bubbler  is acceptable
 (capacity  about 300 milliliters).  (Among-the suppliers  are Corning Glass
 Company and Fisher Scientific  Company.)

    Air   Pump.  The  pump should be capable of  drawing 1  cfm through the
 sampling assembly.

    Thermometer (or Other Temperature-Measuring Device). Thermometer should
 be  capable of  controlling  and measuring  flows with an accuracy  of ± 2
 percent.  The flow  meter should be  calibrated for variation in reading  with
 temperature and pressure so that the appropriate corrections can be applied.

    Thermometer (orOtherTemperature-MeasuringDevice). Thermometer should-
 have an accuracy of ± 2°C.

    Mercury Manometer  (or  Other  Vacuum-Measuring Device).  Manometer
 should have an accuracy of 0.2 inch Hg.

    Buret.  A buret of 25- or  50-milliliter capacity graduated in 0.1-milliliter
 subdivisions, preferably  with  teflon plug,  should be capable of measuring
 volume with an accuracy of 0.05 milliliter.

 Analytical Procedure

    Collection of Samples. Set  up a sampling train consisting of, in order,
 impinger, trap  to protect flow device, flow control device,  flow-metering device,
 temperature  and  vacuum gauge,  and air pump. Measure  75 milliliters of
 absorbing  reagent into  the large  impinger. Aspirate air through the bubbler
 at a rate of 1  cfm for  30 minutes. Note the readings of the vacuum gauge
 and thermometer.  The downstream  flow-metering device  can be empirically
 corrected  to atmospheric conditions \ by conducting a  dummy run  with an
 upstream flow  meter inline that is  open to the atmosphere.  If an integrated
 24-hour air sample is  desired the  sampling rate may be reduced to 1 liter
 per minute. For SO 2concentrations of 0.3 ppm and greater the strength of the

-------
AND MONITORING SO 2
   standard alkali  may be increased or the sampling time shortened.  For con-
   centrations  greater than 0.8 ppm a second impinger should be connected in
   series  so that  a recovery efficiency of 98 percent is maintained.  All probes
   and tubing upstream from the impinger should be pyrex glass, stainless steel,
   or  teflon.  Butt-to-butt connections may be made with short lengths of tygon
   tubing. The collected sample will not decompose on standing; consequently,
   the solution may be titrated  long after sample collection. The sample may be
   stored in the impinger, which has been stoppered or transferred to a stoppered
   glass or polyethylene container.
     Titration. Add three drops of mixed indicator solution and titrate  the solu-
   tion with standard 0.002 normal sodium hydroxide until the color changes
   from red to green. A reagent blank is titrated  in the  same  manner, and this
   result (which should be less  than 0.1 milliliter) is subtracted from  the sample
   liter.

     Calculations. Convert the  volume of air sampled to the volume at standard
   conditions of 25°C, 760 mm Hg:
                      P - Pm)       298.2
       V0   = V x
                       29.97       (t + 273.2)
       Vs  -  Volume of air in liters at standard conditions
       V   =  Volume of air in liters as measured by the meter
       P   =  Barometric pressure in inches of mercury
       Pm  =  Suction at meter in inches of mercury
       T   =  Temperature of sample air in degrees centigrade.
       Results are computed on the basis of the following reaction:

                   S02  + H202 	1> H2S04

     Thus  the  net  titer  of 0.002 .normal  NaOH  (in milliliters) multiplied by
  24.47 gives the microliters of sulfur dioxide. Then the concentration is:

                           ppm SO2 by ^i™™°-   ^ s°2
                                                    * s
  Discussion of Procedure

     The  error for the  combined  sampling and analytical technique is ± 10
  percent in the concentration range below 0.1 ppm with increasing accuracy with
  concentration in the range 0.1 to 1 ppm. The measurements should be reported
  to the nearest 0.01 ppm.

     The presence in  the air sample of strong acidic gases other than SO 2 or
  reactive acid solids such as HC1 and NaHSOs gives erroneously high results,
  whereas  the  presence of alkaline gases or reactive basic solids such as NH3
  and  CaO gives erroneously low  results.  H2SO4 does not interfere since it is
  not appreciably separated from the airstream owing to  its  small particle size
  except, perhaps, when  the relative humidity is greater than 85 percent, which
  could result  in particle  sizes of greater  than 1  micron. SOs  gas,  if present,

-------
12                                           METHODS OF MEASURING
 would be a positive interference. Sulfates do not interfere. CO2 does not inter-
 fere since it is not absorbed in the acid-absorbing reagent. If large amounts of
 solid material are present, a filter may be employed advantageously upstream,
 however, a  loss of SOa may. occur.8 The  extent of this  loss would depend
 upon the composition of the particulate matter and the nature and retentive
 capacity of the filter used. The  acid base indicator is not  included hi the ab-
 sorbing reagent because it tends to decompose during sampling, which leads
 to unreproducible results.

 Recommended Procedure and Specifications
 for Automatic Monitoring Instrument

 ELECTROCONDUCTIVITY ANALYZER*

 Introduction

    This method is applicable to the continuous, automatic sampling, analyzing,
 and recording of SOa concentrations in outside ambient air in the concentration
 range  from about  0.01 to  2  ppm. The upper limit can range from 1 to  20
 ppm, depending upon reagent flow and airflow rates and on electronic ampli-
 fication. Air  is  continuously admitted to  the  absorber where the SO2 in the
 airstream is removed by a continuously flowing liquid absorbent. The electrical
 conductivity of the resulting solution is continuously measured and recorded,
 and the readings  obtained  are proportional to the concentration of S02  in
 the sampled  air. The method is not specific for 862 since other soluble elec-
 trolyte-forming gases and solids affect the results.

 Reagents
    All chemicals used must be A. C. S. analytical-reagent grade.

    Distilled Water.  For  use  as  absorbing solution,  distilled water should be
 prepared by passing it through a cation-anion exchange resin.

    Hydrogen Peroxide-Sulfuric Acid, Alternate Absorbent, 2 x 10"3 molar  H2O2.
    1 x 10'5  normal ^804. Add 2.3 milliliters of 3 percent H2O2 solution
 per liter of 1 x 10"5 normal H2SO4.
   It is  not  intended to  imply that only analyzers using  electroconductivity
 detectors are  suitable for automatic monitoring of SC>2  in air. At the time
 of this  report,  electroconductivity analyzers most nearly fulfilled the criteria
 for evaluation and  the recommended  instrument performance specifications
 set forth herein. Further research on  and development of  automatic  air-
 monitoring  instruments  incorporating  other  principles of detection such as
 potentiometry, photometry, air ionization, thermal  conductivity, and so forth
 should  be encouraged as well as electroconductivity.

-------
AND MONITORING SO 2                                                 13
     2,  4,  5 Trichlorophenate (Dowicide B) or Any Suitable Fungicide. Add 2
  milligrams of absorbing reagent per liter.

     Calibration Reagent  A reagent should  be prepared corresponding  to  1
  ppm SO2 under the conditions  of air  sample flow  rates employed. Add the
  calculated amount  of  0.1 normal H2SO4 solution to  1 liter of unexposed
  absorbing reagent.

  Apparatus

     The apparatus  should consist  of a suitable assembly of sampling probe,
  absorber, regulating  and recording device for airflow,  regulating and re-
  cording device for  liquid flow,  air pump, liquid-metering  pump  or  constant-
  head  device  with capillary tube for dispensing the  absorbent at a constant
  rate, thermostatted cabinet, conductivity electrodes, and conductivity recorder.

     Sampling  Probe.  The sampling  probe  should be made  of pyrex glass,
  316 stainless  steel or teflon  tube  with intake end turned down,  and a loose
  glass wool filter to remove large particulate matter. The inline  prefilter should
  be  mounted  indoors, exterior to  the instrument,  or, if necessary, inside the
  thermostatted  cabinet, to  prevent  condensation of water vapor, which would
  absorb SO2 and result in serious  losses.

     Absorber.  A venturi scrubber or any reagent-air-contacting system capable
  of a scrubbing efficiency of 98 percent or more is acceptable.

     Measuring, Regulating, and Recording Device for Airflow. A flowmeter, needle
  valve, or other flow-measuring, flow-regulating device capable of measuring
  flows with an accuracy of ± 2 percent should be used.  The airflow rate should
  be  monitored continuously  by means  of suitable mechanical and electronic
  circuitry  to cause  characteristic markings to appear on the conductivity re-
  corder chart.
                                     \
     Measuring, Regulating, and Recording Device for Liquid Flow. A metering
  pump, flowmeter, and needle valve or device capable of measuring flows with
  an  accuracy  of  ±  2  percent is  acceptable. The liquid flow rate should be
  monitored continuously  by means of  suitable mechanical and electronic cir-
  cuitry to  cause characteristic markings to appear on the conductivity recorder
  chart.

     Air and Liquid  Pumps. Any air  pump  and liquid pump or combination
  air-liquid  pump is  acceptable that is  capable of  drawing air at a rate of 5
  liters per  minute and liquid at a rate up to 30 miUiliters per minute through
  the absorption-analyzing system under conditions  of continuous operation.
  To assure satisfactory  performance, the  pump should have a much greater
  capacity, about 20-liters-per-minute airflow under the conditions of operation.

-------
14                                            METHODS OF MEASURING
   Thermostatted  Cabinet.  The  reagent feed lines, absorption column, and
 conductivity cells  should be enclosed in an insulated compartment thermo-
 statically maintained at a temperature a few degrees higher than the highest
 ambient  temperature expected.  The temperature  of the  reagent in the con-
 ductivity cells should be continuously monitored by means of suitable electronic
 circuitry to cause characteristic markings to appear on the conductivity recorder
 chart.

   Conductivity  Electrodes. Two  sets of platinum  dip electrodes of suitable
 dimensions, one pair to measure the conductivity of the unreacted  reagent and
 the other that of the reacted reagent, should be provided.

   Conductivity  Recorder.  A zero-to-10 millivolt,  potentiometric,  strip  chart
 recorder  with  30-day chart and scale graduated from zero to 100 and chart
 speed of 1  or  2 inches  per hour, or any instrument capable of recording the
 differential  output of the conductivity cells corresponding to an SO2 concen-
 tration range of zero to 2  ppm with an  accuracy of ± 1 percent  of full scale
 should be provided.

   Reagent Reservoir and Delivery Bottle. Any bottle with sufficient volume
 to contain  a 1 week's  supply of  reagent can be used; the reagent should be
 protected from air pollutants by the insertion of a soda-lime charcoal column
 on the air inlet line.

   Switch-controlled Electronic Check.  A switch-controlled  manual check con-
 taining the proper resistance to cause a deflectionof the recorder corresponding
 to a concentration of 1 ppm SQ% should be used.

   Switch-controlled Electronic Zero Check. A check  should be used to  simulate
 a differential conductivity  between the reference cell and  a known resistance
 equivalent  to  zero ppm SO2.  This also  serves as a check on the purity of
 the unreacted reagent.

 Calibration

   Static  Calibration, Standard  Solutions. The instrument may be calibrated
 by sulfuric acid  solutions of known composition corresponding to definite
 atmospheric SO2  concentration in the  range zero  to 2 ppm.  Solutions cor-
 responding to  0.5, 1.0,  1.5,  and 2.0 ppm SO2  are prepared by  the addition
 of calculated amounts  of  0.1 normal H2 SO4 to the  absorbing reagent. The
 static methods described below  may  be used  when the absorption efficiency
 is known to be  greater than 98 percent.

   1.  Establish  instrument zero  by  introducing  the unexposed absorbing
       reagent  in both  the reference  and sample conductivity  cells; then the

-------
AND MONITORING SO 2                                                 15
        standard solutions  are substituted for the absorbing  reagent in the
        sample conductivity cell only. The instrument reading is checked against
        the standard reagent,  and if necessary, the instrument is adjusted by
        means of the span control to indicate the correct concentration.

    2.  The standard solution is  substituted for the absorbing  reagent and is
        introduced  directly  into the system as in normal operation. SOa-free
        air, obtained  by passage of  air  through a  drying tower containing
        ascarite or  soda lime, is admitted to the analyzer under the same con-
        ditions as in actual sampling. The instrument reading is checked against
        the standard reagent, and the necessary corrections are made. Absorbing
        reagent is maintained in the reference cell.

    Dynamic  Calibration,  Standard Gas  Mixture. The dynamic calibration
  methods described below should be employed to take into  account the scrubbing
  efficiency of the absorbing column under flow conditions. Standard air-SO2
  mixtures may be  prepared in a rigid test chamber, compressed gas cylinder,
  or collapsible mylar or other inert plastic bag  and then introduced into the
  analyzer. The mixtures are prepared by diluting  a measured quantity of pure
  SO2  gas with a known volume of SO2-free air. A measured quantity of SO2
  gas may be  introduced into the  test chamber through a rubber diaphragm
  by means of  a hypodermic syringe and needle. In the compressed-gas-cylinder
  technique, the dilution  of SO2 is accomplished by  introducing a measured
  amount of SO2 by a hypodermic syringe into a partially evacuated, stainless
  steel cylinder  and compressing the mixture by  addition of air  contained in
  a compressed-air cylinder at high pressure until the desired pressure is reached.
  When  a rigid test chamber is employed, corrections  should be applied for
  the diminution of  gas concentration resulting  from the dilution of the test gas
  by influent air during sampling, or a flexible plastic  bag can be put inside
  the  chamber  to receive replacement air. The concentration of SO 2 in the gas
  streams prepared  by  the  above methods should be calibrated by the West-
  Gaeke or hydrogen peroxide methods.

    The  instrument  may also be calibrated  against a manual  method such
  as the  West-Gaeke or  the  hydrogen  peroxide  method. Atmospheric air  or
  synthetically  produced air-SO2 mixtures are introduced simultaneously into
  the  analyzer  and the manual absorber. The  instrument record is adjusted to
  read S02 concentrations as determined by the manual  method.

    Fluctuations in air-liquid flow rate and temperature may occur. Calibration
  curves  showing the effect  of these variations on recorder reading should be
  prepared so that appropriate correction factors can be  applied.
  Procedure
         When a fresh supply of absorbing solution is installed in the apparatus,
         any  air  bubbles that may  form  iii the reagent feed lines must be
         removed.

-------
16                                             METHODS OF MEASURING
    (b)  Check the airflow and liquid flow rates and temperature of the reagent
        in the conductivity cells and make necessary adjustments.

    (c)  Check the ink  supply to the pens hi the recorder. Check the recorder
        battery and output range by suitable electronic test equipment.

    (d)  Lubricate the pump and motors as required.

    (e)  Calibrate the  instrument once  a week  or more or less frequently as
        required.

    (f)  Interpret the charts by using a calibration  curve on a section of the
        recorder paper, or integrate under  the curve by using a planimeter.
        Correct for deviations from  the standard curve due to variations hi
        temperature and airflow or liquid flow. Hourly concentrations may be
        used  to calculate time-concentration values.  Short-time peak concen-
        trations are measured as required.

 Instrument Performance Specifications*

    (a)  A reproducibility  of ± 1 percent  of full-scale  deflection over a 24-
        hour period,

    (b)  an instrument accuracy of ± 2 percent  of full-scale deflection  over
        a 24-hour period,

    (c)  linear response in all ranges  of concentration (90% response time
        not to exceed  1 minute for the concentration range zero to 2 ppm),

    (d)  electronic drift not to exceed 1 percent of full scale  in 24 hours,

    (e)  zero to 2 ppm full-scale range with a sensitivity  of 0.01 ppm,

    (f)  sufficient control of temperature, reagent,  and sample flows to assure
        the accuracy of the calibration technique.
 * Electroconductivity  analyzers  are available  from the following manufac-
 turers: Beckman Instrument Company, Davis Emergency Equipment Company,
 Leeds  & Northrup,  Instrument  Development Company,  Mervyn-Cerl Ltd.,
 and distributed by Gelman Instrument Company. Not all models supplied
 by  manufacturers meet these performance requirements. In the case of some
 of  the  manufacturers none  of  the  models available as of January  1964
 meet specifications.

-------
AND MONITORING SO 2                                                  17
  Discussion

     Electroconductivity,  which  is  measured in terms of the resistance of the
  solution between two electrodes immersed in it, is a property of all ionic solu-
  tions  and is not specific for any particular compound.  It is dependent upon
  the number and type of ions  dissolved in solution.  Soluble gases that yield
  electrolytes  in  solution  cause  the greatest  interference.  All hydrogen halides
  present  would  be measured.  Except  near  special sources  of contamination,
  these  gases are, however,  seldom present in air in appreciable amounts in
  comparison with SO2. Weak  acidic gases such as H2S cause practically no
  interference  because  of  their  slight solubility and  poor conductivity.  If the
  water is free of bases, the carbon dioxide content of air causes no  interference.
  Alkaline gases, such as  ammonia, interfere by neutralizing the  acid and yield
  comparatively low results because the transport number of the hydrogen ion
  is  several times greater than  that of other cations.  Similarly, lime  dusts or
  other  basic solids,  if absorbed,  would cause comparatively  low results for
  S02.  Neutral and  acidic  aerosols such as sodium chloride or sulfuric acid
  would give comparatively  high results depending on their solubility, ioniza-
  tion,  and the ability of the absorption system to remove them from the  air-
  stream,  which,  in this  method, is very poor  unless particle size  is relatively
  large.  Since the particle  size  of sulfuric  acid  mist is small (less than 1 p )
  except perhaps  when the relative humidity is greater than 85 percent,  it is not
  measured appreciably in this method. A special absorber and different operat-
  ing parameters are required  for effective  collection  of sulfuric acid mist. A
  loose  glass  wool filter is used to minimize maintenance time requirements for
  cleaning of absorber and  conductivity  cells, and reducing the interference of
  particulate matter. SOy gas, if present, would  result in a positive  interference.
  Instrument response is a function of gas  concentration, airflow and liquid flow
  rate,  and conductivity cell constants. Response time is a function of liquid
  flow  rate  and  the  distance and volume between the absorber and the con-
  ductivity cells.

     A  change in temperature  of  2°F will alter the conductivity of  a strong
  electrolyte by approximately 2 percent; consequently, for accurate operation,
  the absorption  column and  conductivity  cells  should be  enclosed in an in-
  sulated  compartment thermostatically  maintained  at  a temperature a  few
  degrees higher than the maximum ambient temperature expected.

                             12
     Perley  and  Langsdorf    reported that, at SO2 concentrations of 2 ppm,
  amounts  of CO2   as high as  2,000 ppm had no  influence on the  resultant
  S02 measured.  H2S concentrations of 25 ppm result in  a positive reading of
  0.2 ppm. Since this concentration is rarely approached in atmospheric sampling
  the interference of this weak acid is not appreciable.

     Yocom et al.13  reported on the effect of hydrogen chloride on the operation
  of an electroconductivity analyzer. Concentrations of hydrogen chloride in the

-------
18                                             METHODS OF MEASURING
 range  1.2  to 11.4  ppm were collected at an efficiency of 75 percent.  The
 relationship between concentration of HC1  and its equivalent concentration
 in ppm SO2 was  1:2.9.  From these considerations it was shown that appreci-
 able interference is caused by the presence of HC1,  as would be the case  with
 all hydrogen halides and other strong electrolytes.

    Jacobs,  Braverman,  and  Hochheiser u  adapted an  electroconductivity
 analyzer to the measurement of 862  in outside air in the concentration range
 zero to  0.95 ppm full scale. A dynamic  calibration procedure is described
 by the authors. Synthetic air-SO2  mixtures  are introduced into the instrument
 and  into a bubbler containing the hydrogen peroxide reagent simultaneously.
 In this manner  the  instrument is  standardized in terms of the acidimetric
 hydrogen peroxide method.

    Reece, White,  and Drinker 15  adapted the electroconductivity analyzer to
 the analysis of CS 2 and H2S by attaching a combustion furnace and oxidizing
 these gases to  SO2. The apparatus was also used for the  analysis of sulfur
 compounds and chlorinated hydrocarbons by Thomas et al.16 Pyrolysis and
 oxidation to HC1 and SO2 were accomplished in a silica combustion  tube by
 using a platinum catalyst.

    A detergent solution may be added to the absorbing reagent to help main-
 tain  uniformity of liquid flow and prevent the accumulation  of dust in the  flow
 system. 17

    Commercially  available electroconductivity analyzers designed for the mea-
 surement  of  SO2  in air  are  described  in  a book about  air-sampling
 instruments.18

    Recommended  instruments and procedures for automatic monitoring of SO2
 in air  are contained in published methods manuals on atmospheric sampling
 and  analysis. 19,20
               COMPARISON  OF  METHODS
                          CONSIDERED
 Manual  Methods

    Hochheiser  and co-workers 21  compared the hydrogen peroxide, iodine,
 and  fuchsin-formaldehyde methods  for  the determination  of SO 2  in air by
 taking atmospheric samples in New York City and found that all were equally
 reliable (Table 1).  In the peroxide method,  the  strongly acidic component

-------
AND MONITORING SO 2                                                 19
 in  the  air  is determined rather than acidity attributable to SO 2 alone, but
 there is little difference in the results obtained by these methods,  particularly
 in the lesser concentrations, because the major acidic pollutant is  SO 2 .
    Table 1. COMPARISON AMONG METHODS OF DETERMINING SO2,ppm
       Peroxide method        Iodine method          Fuchsin method
0.30
0.30
0.27
0.23
0.15
0.15
0.15
0.14
0.11
0.09
0.07
0.05
0.05
Average
0.15
0.26
0.23
0.22
0.22
0.14
0.12
0.11
0.11
0.09
0.09
0.07
0.06
0.05

0.13
0.25
0.22
0.22
0.22
0.12
0.10
0.09
0.09
0.07
0.06
0.06
0.05
0.04

0.12
     In accordance with a recommendation of the working party to study methods
  of measuring air pollution used by the organization for European Economic
  Co-operation, the United Kingdom delegation made comparisons among four
  recognized methods of estimating SO2 on a simple routine basis. 22  The West-
  Gaeke, peroxide, Stratman,  and direct  iodine  methods were compared by
  means of  atmospheric  samples obtained at  a measuring site in a  city center
  during the winter period.

    Table 2 shows that the hydrogen peroxide and the West-Gaeke methods gave
  fair agreement when  used  in  routine measurement. The Stratman method
  yielded erratic results in that hi some instances it gave good agreement with
  the two methods mentioned  above, but sometimes its results varied considerably
  from those of the other two. Table 2 shows the tendency of the iodine method
  to read  consistently high.  This tendency  appears to  be owing to carryover
  of iodine from solution by the air stream.

-------
20                                            METHODS OF MEASURING
     Table 2. COMPARISON AMONG METHODS OF DETERMING SO2,
                            ^g/m3 (Ippm-S.OOO^g/m3)
      West-          Hydrogen           Stratman            Direct
      Gaeke          peroxide                                iodine

       465              475                495                 605
       575              630                180                 930
       245              200                285                 410
       245              210                430                 390
       425              330                465                 490
       380              450                480
       840              785                850               1,090
       150              205                215                 400
       245              270                160                 530
       310              240                235                 295
       730              620                215                 880
       175              155                 80                 240
      Average
       390              368                298                 550


     Welch and Terry23 reported that values  obtained  by the H2O2 method
  were approximately  30 percent  lower than values found by the West-Gaeke
  method for concentrations  up to  0.65  ppm.  The relationship between the
  West-Gaeke and the H2O2 method was established by using synthetic mixtures
  of SOj in  air  and a  dynamic calibration procedure.  It is believed that this
  should be investigated further since the collection efficiency of the H2O2 method
  is generally considered to be greater than 90 percent. Relatively close correla-
  tion between the electroconductivity analyzer and the West-Gaekeprocedure was
  obtained on synthetic air-SO2 mixtures in the range of 1  to 3 ppm.
     Paulus, Floyd, and Byers 24  compared the polarographic method of analysis
  with the fuchsin-formaldehyde colorimetric method. The polarographic results
  consistently  averaged from  2  to  7 percent less than the colorimetric results,
  depending upon the concentration range.  A wider deviation was obtained in
  the lower concentration range.

     Terraglio  and  Manganelli 4   studied the  variability  of the H2,O2, West-
  Gaeke,  and  iodine methods.  A  comparative study over  the concentration
  range of 0.10 to 1.3 ppm SO2 prepared synthetically in a  test chamber showed
  that the average results obtained by the acidimetric and  colorimetric methods
  were approximately the same but results of a single determination could vary
  significantly.

-------
AND MONITORING SO 2
                                                                          21
      Comparative recoveries by  the  iodimetric method were found to be lower
   than the values obtained by the above-mentioned methods. The relative per-
   centage recovery by  the  acidimetric  method compared with  the West-Gaeke
   method was  106 percent at  a concentration of 0.1 ppm and decreased to
   97  percent  at 1.3 ppm.  For  the  iodimetric method, the relative percentage
   recovery increased from  70 percent  of  the colorimetric value at 0.1 ppm to
   84  percent at  1.3 ppm SO2. The presence of ozone-oxides of nitrogen inter-
   fered with all three methods as follows:

      a.   Colorimetric method resulted in reduced recovery owing to bleaching
         of the sample by oxides of nitrogen and ozone.

      b.  Acidimetric  method resulted in  increased recovery owing  to titration of
         oxides of nitrogen dissolved in trapping solution.

      c.  Iodimetric method resulted in increased recovery owing to variation in
         the method of oxidation of the sulfur dioxide (Table  3).
       Table 3. INFLUENCE OF OZONE-NITROGEN OXIDES ON THE
                             DETERMINATION OF SO2
Ozone
added,
NO 2
added,
SO2
added,
ppm
0.01
0
0
0.01
0.08
0
0
0.07
0
0
0.07
0.25
0
0
0
0.47
0.46
0.47
0.46
0.47
0.46
SO2
Colorimetric
0
0.47
0.46
0.44
0.38
0.47
0.46
recovered,
Iodimetric
0
0.25
0.38
0.20
0.36
0.26
---
ppm
Acidimetric
0.03
0.48
0.44
0.53
0.69
0.49
0.51
      Selection of the manual  methods  currently used  by the Division of Air
   Pollution  was  based upon  an evaluative review of available literature in-
   formation reported herein and upon the experience of field investigators.

      Reference  to  Table  4 and the material that  follows it shows that of all
   the manual methods reviewed the West-Gaeke and hydrogen peroxide methods
   best approach the criteria for evaluation set forth herein. The  West-Gaeke
   method has  a higher sensitivity, but its calibration and analytical procedure
    are more  difficult.  The absorbents  used and the products formed in the sep-
   aration and  concentration of  the pollutant from the air mass are stable in
   both methods; though  the product formed in the hydrogen peroxide method
   is  more stable;  either  one  may, therefore, be employed with automatic se-

-------
Table 4.   COMPARISON AMONG MANUAL METHODS FOR  THE ANALYSIS OF SO-,  IN AMBIENT AIR
Method
Colorimetric
1. West-Gaeke
2. Fuchsm-
formildehyde
3. Stratman
4, Barium
chloranilate
Impregnated test papers
5. Zinc nitro-
prusside


6. Astrazone pink

Sulfate ion
7. Diazo dye
8. Thorium
bo rate -amaranth
9. Ultraviolet
determination

Acidimetric
10. Hydrogen
peroxide
11. Polarographic
lodimetric
12. Iodine

1 3. Direct iodine

14. Iodine thiosulfate

15. Turbidimetric

Cumulative
16. PbO 2 candle

Sensitivity,
ppm

0.005
0.01

0.01
0.05


0.5



0.05


0.20
0.5

0.03



0.01

0.02

0.01

0-. 10

0.40

0, 40


0. 02 mg SO3 /100
. ,--.2/rtay
Accuracy, -f %,
combined error
sampling and
analytical
technique

10
10

--
--


20



--


10
20

--



10

10

10

15

10

10


15

Reproduci-
bility, + %

3
3

--
--


10



--


3
10

--



3

3

3

5

3

3


7

Inter- Sampling
ferences rate, Collection
(known) liters/min. efficiency, %

NO2, O 1 to 4 95 to 100
NO , O , S, 1 to 4 90 to 100
SH, S2O3
H O, H?S, 1 to 5 90 to 100
PO4, F, Cl 20 to 30 90 to 100


HC1,H2S, 0.30
CS2
only at
high cone
	 0.25


P04,C1,S03 1 to 4 90 to 100
F, PO4,HCO3,SO3 1 to 4 90 to 100

H2S, H.,S04



Acid and alkaline 3 to 28 90 to 100
gases
Oxidants 24 90 to 100

Oxidizing and re- 20 to 30 85 to 100
ducing materials
Oxidizing and re- 3 85 to 100
ducing materials
Oxidizing and re- 3 85 to 100
ducing materials
Fluorides 20 to 30 90 to 100
Particulates
.
--{2 weeks' to 3
•rjonths' exposure)
Sta-
bility Stability
of ab- of pro-
sorb- duct,
ents, days
days

90 3
90 2

90
30 90


90 2 hours



14 Head
directly

90 90
90 90

	



30 90

90 2

90 2

1 Read
directly
1 1

90 90


90 90

to
to




















S
W
ffi
o
o
CO
O

2
M
3>
en
d
73
3
O


-------
AND MONITORING SO 2                                               23
  quential sampling equipment.  Comparable results are obtained by sampling
  and  analyzing known  amounts  of  synthetic mixtures of SO 2 in air. Both
  methods  are  nonspecific, though  the West-Gaeke method is more selective,
  and further studies concerning the relative effects of interfering materials are
  necessary.
  Instrumental Methods
     The performance parameters of the automatic monitoring instruments are
  summarized in  Table  5. Of the analyzers currently commercially available,
  some models of those employing the electroconductivity detection principle
  best  approach the criteria for evaluation. Any instrument employing other
  methods  of detection,  such as  potentiometry,  photometry, and ionization,
  that  meet the performance criteria established for the electroconductivity ana-
  lyzer  would also be  acceptable. The  experience of field investigators has
  demonstrated that the operating parameters specified for the automatic instru-
  mental method  selected  for use that are detailed herein are necessary for
  satisfactory  operation  of the  equipment over a reasonable period of unat-
  tended operation.

     The method is nonspecific since any soluble, electrolyte-forming gases and
  solids affect the result.  It has, however, been •found that, except in special
  situations, the measurements recorded by these analyzers predominantly result
  from atmospheric SO2.
MANUAL  METHODS  CONSIDERED  IN  ADDITION

             TO  RECOMMENDED  METHODS


 Colorimetric Methods


 FUCHSIN-FORMALDEHYDE METHOD25

    SO 2 in the atmosphere is removed and concentrated by scrubbing through
 a sampling solution of 0.1  normal sodium hydroxide and 5 percent glycerol.
 The  subsequent  determination of SO2  is  based on a color reaction first
 developed by Steigmann.26 The chromogenic reagent consists of a mixture
 of basic fuchsin, sulfuric  acid,  and^ formaldehyde, which develops a red-violet
 color in the presence of sulfurous acid.

-------
Range Sensi-
Prmciple of operation3" Absorbent covered, tivity,
ppm ppm


1. Analyzer "A" De-ionized water 0 to 2 0.01

2. Analyzer "B" Hydrogen peroxide- 0 to 5 0.05
sulfuric acid
3. Analyzer "C" 11 0 to 2 0.01
4. Analyzer "D" 0 to 1 0.02
Potentiometric Potassium bromide- 0 to 10 0. 10
sulfuric acid

Photometric
1. Analyzer "A" Starch-iodide 0 to 1 0.01
2. Analyzer "B" . . 0 to 1 0, 01

3. Analyzer "C" 0 to 0.5 0.01
4. Analyzer "D" Pararosaniline- 0 to 5 0002
formaldehyde

Air ionization Aerosol formation 0 to 10
a. Described in what follows.





Reprodu- Stability
cibility, Interference of reagent,
ppm days

Extremely stable; reagent
0.02 Soluble ionic materials (i.e., is circulated through
acids, bases, salts) purification system.
0.10 " 30

0.02 M 30
0.05 11 30
0. 20 Oxidants and reductants Extremely stable; reagent
i.e., H S, mercaptans, is circulated through
olefins , phenols

Oxidants and reductants 7
i.e., H2S, N02, etc. ?
1 1
II 7
NO2 and O3 if present 14
in appreciable
quantities
--






















2
w
H
SB
O
a
en
O
g
&
en
a
1— 1
z
o

-------
AND MONITORING SO2                                                  25
    The  absorption  maximum  is  at  570 millimicrons,  and  the color  is
 independent of temperature in the range 23 to26°C.  Above this range the effect
 of temperature  on the absorbance of the  sample gives increasingly erroneous
 results. A 30-minute color development thermostatted at  25°C was suggested.
 The extent of color stability for longer periods was not reported.

    Inorganic sulfides, thiols, and thiosulfates interfere  with the determination
 and may be removed  as a mercury precipitate by treatment with saturated
 mercuric chloride prior to the addition of the colorimetric reagent.27
    The method has a sensitivity of 0.01 ppm with a 40-liter air sample scrubbed
  through 10.0 milliliters of absorbing solution. The minimum amount detectable
  is 0.1 microgram per milliliter of solution. At a sampling rate of 20 liters per
  hour with a midget fritted bubbler, the collection efficiency is close to 100 per-
  cent. Stang et al., 28 using a Greenburg-Smith all-glass impinger containing
  100 milliliters of 1  percent glycerol in  0.05  normal sodium hydroxide and
  sampling  at  1 cubic foot per minute,  obtained results  indicating from 93 to
  99 percent efficiency based on the amount of  862 collected in two  impingers
  in series.

    Moore, Cole,  and Katz     reported that NO2, if present in the  same con-
  centration  range as SO2, produces  a  negative interference owing to the
  bleaching effect of NO2  on the fuchsin-formaldehyde sulfite color. Corrections
  for the effect  of  NO2  on the colorimetric  S02 values were obtained by con-
  current determination of SOa  by the conductimetric and colorimetric methods
  and of NO2 by the Saltzman method.

    Paulus, Floyd, and Byers 24 found it necessary to establish  a new standard
  curve with each new batch of colorimetric reagent. They also  reported on the
  effect of aging, light, and agitation on the  collected sample. Solution strength
  was determined  after  the first,  third, and sixth days of sample collection.
  Average losses of 6 percent were found after 6  days. The samples that showed
  losses  after the  first and third days were  mostly in the lowest concentration
  range.  No losses due to sunlight or artificial light were found. The effect of
  agitation  is  important  when  samples are shipped  by mail. Losses due to
  agitation  are comparable  to those due to  the  aging experiments and are,
  therefore, not hastened by this condition.

    The  fuchsin-formaldehyde  method for the  determination of SO 2 in air is
  recommended and reported in the ACGIH manual of recommended methods. 30
  An average  deviation of 8 percent was found.  According to ACGIH, when
  concentrations of nitrogen  oxides higher than those of SO2  are anticipated,
  the polarographic method should be used.

-------
26                                               METHODS OF MEASURING
  STRATMAN METHOD31

     Atmospheric  SO2 is initially absorbed on silica gel and then reduced with
  hydrogen to H2S on a platinum  contact catalyst at 700 to 900°C. The H2S
  formed is passed into a bubbler containing 2 percent ammonium molybdate
  in 0.4 normal sulfuric acid.  The resulting blue-violet molybdenum complex
  is then determined colorimetrically with the aid of a Ziess Opton S57 filter.
  According  to  the amount of reagent used, 1 to  200 micrograms of SO2 can
  be determined. Efficiency  of removal of atmospheric 862 is  higher than 90
  percent at a  sampling rate up to  5 liters per minute. If the quantity  of air
  sampled contains more than  300 milligrams of water, it must be dried by
  passage  through  a preliminary drying  tower containing phosphorous pento-
  xide.  It was found that pressures  up to 200 millimeters Hg and temperatures
  between  18 and  40° C  do not affect  the collection efficiency. The method has
  a sensitivity of 0.01 ppm with a  40-liter air sample when  the H2S produced
  in the catalytic desorption process is absorbed in 10  milliliters of the ammon-
  ium molybdate reagent.  The interference of SO3 is eliminated by preceding the
  silica gel absorption equipment by a bubbler containing phosphoric acid. The
  Stratman method is particularly  suitable for relatively short-term sampling
  since moisture interferes when samples are obtained over a longer time.
                                        on  QQ
  BARIUM CHLORANILATE METHOD   '

     The method is based on  the  reaction of solid  barium chloranilate with
  sulfate ion  at pH  4  in  50 percent ethyl alcohol to liberate highly colored
  acid-chloranilate  ion.  The concentration is determined spectrophotometrically
  with the absorption peak at 530  millimicrons. Atmospheric SO2 is removed,
  concentrated, and oxidized to sulfate by scrubbing through 0.5 percent aqueous
  H2O2 solution.  A buffer  solution,  95 percent  ethyl alcohol and 0.1  gram
  barium chloranilate, is added  and the mixture is shaken for 10  minutes. The
  excess barium chloranilate and the  precipitated barium sulfate  are removed
  by filtration. The method has  a sensitivity  of 0.05 ppm with  a 1,000-liter air
  sample scrubbed through 25  milliliters  of absorbing solution. Kanno 32
  reported a collection efficiency  of close to 100 percent at a sampling rate of  5
  liters per minute. The residual presence of H2O2 and CO2 did not  interfere with
  the colorimetric method.  The  accuracy  and the  effect of interfering materials
  were not reported. Phosphates, fluorides, and chlorides are known to interfere
  hi the chloranilate procedure, and a preliminary separation would be required.
  A method for the conversion of the gravimetric  lead peroxide method to
  colorimetric with the use of barium chloranilate is also  described by the author.

  ZINC NITROPRUSSIDE METHOD34

     Air is aspirated  at the rate of 300 milliliters per minute through a test
  paper supported in  a  suitable holder. The determination of SO2  is based on

-------
AND MONITORING SO 2                                                  27
  the formation of a brick-red stain, which is compared with a standard stain
  chart or with a disc of standard tints. The test papers are prepared by soaking
  strips  of filter paper in  ammoniacal  zinc  nitroprusside solution and drying
  at a temperature not exceeding 40°C. The dried test papers are stored in a
  stoppered container in the dark and  are moistened by spraying with  water
  immediately before  use.  It was found that  dry zinc nitroprusside papers are
  stable indefinitely  whereas  papers  prepared with glycerol decompose  after
  about 4 weeks of storage.

    With  a 360-milliliter air  sample, the range of concentrations determinable
  with an accuracy to within ±20 percent is 1 to 20 ppm.  Hands and Bartlett34
  stated that concentrations outside these limits can be measured by increasing
  or decreasing the size of the sample. The stains  are stable for about 2 hours.
  The effect of hydrogen chloride, hydrogen  sulflde, and carbon disulfide on
  color  production was investigated. Hydrogen chloride produced a discoloration
  while  hydrogen sulfide  produced  a positive stain,  although  this effect was
  observed  only  at relatively large concentrations.  Carbon disulfide  at con-
  centrations up to 500 ppm yielded no stain.

    In  an  investigation  of the starch-potassium iodide-potassium iodate  test
  paper method35 the authors  found that, unless strict control was exercised
  during the preparation of the test paper, nonuniform  papers  resulted.  This
  made  the calibration of the procedure difficult.

  ASTRAZONE PINK METHOD36

    Air is drawn through a wet test paper at a rate of 250 milliliters per minute
  until bleaching of the impregnating reagent  is effected, whereupon the volume
  of air taken to produce this  effect is noted. The  test reagent consists of Astra-
  zone Pink FG (Bayer), sodium bicarbonate, and glycerol and is stable in the
  dark for  2 weeks. The  dye solution alone is unchanged after several months.
  One drop of  this reagent is  placed on the filter  paper and allowed to spread
  before sampling. The method has a sensitivity of 0.05 ppm SO2. No informa-
  tion is reported regarding the specificity or accuracy of the method.

  DIAZO DYE METHOD
     Klein37 described a  colorimetric method for the determination of sulfate
  ion that  may be applicable to air  analysis. The sulfate in the  sample is pre-
  cipitated  as benzidine sulfate with benzidine hydrochloride, purified, dissolved
  in 0.2 normal HC1, diazotized, and coupled with N-1-naphthylethylenediaminedi-
  hydrochloride after excess nitrous acid is destroyed. The resultant purple color
  is read  in a colorimeter with a green filter. The color produced is stable for
  at least  12 hours. Phosphate and chloride interfere in the  determination. Phos-

-------
28                                              METHODS OF MEASURING
 phate  should  be completely  removed.  If  the  ratio  of chloride to sulfate is
 greater than  30, incomplete  -precipitation  of  benzidine sulfate occurs. The
 sensitivity of the method is 0.05 milligram in 15 milliliters of sample solution,
 or  to  translate this to air sample concentration, the sensitivity would be 0.2
 ppm for a 90-liter air sample collected in a midget impinger containing H202-
 absorbing reagent. Comparison with the gravimetric sulfate method shows good
 correlation of results. The analytical error is ± 2 percent.

 THORIUM BORATE-AMARANTH DYE METHOD

   A colorimetric procedure for determining  sulfate ion, described by Lambert
 et al.  f8  uses  an insoluble thorium borate-amaranth  dye reagent. This pro-
 cedure  may be  applicable to air  analysis. Sulfate releases  dye molecules from
 the  solid reagent in direct proportion to the concentration of the dye measured
 at  521  millimicrons. Interference by fluoride, phosphate, and bicarbonate is
 eliminated  through the use of lanthanum ion and a weak acid cation exchange
 resin.  It was found that the color  produced was independent of the amount
 of  reagent added and  a function only of the sulfate ion in solution.  Never-
 theless,  it was  found necessary  to  prepare  a calibration curve for each  new
 batch of reagent. Reaction  time is not a factor in the range 1 to  10 minutes.
 With a  10-millimeter cell path, the sensitivity of the method  is 0.2 milligram
 sulfate  contained in 10 milliliters of sample with a reproducibility of 10  per-
 cent. For a 90-liter air sample obtained by collection and  oxidation of  SOa
 to 864 the sensitivity would be 1 ppm. No doubt the sensitivity of the method
 could be increased by increasing the cell path.
 INDIRECT ULTRAVIOLET DETERMINATION OF SO2 BY
 MEANS OF PLUMBOUS ION 39

   SO 2  is  precipitated as lead sulfite with 1 milliliter of lead  acetate solution
 (100 jig/ml) and is determined  indirectly as the plumbous ion remaining in
 solution  at 208 millimicrons.  Plumbous ion  absorbs  at  208  millimicrons
 (molar  absorbance  8,210) and  this permits a  sensitivity of  13  micrograms
 of plumbous  ion per  0.1  absorbance  unit  (equivalent to  8 fig S02)  to be
 attained. The lead precipitate is sufficiently insoluble in water at room tempera-
 ture for plumbous ion not  to be spectrophotometrically determinable from its
 saturated  aqueous solution.40  This method would be sufficiently  sensitive for
 atmospheric analysis and may be applicable to  air analysis by scrubbing an
 air sample through a suitable  absorbent and subsequently using the indirect,
 ultraviolet, spectrophotometer method.

                       On
   Blinn and Gunther    found that under the conditions specified, two moles
 of  SO 2  are needed  to precipitate 1 mole of plumbous ion. Acid gases such
 as  H2S and H2SO4, which precipitate plumbous ion from aqueous solutions,
 and  other materials,  which absorb strongly in the 250-to 190-millimicron
 region, would interfere.

-------
AND MONITORING SO 2                                                  29
  FERROUS-PHENANTHROLINE METHOD

     Stephan and  Lindstrom 41 developed a direct color method for determining
  SO 2 in the atmosphere. The air sample is passed through a wet scrubber con-
  taining ferric-phenanthroline reagent. Sulfite ion oxidizes ferric ion to ferrous
  and produces the colored ferrous-phenanthroline complex ion.

     The color produced is stable for about 3 days. Nitrogen dioxide and ozone
  do not interfere. The lower  limit of detection is 0.05 ppm for  a 100-liter air
  sample collected for 50 minutes  in  75 milliliters  of absorbing reagent. A
  serious disadvantage  of the method  is  that the relatively high temperature
  sensitivity during  sampling  necessitates the use of a constant water bath or
  other regulatory device to control temperature to ±  0.5°C. Because of the
  stability  of  the  product formed, this method  can,  however,  be adapted for
  use with a sequential sampler or for collection of  an integrated 24-hour sample
  provided that the sampling unit is thermostatted.

  POLAROGRAPHIC METHOD 24

     SO 2 in .air is removed and concentrated by scrubbing  through an absorbing
  solution,  contained in a standard, all-glass, Greenburg-Smith impinger, con-
  sisting of 2  percent glycerol in 0.05 normal sodium hydroxide at a rate of 24
  liters  per minute  for 30 minutes. Subsequently an acetate buffer (pH 4) is
  added, and  the combined solution is deaerated  by bubbling  nitrogen through
  the sample contained in an air electrolysis vessel. The flow of nitrogen  is then
  stopped,  and  a  polarogram  is made from -0.35 to -1.00 volt. A sensitivity
  of 0.006 microampere per millimeter is used.

     For a 30-niihute air sample  contained in  75 milliliters  of absorbent, the
  sensitivity is 0.02 ppm. Smaller concentrations could be determined, depending
  largely on  the  ability  to measure the  inked  lines on the polarogram. Air
  samples  can be determined  with  an accuracy of —  10 percent and a repro-
  ducibility of 3 percent.

                                        42
     According to  Kolthoff  and Miller     only one of the two tautomers of
  sulfurous acid is reducible at the dropping mercury electrode, and at pH 4,
  there  is less of  the  reducible than of the nonreducible  tautomer present. The
  fuchsin-formaldehyde colorimetric determination involves a similar situation
  in that only one of the tautomers results in the formation of a red color. The
  same conditions of collection efficiency and reagent stability also apply  since
  the scrubbing media used in the methods are identical.

     Sulfur compounds generally do not interfere with the polarographic method.
  Cystine does not interfere with the analysis but reacts with the SO2 while in

-------
Qn                                            METHODS OF MEASURING
oU
the collecting medium. (The disulflde group is reduced to sulfhydryl by SO2.)
Nitrites  are  reduced at the dropping  mercury  electrode at a more negative
potential than  SO2  and  presumably  would interfere only  if present in very
large concentration.
 lodimetric Methods
 IODINE METHOD 43

   SO 2  in the air is  removed and concentrated by scrubbing through a 0.1
 normal sodium hydroxide solution contained in a midget or standard impinger
 at a  sampling rate of 0.1 and 1 cfm respectively.  The solution is subsequently
 acidified and  titrated with  a standard 0.001 normal iodine  solution. This
 method  of sampling is not particularly suitable for field work unless a mobile
 laboratory  is  available that  permits  titration of the sample within 24 hours
 after  collection.  Negligible losses  of  SO2  strength result from  the storage of
 the sample solution in the concentration range reported for a 24-hour period. 21

   For  a 30-cubic-foot  air sample collected at  1 cfm for 30 minutes,  a sen-
 sitivity of 0.01 ppm is  obtained. Thereproducibilityfor the combined analytical
 procedure is 10 percent in the  range 0.10 to 1.3 ppm with the relative standard
 deviation decreasing  with  increasing  concentrations. 4   It was  found that re-
 coveries by the  iodimetric method in the  range of concentration of 0.10 to
 1.3  ppm were  lower  than  by  West-Gaeke or  hydrogen peroxide methods.
 Oxidizing materials such as nitrogen  dioxide and ozone would result in a de-
 crease in recovery, or amount of SO2 measured,  and the presence of reducing
 materials such as H2S in  the airstream would result in a positive interference.

 DIRECT IODINE METHOD

                         44
   Pearce and Schrenk      described a method in which SO 2 in air is ab-
 sorbed  and oxidized  directly in  a neutral starch-KI-iodine solution. Air is
 sampled through a midget impinger  at 0.1  cfm. The time required for dis-
 appearance  of the  blue color is  noted and  the amount of SO2 present is
 calculated  from  the amount  of iodine originally present. If sampling time is
 too long, the  results  are  relatively high owing to loss of iodine by aeration
 of the absorbing solution. The loss is decreased with increasing amounts of
 KI.  A recovery  of  better  than 83 percent  is reported  and is independent of
 concentration.  The  method is nonspecific  and is applicable in the presence
 of relatively small amounts of other oxidizing or reducing materials. A sensi-
 tivity  of 0.1  ppm  is attained  for  a 0.5-cubic-foot air sample collected for
 5 minutes.

-------
AND MONITORING SO 2                                                  31
    Kitano  45 utilized a second bubbler downstream connected in series and
  containing  standard thiosulfate solution.  Exposure to carbon dioxide and
  oxygen did  not  affect the strength of the thiosulfate solution. In this manner
  any  iodine  lost  during  aeration  would be  trapped by the second bubbler
  and  could  therefore be accountable. The bubbler containing the iodine-ab-
  sorbing reagent was maintained at 0°C.  All-glass tubing was used to prevent
  any  losses   of volatilized  or  entrained  iodine solution. If a hand pump is
  used,  the method is useful for  a  rapid, direct approximation of 862 levels
  in air since no accessory equipment is required in the analysis.

  IODINE-THIOSULFATE METHOD 36>46-48

    S02 in air is collected by absorption in  a fritted bubbler containing 50
  milliliters of 0.005 normal iodine at  5 liters per minute for 10 to 30 minutes.
  The amount of iodine consumed is proportional to the amount of SO 2 collected
  and is determined by titration with 0.005 normal thiosulfate solution. It was
  found that   the  error  due to  evaporation of liquid is relatively small as is
  that due to loss of iodine if 25percent KI is used.  For a 30-minute air sample,
  a sensitivity of 0.4 ppm  is attained.  This may be increased by  reducing the
  strength of the iodine and thiosulfate reagents. Griffin and Skinner 48 suggested
  the use of an air blank obtained by simultaneously sampling through a bubbler
  preceded by a soda-lime tower to remove all SO 2 at the source.

  TUBBIDIMETRIC  METHOD  49>5°

    Matty and Diehl 50 used a phototurbidimetric method developed by Corbett 49
  for the measurement of SO2 in flue gas. SO 2 is removed from the gas stream
  by absorption in hydrogen peroxide reagent and is subsequently determined
  as sulfate turbidimetrically in an acid isopropyl alcohol medium. The turbidity
  is determined 5  minutes  after  the addition of 1 milliliter of 1 percent BaCl2
  solution. Since optical density changes with time owing to the degree of com-
  pleteness of precipitation and changes in crystal size, the time elapsed between
  the addition of precipitant and photoelectric measurement is critical and should
  be standardized.  The range of concentration determinable is 0.1 to 0.4 milli-
  gram of  SO 4 per 25 milliliters when photometric measurement is made with
  a light path of  25  millimeters. A very close correlation with the gravimetric
  sulfate procedure was obtained over this range of concentration. A 30-cubic-
  foot gas  or air  sample may be determined with a sensitivity of 0.4 ppm by
  using these  calibration data. The sensitivity  of the method  may be increased
  by increasing the length  of the cell path or by sampling for a longer period
  of time to obtain a larger sample. The addition of barium chloride in solid
  rather than in solution form was found to yield a more lasting suspension of
  barium sulfate.  51   Volmer and Frohlich 52 stabilized the suspension with an
  acid-glycerine-alchol-gum arable mixture  at a pH of 3.2.

-------
                                               METHODS OF MEASURING
   Mathers  40  described a turbidimetric method in which SO2 is collected in
a  neutral 1  percent lead acetate  solution.  The  amount  of PbSO 3 formed is
determined turbidimetricaUy  at  a  wavelength of 600 millimicrons. The sen-
sitivity of the method is 0.1 milligram as SO3. Materials  other than SO2 that
form insoluble compounds with lead ion interfere in the determination although
PbSO 3 has  a  lower solubility  product  than do  lead  derivatives of other
commonly occurring gases and aerosols such as hydrogen sulfide and sulfuric
acid.
Cumulative Methods
LEAD PEROXIDE CANDLE METHOD

   The lead peroxide (PbO2) method of measuring the extent of atmospheric
pollution  by SO 2 was  developed in England by the Department of Scientific
and Industrial Research (DSIR) in  1932  and has been used extensively there
ever since; it has been applied only recently in this country.  '

   The object was  to provide an index of the activity of SO 2 in the atmos-
phere  as  a  measure  of an aging  effect on fabric and  buildings and of its
effect  on  the  corrosion  of  metals.  The method is based on measuring the
sulfation  caused by gaseous  SO2  in ambient air  by exposing PbO2 paste.
It is a cumulative method similar to the usual measurement of dust fall. The
candle used in England consists of a porcelain cylinder about 10 centimeters
in circumference. A  10 x 10-centimeter piece of cotton gauze is wrapped  around
the  porcelain form  as  reinforcement and the active reagent is applied. The
active reagent is applied in  the  form  of  a paste consisting of 8 grams of
PbO2  in  about 5 milliliters  of a gum tragacanth  solution prepared by dis-
solving the gum  in ethyl alcohol and diluting with distilled  water. The candle
is exposed in  a  shelter,  which protects the reactive surface from rain, for a
period of 1 month; shorter or longer exposure periods may be used depending
upon the SO2 activity of the atmosphere. After exposure, the material is stripped
from the  candle  with sodium  carbonate,  and the amount  of sulfate is deter-
mined by  the standard gravimetric procedure. The results are reported as milli-
grams of SO3 per 100 square centimeters of PbO2 per day.

   Wilsdon  and McConnell 55  indicated  that  the rate of sulfate formation is
proportional to SO2 concentrations in the atmosphere,  at least up to  15 per-
cent conversion of  the reactive material. From experiments in a wind tunnel,
the rate of  reaction was  found to vary inversely  as the 4th root of the wind
velocity. An increase in  temperature  of 1°C increased the reaction rate about

-------
AND MONITORING SO 2                                                 33
 0.4 percent. The reaction rate also increased considerably when the surface
 was  wet. Conversion to PbSO4 was found to be a function of PbOa particle
 size.  The  authors noted a  change  in  reaction rate  with different batches of
 Pb02.  In the work done in  England, a large batch of PbO2 sufficient to last
 for several years  was obtained.  The  results obtained by this method corre-
 lated very well with data secured by other methods (volumetric) when the results
 were corrected for wind  velocity and temperature.  These experiments were
 conducted at 862 concentrations  (30 to 300 ppm)  much larger  than those
 in the atmosphere.

    Parker  and  Richards 56  estimated that errors of sampling and analysis
 are about 10 percent. Eight PbO2 cylinders  were exposed simultaneously for
 a  period  of 6  months during 1948-1949.  The mean rate of sulfation  was
 2.6 milligrams of SO 3 per 100 square centimeters per day and the standard
 deviation of one observation was 7 percent of the mean.

    Thomas  and  Davidson  57  employed PbO2 cylinders  to  obtain relative
 sulfation values at selected  sites  in  the vicinity of large, coal-burning steam
 plants. No deterioration in the rate of reactivity of PbO2 with SO.2  was noted
 in periods  of  exposure as long a-, 4 months. A relatively low degree of  cor-
 relation was obtained for sulfa.'!in rates and 862  dosage as measured by
 the Thomas Autometer. This %j.s attributed principally to the typically low
 average SO 2 dosage at most sites near a single source.  At the site of maximum
 exposure, the average SO2 concentration was 0.02 ppm. Rather than  compare
 field values with a sealed laboratory control, control cylinders were  operated
 at remote sites  60 to 70 miles distant from a sulfur dioxide source. Sulfation
 rates varied from  0.02 to 0.04 milligram  of SO 3 per 100 square centimeters
 per day. This  value  of about 0.03  milligram  of SOs per 100 square centi-
 meters per day is considered to  be a realistic value for clean air, which is
 an order of magnitude greater than that established from sealed-source  lab-
 oratory controls. The basic cylinder employed at TVA consists of an ordinary
 8-ounce,  short-form glass jar. Eight grams of PbO2 paste are painted on a
 100-square-centimeter band  of  cotton  gauze stapled  around the  glass jar.
 Freshly coated cylinders are dried overnight in a desiccator and screwed into
 the smaller of  two concentric, bradded and soldered-metal jar tops. This
 assembly is screwed into a wide-mouthed, 32-ounce glass jar, which  serves  as
 a  convenient carrier for shipment  and storage. At the site of the field station,
 the inner  small  jar  with  PbOa  coating is inverted and screwed into a jar
 top permanently mounted on the base of a  louvered shelter. The shelter is
 mounted on a 4-foot post. Supports such as utility poles  may be used.

    Foran  et  al.  58  found that  measurement of SO2 activity by means of
 PbOj  candles  was well  suited to measuring  relative concentrations of SO2  in
 conjunction with metal corrosion studies. Severity of corrosion of  zinc and stain-

-------
                                               METHODS OF MEASURING
less steel panels closely correlated with SO2 dosage as measured by the Pb02
candle method. Wilkins 59 compared the relative values of SC>2 concentrations
determined by the H2O2 and the PbO2 candle methods. A close correlation of
sulfation rate with SO2 concentration  was  obtained. The conversion factor
from  ppm  to sulfation rate  was  milligrams per  100 square centimeters per
day  x  0.04 = S02,  ppm.  It was found that the factor by which the Ph02
reading must be multiplied to  give the concentration of SO2 varied from month
to month at any given site and from site to site for any given month.8 Yearly
means for each  of seven sites in and around London were obtained by divid-
ing the concentration  of  SO2 in  micrograms per cubic meter by  the rate of
sulfation of PbO2 in milligrams SO3 per  100 square centimeters per day by
the H2O2 method. The yearly means of the sites varied from 63 to 172 with
an average value of  112. These results show no simple connection between
concentration of SO2  and PbO2  readings.  The author concluded that Pb02
readings at any one site  should not be used to give an indication of change
in concentration from one month to another. Comparisons between one year
and another  are also not very precise, though they may be useful in defining
areas of gross  pollution. Nevertheless, if trends over a number of years are
considered,  for  example,  by a comparison of one  5-year average with the
next  or of 5-year running averages, the variation  due to wind, weather, etc.,
tends  to become small and the measurement of SO2  more precise;  the 5-year
average has long been recommended by the DSIR for this purpose.

   In any district of limited size, for example, the area surrounding a particular
source such as a power station, it is a reasonable assumption that the cylinders
would be exposed to the same weather conditions — wind, humidity, temperature—
so that the rate of  sulfation should bear  the same relation to concentrations
of SO 2 for  each.  The pattern of  concentration so  obtained should, therefore,
be valid, even though the  absolute value for each month can be  obtained only
by comparison with data obtained from other SO2apparatus.

TEST PAPER METHOD

   Buck 60  described  a  cumulative  method for  the  determination of S02.
Filter paper  is impregnated  with  a  mixture  of KHCOs - H2O - glycerine and
the treated  paper is then exposed to ambient air for  100 hours.  After exposure
the material  is stripped from the filter paper, and the amount of sulfate is
determined by the standard  gravimetric  procedure.  Further  investigation is
necessary to determine sulfation values relative to SO2  concentration.

   Pate et al .61 used membrane filters impregnated with potassium bicarbonate
to collect SO2  in  aspirated air  samples.  A 95 percent collection efficiency
was obtained in the concentration range 0.1 to 10 ppm.

   Hugen 62  sampled SO2 in air using Whatman No.  1  filters impregnated
with potassium hydroxide and glycerol. A collection efficiency of greater than
95 percent  was  obtained  at relative  humidities of  greater than  25 percent,

-------
AND MONITORING SO 2                                                35
  but below  25 percent, collection efficiencies  dropped sharply.  SO2 was sub-
  sequently  analyzed  by washing the  filter  with water, neutralizing with HC1,
  and applying the West-Gaeke  colorimetric method.  The stability of the po-
  tassium sulfite formed on the filter was poor in the presence of water vapor,
  but the filter could be stored for 2 weeks without appreciable oxidation if the
  filters were stored in a dry atmosphere.
                    r*n
  Detector Tubes
    Gas detector tubes  are generally nonspecific  and not sufficiently sensitive
  to measure  concentrations  of pollutants found  in  ambient ah-. The present
  state  of gas-detecting  tubes is that they are  semiquantitative devices useful
  for preliminary survey and screening in industrial hygiene work. Test kits
  are being used more  and more  in industrial hygiene investigations because
  they are  relatively inexpensive and capable of  detecting hazardous concen-
  trations immediately so that corrective measures  can be taken. Detector tubes
  are, however, subject to interferences, and the findings are frequently estimates
  only,  and abuse is, therefore, possible in  the hands  of untrained personnel.
  Techniques  for the  calibration  of  gas-detecting  tubes are  described by
  Kusnetz et al. 64
     INSTRUMENTAL METHODS  CONSIDERED
                       IN ADDITION TO
               RECOMMENDED  METHODS
 Potentiometric Method 20'65'66
   The apparatus consists of a sampling probe, flow control device, absorption-
 titration  cell,  current-generating electrodes,  oxidation-reduction-sensing  elec-
 trode system,  amplifier, milliampere recorder, and gas pump.  Air is drawn
 continuously through the titration cell at a fixed rate of approximately 1 liter
 per minute. The zero level  is automatically recorded periodically by passage
 of the  air sample  through a charcoal-soda lime filter. SO 2 in the measured
 air stream is  absorbed in an acidified  bromide solution contained in the

-------
36                                             METHODS OF MEASURING
 titration cell.  The  instrument  is initially  adjusted to generate continuously a
 comparatively low level of bromine in the  acid-bromide reagent.  A pair of
 electrolyzing electrodes is used in which bromine is generated at one electrode
 and hydrogen is evolved  in the second  electrode. Any compound  in the air
 stream that  is oxidized by bromine will proportionately reduce the initially
 selected bromine concentration. This reduction in bromine concentration changes
 the oxidation  reduction potential of the reagent, which is immediately sensed
 by the appropriate sensor electrode system.  This, in turn, electronically calls
 for generation of sufficient additional bromine to maintain the  original bromine
 concentration.  The electric current required to generate this additional bromine
 is a measure of the reducing gas in the atmosphere.

    Oxidizable sulfur  compounds  other  than SO2 such  as H2S, mercaptans,
 organic sulfides, and disulfides are recorded  by the  analyzer.  Some gases
 such  as  olefins,  diolefins, and phenolic compounds would be titrated to  a
 limited degree. The presence  of these interferences would yield relatively high
 results  for  SO2. Chlorine, bromine, chlorine  dioxide, nitrogen dioxide, or
 ozone would reduce the bromine demand and  would be manifested by com-
 paratively low results for  SO2. It would be possible to conduct a prior sep-
 aration of  an atmospheric  mixture of sulfur-containing  compounds before
 passing the sample into the analyzer. 67 An automatic  multiple-selector valve
 operating on a timed sequence could pass the air sample sequentially through
 various filters and scrubbers as follows:

     1.   Bismuth subcarbonate-H2SO4 solution to remove H2S;
    2.   potassium dichromate solution to remove H2S and SO2;
    3.   alkaline CdSO4 solution to remove H2 S, SO2, and RSH;
    4.   activated charcoal-soda lime to remove all reactive constituents.

    Pyrolysis offers  the  possibility  of converting mixtures of sulfur gases  to
 either H2S or SO2.  Field and Oldach 6a converted a mixture of sulfur com-
 pounds to H2Sbypassingasampledstream of air with hydrogen over alumina
 at  900° C.  Thomas  et al. 69  oxidized  H2S,  mercaptans, and other sulfur
 compounds  to SO2  by passing the sampled air through a  silica combustion
 tube containing an electrically heated, spiral, platinum wire.

    The range  of SO2  detectable  by  the instrument  is  0.1 to  10 ppm with a
 sensitivity of 0.1  ppm and a reproducibility of 0.2 ppm.  A 90 percent response
 to a change in concentration is effected in 30 seconds.

    Nader and Dolphin 70 have developed a circuit modification that increases
 instrument  sensitivity by  a factor  of 10.  This has, however, resulted in ex-
 cessive background  noise  level  and zero drift.  Development of an accessory
 circuit to arrange for automatic adjustment of the generating current is currently
 in  process  and  will  permit the unattended, continuous  monitoring of atmos-
 pheric  SO2  at the increased  sensitivity.  McKee  and  Rollwitz 71 pursued a
 similar modification  and increased instrument sensitivity to one scale division

-------
AND MONITORING SO 2                                                   37
  per 0.01 ppm SO2; full-scale deflection corresponded to 0.75 ppm. A potentio-
  metric  recorder was substituted for the  recording milliameter. The noise level
  was  reduced to a satisfactory level by  an electronic filter. Zero drift was still
  the greatest difficulty encountered, and occasionally no record was obtained
  when the zero point drifted off the scale.

     The apparatus  is calibrated against standard mixtures  of inert gas with
  S02. The calibrating mixtures  may be  prepared in a tank of known capacity
  pressured to from 100 to 200 psig containing metered volumes of SO,j gas.20
  SO2   cylinders prepared in this manner will decrease in  strength by 5 to 10
  percent  in  2  months and should  therefore be used within this time. Standard
  SOj  mixtures may also be  prepared in gas test chambers or  in  collapsible
  mylar  or teflon bags and then  fed into the  analyzer at the normal rate of
  flow. Several calibration points within the full-scale reading of the  instrument
  should be obtained.

     Carpenter  and  Sparkman  72 adapted the analyzer for field mobile sampl-
  ing in a motor vehicle and in a helicopter. The  analyzer  has also been oper-
  ated  in  a helicopter by  Gartrell and Carpenter  73  in the study of dispersion
  patterns by measuring S02 concentration in a plume.
  Photometric Methods
  ANALYZER "A"

    Katz    '   described a continuous, automatic analyzer for SO2 that employs
  photoelectric cells and  a recording potentiometer to indicate the increase in
  light transmission  of blue  starch-iodine  solutions after aspiration with con-
  taminated  air.  The instrument  has a  range of 0.01 to 1 ppm or more, de-
  pending upon the volume and concentration of solution in the  absorbers  and
  the  amount  of  air sample passed through the  solution. This instrument is
  not commercially available. The gas bubblers consist of pyrex salvarsan tubes
  of 350-milliliter  capacity. The  instrument operates on  a 2-minute time cycle,
  the  solution in  one absorber  being  aspirated at a measured flow rate for
  this  interval after which the air is passed through  the second absorber, by the
  opening  and closing of appropriate,  cam-operated,  poppet valves.  The as-
  piration is continued alternately for eight successive operations in each absorber.
  At the end of this, period of about 32  minutes, the solution in each absorber
  is drained  successively,  and fresh solution  is delivered from a stock bottle
  of 20-liter  capacity. The light, transmission through  the blue starch-iodine

-------
                                              METHODS OF MEASURING
oo
solution is  recorded during  the 2-minute quiescent period following each as-
piration cycle by means  of a  recording potentiometer with a scale range
of 0  to 50 millivolts.  The  change  in concentration of the blue starch-iodine
solution is determined by using two photronic cells to measure the light trans-
mission through  each  absorber  from a  constant source supplied by a 20-
watt lamp.  The  light source is mounted midway between the absorbers, and
the photo cells are  mounted behind  each  one. Output  current from the cells
is passed  through  a standard resistance  box, and the voltage drop across
the O-to-500-ohm terminals is measured by means of the recording potentiometer.

   The  recorder  is  calibrated by determining the potentiometer reading for
a  series of starch-iodine solutions varying in normality  from 8 x 10~° to  2
x  10~5  during the operation of the  analyzer  with  SO2 -free air.  Reproducible
results  are obtained on  iodine solutions of equivalent normality. Experiments
indicate that  the  starch-iodine  solution used  remains  stable after  aspiration
with  SO2 -free  air  at rates of about  10 liters per minute for periods of 30
minutes. Only a high grade  of starch, which gives the characteristic  blue color
with dilute iodine solution, is suitable for this work.  Grades that yield a purple
tinge  are unsatisfactory. The stock solutions are stable for  1 week when stored
in a  dark,  cool  room provided the vessels are maintained in a clean, sterile
condition.

   Sulfur trioxide, sulfuric acid mist or sulfate, and unsaturated hydrocarbons
do not  interfere  in  this method. Oxidizing and reducing materials such  as
H%S  and  NO2  do, however, interfere,  and  the method  is not applicable in
atmospheres  where   these materials  are  likely to  be  present in appreciable
quantities.

ANALYZER"B"

   Adams, Dana,  and Koppe 76  developed a versatile, continuous,  automatic,
photometric  analyzer based on   the dosimeter sampling principle. A small
volume  of  reagent   sensitive  to  a  particular ion  or  class of compounds  is
circulated continuously in the air-reagent contactor system until a preselected
concentration of  the  pollutant under study  is accumulated. When this has
been  reached, the instrument  automatically discharges  the spent reagent and
injects a measured volume of fresh reagent. Sampling  is then continued until
another  cycle is  completed  by the accumulation  of the standard quantity of
pollutant.  The pollutant  concentrations  may be  recorded  as  either a tape
record  of the time  necessary to  accumulate  each equivalent quantity of the
pollutant under study  or as a continuously  integrated recorder chart of the
rate  of  accumulation  of the pollutant  during  each dosimeter period. This
instrument is suitable for the automatic recording of any  pollutant for which
a suitable colorimetric reagent can be developed.

   A starch-iodine reagent is used for the photometric determination of atmos-
pheric S02. It was found that the partially bleached blue color was more un-
stable under conditions  of continuous aeration than  was the  unbleached blue

-------
AND MONITORING SO 2                                                 39
  color.  A  reasonably  stable reagent that may  be  aerated  at approximately
  0.1 cfm for periods of more than 12 hours  was produced by the addition of
  small quantities of N-acetyl-p-aminophenol  and mannitol. It was also found
  that the  blue  color  of  the reagent was extremely sensitive to temperature.
  Furthermore,  the color was found  to  deteriorate rapidly when stored in the
  dark  at elevated temperatures of 120°F. To minimize these effects, the dosi-
  meter cabinet was  thermostatted at 75°F with a refrigerative compressor to
  maintain this temperature. Light absorbance measurements are conducted with
  a  50-watt projection lamp and an optical filter peaking at 575 millimicrons.

     The prepared  reagent  is stored in a black reagent bottle in the  automatic
  analyzer. A drop of mercury is kept in the  storage bottle to act  as a  fungicide.
  The reagent  has  been found to remain stable up to  1 week when handled in
  this manner. A soda  lime-activated charcoal trap  on the  reagent bottle's air
  inlet protects the reagent from possible contamination by air pollutants.

     A countercurrent type of air-reagent-contacting system is employed in which
  a  small batch  of reagent  is continuously circulated through an optical  flow
  cell. Small evaporation losses  are continuously sensed and dropwise  additions
  of  distilled water are  automatically made to maintain the  original volume.
  The sampled  air is drawn through a  rotameter, and  the air volume is  reg-
  ulated  by  means   of a built-in air  bypass on the vacuum side  of the air
  pump. Air entering  under suction lifts the reagent, which has fallen by gravity
  through the optical  path,  up  through  the contacting column.  The scrubbed
  air is exhausted through a tube at the top  of the system  to  the vacuum pump,
  and the reagent is maintained within the optical cell by means of the capillary
  tip at the  lower end of the cell. The cycling of the reagent with the  stream of
  continuously sampled air continues until:

     (1) A pre-selected concentration of the pollutant is accumulated, or
     (2) a maximum time over which the reagent remains stable is exceeded,  or
     (3) the reset button is manually operated.

     Further investigation  should be conducted to establish the practicality of
  the system  under field conditions for  the measurement  of SO2. Instrument
  performance  data have been  supplied  by the manufacturer for the measure-
  ment  of  atmospheric  fluorides but no  such  specifications have been made
  available as yet for SO2 analysis.

  ANALYZER"C"

     The Portable SO 2 Meter was developed in the Central  Electricity Research
  Laboratory by Cummings and Redfearn. 77  SO 2 in air reacts with a starch-
  iodine reagent in  a countercurrent absorption column.  The amounts of  light
  absorbed  by the unchanged and the partially  decolorized reagent are com-
  pared by  photoelectric cells connected to a galvanometer. It was found that
  the reagent as recommended by  Katz  74  was stable  for several days when
  kept in stoppered  polyethylene bottles.  The light  absorption  of  batches of

-------
40                                            METHODS OF MEASURING
reagent made  over a period of  1  year varied by less than 1 percent. With
the countercurrent  absorption column and the rates of air and reagent flow
used  for this instrument,  no iodine was volatilized from  the solution by  the
incoming air.

   The  starch-iodine  reagent contained in  a 1-liter aspirator bottle,  which is
fitted  with  a constant-head device, flows by gravity through a control valve
into the bottom of  a liquid cell. From here, the reagent passes through a
rotameter,  which indicates the  flow rate to the top of the  absorption column
where it flows down the spiral path.  To avoid  corrosion by the starch-iodine
reagent, a korannite float and tantalum spring float stops are used.

   Air is drawn by means of a DC blower into the bottom of  the absorption
column where  the  SO2  reacts with the reagent. If a check on the zero is re-
quired,  air is first drawn through soda-lime by means  of stopcocks.  Air leaves
the top of  the absorber  via a rotameter and blower.  The partially decolorized
reagent flows from the  bottom of the absorber into the bottom of the glass
liquid cell and then  runs to a waste bottle.

   Comparison of the absorbance of the reacted to unreacted reagent is made
in a  photocell compartment  consisting of  a single light source (6 volts,  0.2
amp), condensers to  form parallel beams of light, red filters,  and two photo-
cells mounted behind the glass cells.  The unbalanced output of the photocells
is measured by  a  galvanometer.  The meter is sensitive to 0.01 ppm in  the
range 0 to  0.50 ppm.  This range may be extended by reducing the flow of
air to the instrument; the sensitivity decreases in proportion to  the increase in
range. The readings  obtained must be regarded as 2-minute running averages
because there  is  a time  lag of  about 2 minutes before a steady reading is
obtained for a particular 862 concentration.

   If  the instrument is operated in a motor  vehicle, power can be drawn from
the 12-volt battery.  For use away from roads a 12-volt battery can be in-
corporated within the instrument case.

   This instrument was  designed  for rapid measurement at various points in
a selected  area;  thus, there is  always an  operator with the instrument, and
a recorder  is  unnecessary.  Since the total output of the photocell is  about
0.25  millivolt,  it would not be practical to feed the output  to  a normal recorder.

ANALYZER"D"

   Helwig and Gordon 78 converted an automatic, conductimetric, SO 2  analyzer
to an automatic, continuous-recording, colorimetric analyzer by replacing the
conductivity cells with colorimeter cells having a 2-centimeter path and a mer-
cury  lamp  light  source.  A  synchroverter  was added  and necessary circuit
changes were made  to  use  a 0-  to  10-millivolt recorder. This instrument is
not commercially available.

-------
AND MONITORING SO 2                                                 41
     The pararosaniline-formaldehyde reagent developed by West and Gaeke 1
   was selected as the chromophoric  reagent for use in the automatic recorder.
   It was found that the rate and magnitude of  the color response to  SO2 were
   greater in the absence  of sodium  tetrachloromercurate.  Further increase in
   sensitivity  was obtained by decreasing the dye  concentration to one half of
   that suggested by the  authors. The mixed reagent was stable for 2 weeks.

     Collection efficiency  of the  countercurrent  absorption  column varied with
   airflow rate. At an airflow rate of  0.25  liter  per minute, the collection effic-
   iency  was  found to be 95 to 97 percent. Efficiency dropped to  73 to 75 per-
   cent at a flow rate of  1 liter  per minute. A system was established to detect
   S02 in the concentration range 0 to 5 ppm. By varying reagent and sample
   flow, the scale can be reduced or expanded.

     Static calibration of the instrument was conducted with standard  solutions
   of sodium  meta bisulfite equivalent to 1 to  5 ppm of SO 2 on the basis of an
   airflow of  0.25  liter per minute and a  reagent flow of 3.3 milliliters  per
   minute.

     An  instrument  sensitivity of 0.02 ppm  and excellent reproducibility were
   reported in the l-to-5-ppm range of concentration. NO2 and Oa in concen-
   trations  of  less  than   1  ppm  produced negligible  interference  on  SOa
   measurement.
   Air  lonization  Method
     The aerosol ionization detector consists essentially of an ionization chamber
  with a stainless steel cylinder serving as an outer electrode  and a heavy wire
  as the inner electrode. Fifty to 100 micrograms of radium is placed along the
  surface of the cylinder  to  serve  as an alpha source. The  two electrodes are
  under a voltage  potential difference that causes a flow of ion current between
  them.  The  presence of  small concentrations  of particulate matter causes  a
  marked drop in the flow of ion current.

     A compensated detection system is used in the analyzer. The sample stream
  is  divided  between  two  similar ion chambers.  The difference in ion current
  produced by the unreacted and reacted sample stream is detected by an elec-
  trometer measuring  circuit and is fed into a suitable recorder.

-------
 42                                             METHODS OF MEASURING
   To use this principle for the determination of gases and vapors, it is neces-
sary to form small particles from the material being analyzed. Three methods
may be used to  accomplish this: (1)  Reaction of the material with a reagent
to produce particulate matter, (2) pyrolysis  or irradiation with ultraviolet
light, or (3) a combination of both.

   The  instrument is sensitized to SO 2 by pyrolysis in the presence of CuO.
A gas-solid reaction occurs and copper sulfate is formed.

   An instrument range of 0 to lOppm SO2  is obtained with this sensitization
method. The method is nonspecific since any material cap able of salt formation
with CuO at elevated temperatures will interfere. Additional investigation is
required concerning instrument performance  under conditions of continuous
air monitoring, reproducibility, sensitivity, and the relative effects of interfering
materials.

-------
AND MONITORING SO 9                                               43
                           REFERENCES
   1.  West, P.  W., Gaeke, G.  C., Fixation of Sulfur Dioxide as Disulfitomer-
      curate  (II),  Subsequent Colorimetric  Estimation, Anal. Chem., 28:1916.
      1956.

   2.  Nauman,  R. V., West,  F. W.,  Tron, F., Gaeke, G.  C., Spectrophoto-
      metric  Study  of the Schiff Reaction  as  Applied to the  Quantitative
      Determination of Sulfur Dioxide, Anal. Chem., 32:1307.  1960.

   3.  McCaldin, R. O., Hendrickson, E. R.,  Use of a Gas Chamber for Testing
      Air Samplers, J. Am. Ind. Hyg. Assoc., 20:509. 1959.

   4.  Terraglio, F.  P., Manganelli, R.  M., Laboratory Evaluation of 862
      Methods, Anal. Chem., 34:675.  1962.

   5.  Perry,  W. H.,  Tabor, E. C., National Air Sampling Network Measure-
      ment of SO2  and NO2, Arch. Environmental Health, 4:44.  1962.

   6.  West, P. W., Ordoveza, F.,  Elimination of Nitrogen Dioxide Interference
      in the Determination of  Sulfur Dioxide,  Anal.  Chem., 34:1324. 1962.

   7.  Zurlo, N., Griffini,  A.  M., Measurement  of  Sulfur Dioxide Content of
      the Air in the Presence of Nitrogen and Heavy Metals, Med. d. Lavoro,
      5:330. 1962.

   8.  Craxford, S. R., Slimming, D.  W., Wilkins, E. T., The Measurement of
      Atmospheric Pollution: The  Accuracy  of the Instruments and the Signifi-
      cance of the Results, Proceeding of the Harrogate Conference.  London,
      England. 1960.

   9.  Pate, J. B.,  Lodge, J.  P.,  Wartburg, A.  F., Effect of Pararosaniline in
      the Trace Determination of Sulfur  Dioxide, Anal. Chem., 34:1660. 1962.

  10.  Jacobs, M.  B.,  The  Chemical Analysis of Air Pollutants,  Interscience,
      New York. 1960.

  11.  Jacobs, M. B., Greenburg, L., Sulphur Dioxide in New York  City Atmos-
      phere, Ind. Eng. Chem., 48:1517. 1956.

  12.  Perley, G. A.,  Langsdorf,  B.  T.,  Problems in the Recording of Sulfur
      Dioxide in Polluted Atmospheres,  Chap.  69, Proceedings of the United
      States Technical Conference in Air Pollution, L. C. McCabe, ed., McGraw-
      Hill,  New York. 1952.

  13.  Yocum, J.  E.,  Richardson, R. L.,  Saslaw, X. M.,  Chapman, S., The
      Effect of Hydrogen Chloride on the Operation of the Thomas Autometer,
      Proceedings 49th Annual Meeting of the Air Pollution Control  Association,
      Buffalo, N. Y. 1956.

-------
44                                            METHODS OF MEASURING
 14.   Jacobs,  M. B., Braverman, M. M., Hochheiser, S., Ultrasensitive Con-
      ductometric Measurement of Sulfur Dioxide in Air,  Instrument  Society
      of America Meeting, Paper No. 56-32-3, Sept. 1956.

 15.   Reece, G.  M.,  White,  B., Drinker, P.,  Determination  and Recording of
      Carbon  Disulfide  and Hydrogen Sulfide in the Viscose-Rayon  Industry,
      J. Ind. Hyg. Tox., 22:416. 1940.

 16.   Thomas, M. D., Ivie, J. O., Abersold, N. N., Hendricks, R. H., Automatic
      Apparatus for  Determination of Small Concentrations of Sulphur Dioxide
      in Air Application to  Hydrogen Sulfide, Mercaptans, and Other  Sulfur
      and Chlorine Compounds,  Ind.  Eng.  Chem. Anal. Ed., 15:287.  1943.

 17.   Cummings, W. G.,  Redfearn, M. W.,  Instruments for Measuring Small
      Quantities  of Sulfur  Dioxide in  the Atmosphere, J. Inst. of  Fuel, 30:628.
      1957.

 18.   Air  Sampling  Instruments, American  Conference of Governmental In-
      dustrial Hygienists, Cincinnati, Ohio. 1962.

 19.   Recommended  Methods in Air  Pollution Measurements, California State
      Department of Public Health, Berkeley, Calif. 1960.

 20.   ASTM  Standards on  Methods  of Atmospheric Sampling and Analysis,
      American Society for Testing Materials, Philadelphia, Pa.  2nd Ed.,  1962.

 21.   Hochheiser, S., Braverman, M.  M.,  Jacobs,  M.  B.,  Comparison of
      Methods for the Determination of Atmospheric Sulphur Dioxide, presented
      at  Metropolitan L.  I.  Subsection,  New York  Section, ACS Meeting in
      miniature,  Feb. 1955.

 22.   Paper  EPA/AR/4283,  Routine  Methods for Estimating Sulphur Dioxide
      in the Air, Organization for European Economic  Cooperation, European
      Productivity Agency, Paris, France. 1961.

 23.   Welch,  A.  F.,  Terry, J. P., Development in the Measurement of Atmos-
      pheric Sulfur Dioxide, J. Am. Ind. Hyg. Assoc., 21:316.  1960.

 24.   Paulus,  H. J., Floyd, E.  P.,  Byers,  P.  H., Determination of Sulphur
      Dioxide  in Atmospheric Samples. Comparison of a Colorimetric  and a
      Polarographic  Method, J. Am.  Ind.  Hyg. Assoc.  Quarterly, 15:4.  1954.

 25.   Urone,  P.  F.,  Boggs, W.  E., Acid-Bleached Fuchsin in Determination of
      Sulfur Dioxide in the Atmosphere, Anal. Chem., 23:1517. 1951.

 26.   Steigmann, A., A New Color Reaction for Sulfurous Acid, The Thiol
      Group and Formaldehyde, Anal. Chem., 22:493. 1950.

 27.   Grant, W.  N.,  Colorimetric Determination of Sulphur Dioxide, Ind. Eng.
      Chem. Anal. Ed., 19:245.  1947.

-------
AND MONITORING SO 2                                                 45
  28.  Stang, A.  M., Zatek, J. E., Robson, C. D., A Colorimetric Method for
      the Determination  of Sulfur Dioxide  in Air,  Am. Ind.  Hyg.  Assoc.
      Quarterly,  12:5. 1951.

  29.  Moore, G.  E.,  Cole, A. F. W., Katz, M., The Concurrent Determination
      of  Sulphur Dioxide  and  Nitrogen  Dioxide  in the Atmosphere, J.  Air
      Poll. Control Assoc., 7:25. 1957.

  30.  Determination of Sulphur Dioxide in Air, Fuchsin-Formaldehyde Method,
      American  Conference of Governmental Industrial  Hygienists,  Methods
      Manual, Cincinnati, Ohio. 1958.

  31.  Stratman,  H., Micro analytical Methods for the  Determination of Sulphur
      Dioxide, Mikrochimica Acta, 6:668. 1954.

  32.  Kanno, S., The Colorimetric Determination of Sulfur  Oxides in the Atmos-
      phere, Intern. J. Air Poll., 1:231. 1959.

  33.  Bertolocini, R. J.,  Barney, J. E.,  Colorimetric Determination of  Sulfate
      Using Barium Chloranilate, Anal. Chem., 29:281. 1957.

  34.  Hands,  G.  C., Bartlett,  A.  F., A Field Method for  the Determination of
      Sulphur Dioxide in Air, Analyst, 85:147. 1960.

  35.  Dept. of Sci. and Ind. Research. DSIR Methods for the Detection of Toxic
      Gases   in   Industry,  Leaflet No.  3, H. M. Stationery Office,  London,
      England. 1939.

  36.  Liddell,  E.  F., A Reagent  for Sulphur Dioxide, Analyst, 80:901. 1955.

  37.  Klein, B.,  Microdetermination of Sulfate. A Colorimetric Estimation of
      the Benzidine Sulfate Precipitate, Ind.  Eng. Chem.  Anal.  Ed., 16:536.
      1944.

  38.  Lambert, J. L., Yaida, S. K., Grother, M. P., Colorimetric Determination
      of Sulfate Ion, Anal. Chem.,  27:800. 1955.

  39.  Blinn, R. C.,  Gunther,  F.  H.,  Indirect Ultra-Violet Spectrophotometric
      Determination of Sulphur Dioxide by Means of Plumbous Ion, Analyst,
      86:675. 1961.

  40.  Mathers, A. P.,  Rapid Determination  of Sulphur  Dioxide in Wines, J.
      Assoc. of Agric. Chem., 32:745. 1941.

  41.  Stephen, B. G., Lindstrom, F., Spectrophotometric Determination of Sulfur
      Dioxide Suitable  for Atmospheric Analysis, Anal. Chem.,  36:1308. 1964.

  42.  Kolthoff, I. M.,  Miller,  C. S.,  The  Reduction  of Sulfurous Acid at the
      Dropping Mercury Electrode, J. Am. Chem. Spc., 63:2818. 1941.

  43.  Smith, R.  B., Fries,  B.  S. T.,  Portable Motor-driven Impinger Unit for
      Determination of Sulfur Dioxide, J. Ind. Hyg., 13:338. 1931.

-------
46                                             METHODS OF MEASURING
 44.  Pearce,  S.  J., Schrenk,  H.  H., Determination of Sulfur Dioxide in Air
      by  Means of the Midget  Impinger, Bureau of Mines  RI 4282. 1948.

 45.  Kitano, Y.,  Takakuwa, H.,  Determination of  Hydrogen Sulfide and
      Sulfur  Dioxide  in Air, Errors in lodometry, Japan Analyst, 3:7. 1954.

 46.  Jacobs, M. B.,  The Analytical  Chemistry of Industrial Poisons, Hazards
      and Solvents, Interscience, New York. 1959.

 47.  Elkins,  H.  B.,  The  Chemistry of  Industrial  Toxicology, Wiley, New
      York. 1959.

 48.  Griffin,  S. W., Skinner, W. W., Small Amounts of Sulfur Dioxide in the
      Atmosphere. Improved Methods for the Determination of Sulphur Dioxide
      when Present in  Low  Concentrations in Air, Ind. Eng.  Chem., 24:862.
      1932.

 49.  Corbett, P. F., A Phototurbidimetric  Method  for the Estimation of Sulfur
      Trioxide in the Presence of Sulfur Dioxide, J. Soc. of Chem. Ind., 67:227.
      1948.

 50.  Matty,  R.  E.,  Diehl,  E. K., Measuring  Flue  Gas Sulfur Dioxide and
      Sulfur Trioxide, Power, 101:94.  1957.

 51.  Treon, J.  F.,  Crutchfield, W.  E., Rapid Turbidimetric  Method for the
      Determination of Sulfates,  Ind. Eng. Chem. Anal.  Ed., 14:119. 1942.

 52.  Volmer, W., Frohlich,  F. Z., Turbidimetric Determination of  Sulfate,
      Anal. Chem., 126:414, 1944.

 53.  Department  of Scientific  and Industrial  Research,  The  Investigation of
      Atmospheric Pollution, 18th Report,  1931-1932, H. M. Stationery Office,
      London, England.

 54.  Ibid, 20th Report, 1933-1934.

 55.  Wilsdon,  B.  H., McConnell, F.  J.,  The  Measurement  of Atmospheric
      Sulfur Pollution by Means of Lead Peroxide, J. Soc. Chem., Ind., 53:385.
      1934.

 56.  Parker,  A.,  Richards,  S. H., Instruments  Used for the Measurement of
      Atmospheric  Pollution in Great  Britain,  Chap. 67;  Proceeding of the
      United States Technical Conference in Air Pollution, L. C. McCabe, ed.,
      McGraw-Hill. 1952.

 57.  Thomas, F. W.,  Davidson, C.  M., Monitoring Sulfur Dioxide with Lead
      Peroxide Cylinders, J. Air Poll. Control Assoc., 11:24.  1961.

 58.  Foran,  M.  R.,  Gibbons, E.  V., Wellington,  J. R.,  The  Measurement of
      Atmospheric Sulfur Dioxide  and Chlorides, Chemistry in Canada  10'33.
      1958.

-------
AND MONITORING SO 2                                                 47
 59.  Wilkins, E.  T.,  Air Pollution  in  London Smog, Mech.  Eng.,  76:426.
     1954.

 60.  Buck, V. M., Methods for the Determination of Hydrogen Fluoride and
     Sulfur Dioxide in the Atmosphere, Staub, 21:227. 1961.

 61.  Pate,  J.  B., et al., The Use of  Impregnated Filters to Collect Traces of
     Gases in the Atmosphere, Anal. Chem. Acta., 28:341. 1963.

 62.  Hugen, C., The Sampling of Sulfur Dioxide in Air with Impregnated
     Filters, Anal. Chem. Acta,  28:349. 1963.

 63.  Gisclard, J., The  Use of Detectors and Test Kits in Industrial Hygiene
     investigations, AMA Arch. Ind. Health, 21:250. 1960.

 64.  .Kusnetz, H.  L., Saltzman, B. E., Lanier,  M. E.,  Calibration and  Eval-
     uation of Gas Detecting Tubes, Amer.  Ind. Hyg. Assoc. J., 21:361.  1960.

 65.  Dickinson,  J.  E.,  The Operation and the  Use of  Titrilog and the  Auto-
     meter, Proceedings 49th Annual Meeting of Air Poll. Control Assoc.  1956.

 66.  Giever, P. M., Cook, W. A., Automatic Recording Instruments As Applied
     to Air  Analysis,  Arch,  of Ind. Health and Occ. Med., 21:233.  1960.

 67.  Washburn, H. W., Austin, R. R., The Continuous Measurement of Sulfur
     Dioxide  and Hydrogen  Sulfide Concentrations by Automatic Titration,
     Chap.  72,  Proceedings of the United  States Technical Conference in Air
     Pollution, L. C. McCabe, ed., McGraw-Hill.  1952.

 68.  Field,  E.,  Oldach,  C.  S., Conversion of Organic  Sulfur to Hydrogen
     Sulfide for Analysis, Ind. Eng. Chem. Anal. Ed., 18:668. 1946.

 69.  Thomas, M. D., Ivie, J. O., Abersold, N. N., Hendricks, R. H., Automatic
     Apparatus  for Determination of  Small Concentrations of Sulfur Dioxide
      in Air.  Application  to Hydrogen Sulfide,  Mercajptans and Other Sulfur
     and  Chlorine  Compounds,  Ind.  Eng. Chem. Anal.  Ed., 15:287.  1943.

 70.  Nader,  J.  S.,  Dolphin, J.  L., Improved Titrilog Sensitivity, J. Air Poll.
     Control Assoc., 8:336. 1959.

 71.  McKee,  H. C.,  Rollwitz,  W.  L.,  Improved  Titrilog Sensitivity, Field
     Performance and  Evaluation, J. Air Poll. Control Assoc., 8:338.  1959.

 72.  Carpenter,  S.  B.,  Sparkman,   R.  E.,  Adaptation of  Titrilog for Field
     Mobile Sampling, J. Air Poll. Control Assoc., 5:195. 1956.

 73.  Gartrell, F. E., Carpenter,  S. B., Aerial Sampling by Helicopter, A  Mejhod
     for the Study of Diffusion Patterns, J. Amer. Meteorol. Soc., 12:215.  1955.

-------
48
  74.  Katz, M., The Photoelectric Determination of Atmospheric Sulphur Dioxide
      by Dilute Starch-Iodine Solutions, Chap.  71, Proceedings of the United
      States Technical Conference in Air Pollution, L. C. McCabe, Ed., McGraw-
      Hill, New York. 1952.

  75.  Katz,  M., Photoelectric Determination  of Atmospheric  Sulfur Dioxide
      Employing Dilute Starch-Iodine  Solution,  Anal. Chem., 22:1040. 1950.

  76.  Adams,  D.   F.,  Dana,  H. J., Koppe, E.K.,  Universal Air Pollutant
      Analyzer, USPHS Contract No. 66512. Robert A. Taft Sanitary Engineer-
      ing Center, Cincinnati, Ohio. 1957.

  77.  Cummings, W.  G., Redfearn, M.  W., Instruments for Measuring Small
      Quantities of  Sulfur Dioxide in  the Atmosphere,  J. Inst. of Fuel, 30:628.
      1957.

  78.  Helwig,  H.   H.,  Gordon, C. L., Colorimetric  Methods for  Continuous
      Recording Analysis of Atmospheric Sulphur Dioxide, Anal. Chem., 34:1660.
      1962.

  79.  Strange,  J. P.,  Ball, K. E.,  Barnes, D.  O., Continuous Parts Per Billion
      Recorder  for Air Contaminants,  Proceedings  of the Air Poll. Control
      Assoc., Cincinnati, Ohio. May 22-26. 1960.
                                                               GPO  937-643

-------
BIBLIOGRAPHIC:  Hochheiser, Seymour. Methods of mea-
   suring and monitoring atmospheric sulfur dioxide.'PHS
   Publ. No. 999-AP-6. 1964.48 pp.

ABSTRACT:  A literature review of methodology relating to
   the measurement of atmospheric sulfur dioxide, a detailed
   description  of recommended  methods, and criteria  for
   selection of recommended methods are presented in this
   report. This publication is intended to serve as a resource
   document for those involved in measurement of pollution
   and in research on new or unproved methods, and for those
   who seek to bring about widespread agreement in. matters
   concerning measurement of pollution.
 BIBLIOGRAPHIC:  Hochheiser,  Seymour. Methods of mea-
    suring and monitoring  atmospheric  sulfur dioxide. PHS
    Publ. No. 999-AP-6. 1964. 48 pp.

 ABSTRACT:  A literature review of methodology relating to
    the measurement of atmospheric sulfur dioxide, a detailed
    description of recommended  methods, and criteria for
    selection of recommended  methods are presented in this
    report. This publication  is  intended to serve as a resource
    document for  those involved in  measurement of pollution
    and in research on new or improved methods, and for those
    who seek to bring about widespread agreement in matters
    concerning measurement of pollution.
ACCESSION NO.
KEY WORDS:
   Sulfur Dioxide

   Measurement and
   Monitoring

   Recommended Methods

   Criteria for
   Selection

   Literature Review

   Ajf Pollution

   Manual and Automatic
   Methods
 ACCESSION NO.

 KEY WORDS:

    Sulfur Dioxide

    Measurement and
    Monitoring

    Recommended Methods.

    Criteria for
    Selection

    Literature Review

    Air Pollution

    Manual and Automatic
    Methods

-------