EPA-650/4-74-036
     AUGUST 1974
Environmental Monitoring Series

SSi'S

-------
                                              EPA-650/4-74-036
N02  ACTINOMETER  FOR  FIELD  USE
                      Prepared by

            D.E. Burch, R.C. Bean, andF.J. Gates

            Philco-Ford Co. , Aeronutronic Division,
                       Ford Road
               Newport Beach, California 92663
                  Contract No. 68-02-0798
                     ROAP No. 56AAI
                Program Element No. 1A1003
               EPA Project Officer: P.L. Hanst

              Chemistry and Physics Laboratory
            National Environmental Research Center
         Research Triangle Park, North Carolina 27711
                      Prepared for

           OFFICE OF RESEARCH AND DEVELOPMENT
          U.S. ENVIRONMENTAL PROTECTION AGENCY
                 WASHINGTON, D.C. 20460

                      August 1974

-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication.  Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
                                   11

-------
                    ABSTRACT
     Solar radiant energy in the ultraviolet and
short-wave length visible dissociate N02 in the
atmosphere to produce NO and 02•  This photolytic
reaction plays an important role in the formation
of photochemical smog, and information about the
amount of actinic energy available in the lower
atmosphere is required for the development of
mathematical models of the atmospheric processes.
This report describes the development and testing
of an actinometer designed to measure the actinic
energy available for the photolytic dissociation
of N02-  A spherical test bulb contains a mixture
of NO? and 0~ when it has been in the dark for
several minutes.  When the bulb is exposed to
solar energy NO is formed; its concentration is
monitored by gas-cell correlation methods involving
the infrared absorption by NO.  A shutter periodically
shades the test bulb from the sun for a one-minute
period each two minutes.  During the shaded period,
part of the NO recombines with 02 to form N02-  The
cyclic change in the NO concentration is related
to the actinic irradiance.
                      iii

-------
                              CONTENTS







                                                                 Page





Abstract                                                          iii




List of Figures                                                    v





Sections




1       Introduction                                              1-1




2       Preliminary Tests                                         2-1




3       Description of Actinometer                                3-1




A       Tests Results and Instrument Performance                  4-1




5       Electrical Diagrams and Details of Components              5-1




6       References                                                6-1
                                  IV

-------
                               FIGURES

Wo.                                                              Page

 1-1     Spectral plots of the absorption coefficiento            1-5
         for N02 and N20^ between 3000 cm"1 and 4500 A

 2-1     Optical diagram of actinometer used for pre-             2-2
         liminary tests

 2-2     Response of breadboard model NC>2 actinometer to          2-5
         sunlight

 2-3     Response of N02 actinometer to artificial uv and         2-6
         visible light at different intensities and
         different position in the test bulb

 2-4     Plots of V" and AV" vs relative actinic irradiance       2-8

 2-5     Plots of V" and AV" vs 02 partial pressure in the
         test bulb for various partial pressures of N02
                                                o
 2-6     Transmittance of NO, from 4300 to 4500 A                 2-12

 3-1     Optical diagrams of the optics assembly of the           3-2
         actinometer

 3-2     Two views of the optics assembly of the actinometer      3-3

 3-3     Photograph of the control assembly, amplifier and        3-4
         recorder

 4-1     Response of actinometer to solar energy with the         4-2
         shutter in the automatic mode

 4-2     Recorder tracings of actinometer output when the         4-3
         sky contains broken clouds

-------
                           FIGURES  (contd.)


Mo.                                                             Page

4-3       Plots  of AV"  vs  relative actinic  irradiance             4-5

4-4       Response of actinometer to different absorber           4-8
          thicknesses of NO in  a mixture  of NO + ^ at
          a  total pressure of  120 torr

5-1       Simplified schematic  diagram  for  ac power               5-4

5-2       Simplified diagram of interconnecting cables            5-5

5-3       Wiring diagram of electrical  components on              5-5
          the  optics assembly

5-4       Wiring diagram of timer and control assembly            5-6

5-5       Schematic diagram of  preamplifier used with             5-7
          the  InSb detector

5-6       Schematic diagram of  the 350  Hz reference-              5-7
          pickup

5-7       Power  supply  for thermo-electric  cooler for             5-8
          the  detector

5-8       Power  supply  for preamp and reference circuits          5-8
                                     vi

-------
                                SECTION 1
                               INTRODUCTION
BACKGROUND

     Nitrogen dioxide is readily photolyzed to NO and 0- by short-wave
visible and near-ultraviolet light.  Measurement of the rate of photolysis
of N©2 in polluted air is necessary in order to understand the complex
photochemical processes in the atmosphere.  Photolysis of N0~ starts a
series of chemical reactions that yield ozone, peroxyacyl nitrates, and
other objectionable species that make up photochemical smog.  There is
presently a need for a field instrument that can measure the amount of
light available to photolyze N02 for purposes of developing mathematical
models of the atmospheric processes.

     Laboratory methods for measuring the intensity of light producing
photolysis usually involve a photochemical reaction chamber containing
N02 and an inert gas.  The photolysis is measured by monitoring the dis-
appearance of N02«  Such a method is impractical for a field instrument
because the gas must be replaced after each measurement.  Photoelectric
instruments have generally been used in the field because they are simple
and can be used continuously.  However, this type of instrument suffers
two serious drawbacks that make it difficult to relate a reading to the
rate of N02 photolysis.

     Geometric corrections and spectral response corrections must be
applied to the reading of the photoelectric sensor.  The response of a
photodetector is highly directional, with maximum response to light rays
perpendicular to the surface.  It is clear that correcting for the direc-
tional response would be quite complicated and would require considerable
knowledge of the angular variation of the solar irradiance, which depends
on the solar zenith angle and on the conditions of clouds and haze.
                                     1-1

-------
     Spectral corrections are required because it is not possible to pro-
duce a photodetector  surface with a spectral response that matches the
absorption spectrum of N02 •  The rate of photolysis due to light in a
narrow spectral interval is proportional to the product of the solar spec-
tral irradiance and the absorptivity of the NO,*  Similarly,  the response
of a photodetector due to the same narrow spectral interval is proportional
to the product of the solar irradiance and the spectral responsivity of
the detector.  Because the relative spectral responsivity of  a photodetector
may change and the spectrum of sunlight reaching the lower atmosphere is
variable, the spectral correction is also very difficult.
SUGGESTED INSTRUMENT  FEATURES

     An ideal N0« actinometer would have exactly the same spectral re-
sponse as the N02 absorption in  the- atmosphere and would be equally sensi-
tive to light arriving  from any  direction.  Hanst* has suggested that this
ideal actinometer can be approximated by an instrument built along the
following lines.  A spherical receiver, which we shall call a test bulb,
is made of quartz, which is transparent to all wavelengths of actinic
radiant energy reaching the lower atmosphere.  A photosensitive mixture
of N02 + Oo» and possibly an inert gas such as ^ > is contained in the
test bulb.  The system  operation depends on the chemistry of N02 photolysis
and the reverse reaction given by


          N02 + hv  *  NO + (1/2) 02.                              (1-1)


The bulb is mounted in  the open, well above any parts of the instrument,
so it can receive radiant energy from all directions except for a few
steradians below.  Most types of earth surface reflect weakly in the uv
and visible; therefore, it is unlikely that a significant fraction of the
actinic energy reaching a point  in the lower atmosphere comes from the
lower hemisphere.  Thus, the gas in the bulb will effectively integrate
the effects of the radiant energy arriving from all directions.

     The problem of spectral response errors inherent in photoelectric
detectors is greatly  reduced by  having the N02 itself be the absorber of
actinic light in the  actinometer.  Since the absorptivity of NC>2 is not
the sar.ie for all wavelengths of  actinic light, the correct spectral response
can be maintained only  if the NOo concentration in the test bulb is low
enough that the gas is  optically thin at all wavelengths.

     A new feature of the instrument suggested by Hanst involves the addi-
tion of 02 to the test  bulb along with the NO,•  In the presence of the
excess 02 the reverse reaction takes place to form N02 when the test bulb
                                     1-2

-------
 is shaded from actinic  light.   The  second  new  feature  of  the suggested
 instrument is  a shutter  that  periodically  opens  to  expose the  test bulb
 to actinic light then closes  to allow  the  reverse reaction to  occur.  The
 increase in the concentration of the NO, or  the  decrease  in N02, during
 the open portion of each cycle of the  shutter  is monitored spectroscopi-
 cally.   The monitor employs a gas correlation  cell  that is used as a
 selective filter.   The correlation  cell contains either NO or  N0~, which-
 ever is being  monitored  in the test bulb.

      We have previously  employed gas correlation methods  to monitor NO
 at concentrations  of interest for the  present  purposes.   The fundamental
 vibration-rotation absorption band  near 5.4 |j,m is employed.  The fractional
 change  in the  NO concentration in the  test bulb  is  expected to be quite
 large during each  period of the cyclic shutter.  On the other  hand, only
 a small fraction of the  N02 is expected to dissociate  during each cycle.
 Therefore,  it  seemed likely that the NO concentration  could be measured
 more accurately than the N02  concentration.  Monitoring N02 has one ad-
 vantage in  that visible  absorption  bands could be employed  to  take
 advantage of simple and  efficient sources and  detectors.


 SUMMARY OF  WORK PERFORMED

      The objectives  of the work reported here have  been to  demonstrate
 the feasibility of  an N02 actinometer based on Hanst's recommendations,
 to perform  experiments to determine the required design parameters, and
 to build and demonstrate  an instrument for field use and  evaluation.  The
 preliminary tests  performed before  the final instrument was designed are
 discussed in Section 2.   Most  of  the tests were  performed  indoors under
 controlled  conditions with a uv lamp serving as  the source  of actinic
 energy.   Gas cell correlation methods were used  to monitor  the NO con-
 centration  in  a  cylindrical test  bulb.  A cylindrical bulb, rather than
 a  spherical one, was used because it facilitated the investigation of
 possible wall  effects and geometrical effects in the test bulb.  Besides
 demonstrating  the feasibility  of  the instrument, the preliminary tests
 provided  information that was quite valuable in  the design of the acti-
 nometer  to  be built  and demonstrated as part of  the present contract.

     Limited experimental data  obtained on the measurement of N02 by gas
 correlation methods  tended to confirm our original belief  that NO could
 be  better monitored  in an actinometer test bulb than could NO,.  A few
 data were obtained by using hot NO as a source of radiant  energy when
monitoring  NO  in a separate sample.   As expected, a given small sample of
NO  produces a  larger fractional change in the detector signal when hot
NO  is used as an energy source  in the place of a continuum source,   Better
discrimination against interfering HgO absorption can also be achieved.
However, simple NO cells  cannot be operated at temperatures as  high as
conventional continuum sources; therefore,  their  spectral  radiance  is  less,
                                    1-3

-------
even at wavelengths  of  high NO emissivity.  We have chosen a hot ceramic
source for  the  actinometer because  of  its simplicity and the improved
signal-to-noise ratio obtainable  with  it.

     Section  3  describes  the actinometer built for laboratory and field
use.  It  has  been  designed so that  components can be interchanged and gas
mixtures  in the correlation cell  and test bulb can be modified.  Only a
limited amount  of  test  data represented by  those summarized in Section 4
have been obtained.  Although the practicality of the instrument for field
use has been  demonstrated and the tests indicate that it performs ade-
quately,  subsequent  tests to be performed by EPA scientists will probably
lead to suggestions  for minor design changes.


SPECTRAL  CHARACTERISTICS  OF N02 ABSORPTION

     When actinic  radiant energy  strikes a  layer of NOo, the absorption is
a function  of the  wavelength with strong absorption throughout much of
the visible and near uv.   ^0^, a compound  that is always associated with
N02,  is  transparent  in  the  visible but  absorbs in the uv.  Fig. 1-1 shows
plots of the absorption coefficients  as given by Leighton2 for both species
over  this  portion  of the spectrum.  The spectral curves shown in Fig. 1-1
were  not obtained  with  sufficiently good resolution to show all of the
vibrational structure that  exists at  the longer wavelengths.  Below about
3700A the  bands  are  diffuse with essentially no unresolved structure.

      If  I  ^ photons of actinic energy  of • wavelength X in a parallel beam
are incident on  each cm  per second,  the rate of absorption of the photons
is given by

                        -3-1      P
           N (photons cm  sec  )  =    or c I  , dX,                 (1-2)
           a                        J       o , A.


where a, the absorption coefficient,  is expressed in cm2 /molecule and is a
function of wavelength,  and C is the  concentration of the absorbing gas
species  in molecules /cm3.   Equation (1-2) is restricted to optically thin
samples  so that  the  incident intensity  is essentially the same at all
points in  the absorbing layer.  If the  sample is too thick, the intensity
is less  near the surface where the beam leaves the absorbing layer than it
is on the  front  surface where the beam  is incident.  Additional data on
N02 absorption in  the visible appear  in Section 2.

     The absorption  coefficient shown as the right-hand  ordinate scale
of Fig.  1-1 is expressed in atm'^m"1 and is convenient when relating the
transm'.ttance to the amount of N02 in the optical path.  If p is the par-
tial pressure of the absorbing gas in atmospheres, 8 is the temperature in
degrees  Kelvin,  and  L is the geometrical path length through the gas,
                                     1-4

-------
               =   exp
-[•
X (atm^cm-^p
                                          p(atm) L(cm) 273/9~|.     (1-3)
     The  ratio  of  the  concentration of N?0.  to that of NCL decreases with
 increasing  temperature and increases in proportion to the NO? pressure.
 Because of  the  very  low  partial pressure of  NO? in the atmosphere, the
 fraction  that appears  as ^A is extr~ertely low artd»  therefore,  negligible,
 However,  the N^O,  absorption may be significant in laboratory samples or
 in  test cells containing higher concentrations of the gases.  The rela-
 tive equilibrium concentrations of NOj and N_0, can be determined from
 data given  by Giauque  and Kemp^.

     Most of the absorbed radiant energy of  wavelengths greater than
 approximately 4100A  produces changes in the  electronic and vibrational
 energies  of the absorbing molecules, but it  does not produce dissociation.
 However,  a  large fraction of the energy of shorter wavelengths produces
 photolysis  of the  NO,  molecules in accordance with Eq. (1-1).  The quantum
 yields of photolysis,  according to Leighton  at 4 representative wave-
 lengths are as  follows:  3130A, 0.97; 3660A, 0.92; 4047A, 0.36; and 4358A,
 0.00.  Because  of  the  low quantum yield of photolysis at the longer wave-
 lengths and the low  spectral radiance of the sun at short wavelengths,
 most of the photolysis of atmospheric N0« is due to radiant energy between
 approximately 3200A  and  4200A.  By utilizing the quantum efficiency for
 dissociation, the  spectral absorption characteristics of N02, and the
 actinic irradiance,  the  rate constant for dissociation reaction of N02 in
 the atmosphere  can be  calculated.  A value of 0.37 min"1 (0.0062 sec'l) is
 accepted  for typical clear atmospheric conditions^.
o
0)
i— i
o
e
*~s
o
    4  x 10
           -19
    2 x 10
          -19
             3000
                              3500
                                                4000
                                                                  4500
                                                                and N_0
FIG. 1-1   Spectral plots of the absorption coefficient for NO,        .
between 3000 cm'1 and 4500 X.     The ordinate scale shown on Ehe rigfit is
convenient when the partial pressures of the gases are measured in atmospheres
and the optical path lengths are in cm.
                                     1-5

-------
                                 SECTION 2
                             PRELIMINARY TESTS
INSTRUMENTATION AND PROCEDURES

     An optical diagram of the instrument used for the preliminary tests
is shown in Fig. 2-1.  The test bulb is filled with a mixture of NO- and
Op-  Actinic radiant energy in the near uv and visible from lamp A causes
a photochemical reaction that produces NO in the test bulb.  The actinic
energy can be blocked from the test bulb by shutter Sh.  The remainder
of the instrument is a gas-correlation cell analyzer that employs infra-
red energy to monitor the concentration of the NO in the test bulb.

     Images of the radiant energy source are formed near the center of
the test bulb and on the slit.  A narrow beam of nearly-collimated energy
is formed by the combination of the slit and the mask on the flat mirror
near the source.  After the beam passes through the slit, a rotating mirror
chopper alternately directs it through one of two paths to a bandpass
filter and detector.  The 0.15u-m wide bandpass of the filter centered near
5.3u-m includes a portion of the strong, fundamental band of NO.  Approxi-
mately 640 torr of NO in the 1 cm correlation cell is essentially opaque
near the centers of the NO absorption lines and absorbs approximately 72%
of ihe radiant energy in the bandpass of the filter.  With no NO in the
test bulb and with the blocking shutter B removed from the beam through
the attenuator, the attenuator is adjusted so that the average transmittances
of the two alternate beams are the same.  Thus, the component of the de-
tector signal at 170 Hz, the frequency of the rotating mirror chopper, is
made zero.

     When NO is formed in the test bulb, it absorbs a portion of the radi-
ant energy leaving the detector.  Because the NO in the test bulb absorbs
                                    2-1

-------
                                                                          Det-
                   M
       (r
10
i
IS)
                           Test Bulb
                                                                Reference
Synchronous
Amplifier
  and
Gain Control
                                                                              Strip Chart
                                                                               Recorder
      FIG.  2-1.   Optical diagram of actinometer used for preliminary tests.  S, infrared source.
                 M,  mask to limit IR beam size.  Ch, chopper.  CC, correlation cell containing NO.
                 F;  bandpass filter centered near 5.3 urn.  Det, liquid-nitrogen cooled InSb
                 detector.  Att, adjustable attenuator.  B, blocking shutter that can be moved
                 into or out of the beam.  A, source of uv and visible actinic energy.  Sh, shutter
                 to shade the test bulb from the actinic energy.  When the instrument was used to
                 investigate solar energy, the sky  replaced A.

-------
most at wavelengths where the NO in the correlation cell is  opaque,  the NO
in the test bulb has less effect on the beam passing through the correlation
cell than on the beam passing through the attenuator.   The two alternate
beams are no longer balanced and a detector signal at  170 Hz results.   The
percentage of misbalance between the two alternate beams is  a measure  of the
concentration of NO in the test bulb.  The spectroscopic principles  of
operation of gas cell correlation instruments have been described in detail
in References 4 and 5.

     In order to normalize the detector voltage, all of the  NO is removed
from the test bulb and the blocking shutter is moved so that it blocks
the beam through the attenuator.  The amplifier output for this condition
is defined as VD.  The blocking shutter is removed from the  beam and the
attenuator is adjusted to produce zero amplifier output when no NO is
present in the test bulb.  As NO is formed in the test bulb, or introduced
into it, a detector signal Va results.  The relationship between the ratio
Va/V,  and the amount of NO in the test bulb depends on the spectral  band-
pass as well as on the length of the correlation cell and the amount of
NO in it.  This ratio, V /V,  is defined as the normalized voltage V" and
is nearly independent of source brightness or detector sensitivity pro-
vided neither of these parameters change after Vb has  been measured.

     A photoelectric pick-up produces an electric signal at  the frequency
of the mirror chopper and synchronized with it.  This  signal serves  as a
reference for the phase-lock amplifier (Princeton Applied Research Model
JB-5).  The amplifier output is recorded on a strip chart recorder.   Elec-
trical noise due to the detector and amplifiers is less than 10"4Vb  and
does not contribute significantly to the uncertainty in the measurements.

     The test bulb is cylindrical, 5 cm diameter by 10 cm long, with a
quartz body and Ca?2 windows.  Both the actinic source and the test bulb
can be moved in a horizontal plane perpendicular to the narrow monitoring
beam of infrared energy passing through the test bulb.  Thus, the infrared
beam can sample different portions of the test bulb to check for non-
uniformity in the NO concentration.  If mixing is not rapid, non-uniformi-
ties might be expected to result from gradients in the intensity of the
actinic energy.  For test bulb mixtures containing enough N02 that the
absorption depth of the actinic energy is less than the diameter of the
test bulb, the actinic energy density is less on the side of the test bulb
away from the actinic source than it is on the opposite side.  Reflection
from the quartz wall can also produce non-uniform actinic energy density
within the test bulb.  For example, a line of high density may be expected
along the cylindrical test bulb midway between the center and the wall
farthest from the actinic source.  Reflection of the nearly-parallel beam
of actinic energy is focused along this line.  By using a cylindrical
test bulb, rather than a spherical one, it is possible to direct the infra-
red monitoring beam through a path parallel to the wall of the bulb and at
different distances from it.  Therefore, the actinic energy density is
                                   2-3

-------
nearly constant along the beam, and the effect of some of the geometrical
factors contributing to non-uniform actinic energy density can be investi-
gated more easily  than with a spherical test bulb.  Results obtained  with
the cylindrical test bulb have been used in determining the optimum dia-
meter of a spherical test bulb and the NC^ + CL mixture to be contained in
it.  The non-uniform directional sensitivity of the cylindrical test  bulb
does not create a  serious problem for preliminary tests.

      The irradiance by the artificial actinic source was approximately
the same as that to be expected by solar energy near sea level on a typi-
cal clear day.  Out-of-door measurements were made with the sun shining
directly on the cylindrical test bulb and the artificial actinic source
removed.

RESULTS OF PRELIMINARY TESTS WITH ACTINOMETER

      An example of the recorder traces of the actinometer output during
exposure of the test bulb to sunlight is given in Fig. 2-2.  Before the
trace was started,  the test bulb had been shielded from actinic energy
sufficiently  long  for all of the NO to disappear.  At t = 0, the shutter
was opened and  the test bulb was exposed to sunlight; the increase in V"
is due to the formation of NO in the test bulb.  After approximately
2 minutes of  exposure to the sun, the NO concentration was approaching
equilibrium.  A slight influence of a very thin cloud between the sun and
the actinometer can be seen from about t = 2 to t = 5 min.  At approxi-
mately t = 5  min.,  the intermittent shutter was turned on so that the
test bulb was alternately dark for 30 seconds, then exposed to the sun for
30 seconds.   We note  that the amplitude, A", of the variation in V" due
to the changes  in  NO  associated with the intermittent shutter stabilized
after only  1  or 2  minutes of intermittent exposure.  Other data not shown
indicate that the  drift in  the amplitude when the sky was clear was less
than the drift  in  the steady-state value of V" when the shutter was left
open for continuous exposure to the sun.

      With artificial actinic energy from a "black light" source
(Westinghouse H44-GS mercury vapor spot  lamp) substituted for sunlight,
responses such  as  those of  Fig. 2-3 were obtained.  At  t = 0, there was
essentially no  NO  in  the  test bulb because the shutter  had been closed
for several minutes to block the actinic energy.  At about t = 2 min., the
snutter was opened, exposing the test bulb to the artificial actinic  source
placed a distance  D = 20  cm from the center of the test bulb.  The shutter
was then closed at about  t  = 6 min., and D was varied by moving the arti-
ficial source as  indicated.  The increased response after  the shutter was
re-opened resulted from  the increase in  irradiance associated with the
decrease in D.
                                     2-4

-------
          0.02 .
   e
ro
 i
          0.01 .
                                                                  Intermittent Shutter
                                                            TIME (min)

              FIG. 2-2.  Response of breadboard model N09 actinometer to sunlight.  The test bulb was filled with
                         55 torr N02 plus 155 torr 02>  Exposure started at 12:29 P.M. (PDT) on June 29, 1973 near
                         Newport Beach, Ca. at an elevation of approximately 50 m above sea level.  The sky was mostly
                         clear, but hazy with occasional thin clouds.  The quantities AV" and V" for intermittent
                         shutter operation are illustrated.

-------
  0.06
  0.05
  0.04
 V"
  0.03
  0.02
  0.01
                                                           -l.fc
                         30                 60
                              Time (min)
90
FIG. 2-3.  Response of N0» actinometer to artificial uv and visible
           light at different  intensities and different positions
           in the test bulb.   D  is the distance from the  center of
           the test bulb to  the  face of  the black light lamp used as
           the artificial actinic light  source.  The distance from
           the axis of the test  bulb to  the monitoring IR beam in cm
           is denoted by R;  positive R indicates that the monitoring
           beam is between the axis of the test bulb and the actinic
           energy source, and  negative R indicates that the moni-
           toring beam is displaced in the opposite direction.  The
           intermittent shutter  is open  30 sec and closed 30 sec of
           each cycle.  The  test bulb was filled with 55 torr NO.
           plus 105 torr 09.
                                     2-6

-------
      At about t = 32 min., the actinic source and test bulb were moved
as a unit relative to the IR monitoring beam to check for non-uniformities
in the NO concentration in the test bulb.  The apparent shift in response
when the test bulb was moved is due primarily to a "zero" shift arising
from non-uniformity in the CaF_ windows on the test bulb.  The amount of
zero shift due to the windows was measured with the test bulb evacuated.
After the shift was accounted for, the results indicated there was no
significant difference in the NO concentration at the three different
positions of the test bulb for the N02 and 02 mixture given in the legend
of Fig. 2-3.  When the intermittent shutter was used, the amplitude AV"
was also essentially the same at all three values of R when D remained
fixed.  This provides additional evidence that the NO is mixed under these
conditions.

      Similar tests for non-uniform NO concentrations were also made with
higher and lower pressures of N02 and 02 in the test bulb.  With N02 pres-
sures greater than approximately 100 torr and 02 pressures greater than
approximately 150 torr, there was evidence that the NO concentration was
higher on the side of the test bulb toward the actinic light source.  At
the higher N02 pressures, a large fraction of the actinic energy is ab-
sorbed and forms NO before it has penetrated through the cell.  The higher
pressure also reduces the mixing rate in the test bulb.  Thus, there are
upper limits on the pressures of N02 and 02 that can be used in a spherical
test bulb of a field instrument in order that the concentration along the
diametric path of the monitoring beam represent the average for the entire
bulb.  This is required if the response is to be independent of the direc-
tion of the incident actinic energy.

      From Fig. 2-3 we see that the steady-state values of V" obtained
with the shutter open and the amplitude AV" of this quantity when the
shutter is intermittent both increase as the actinic irradiance is increased
by decreasing D.  The dependence of both V" and AV" on the actinic irradi-
ance are shown in Fig. 2-4 for the conditions represented by Fig. 2-3.  The
relative irradiances were measured with the probe of a "lite-Mike" (EG&G,
Inc. Model 560B) at the normal position of the center of the test bulb.
The steady state value of V" tends to saturate at much lower actinic
irradiance than does Av" obtained with the intermittent shutter.  A near-
linear relationship exists between V" and the actinic irradiance over a
wide range of irradiance values.

      Figure 2-5 illustrates the dependence of the actinometer response on
the partial pressures of NO- and 02«  Values of AV" increase nearly in
proportion to the %QO «ith little or no 02 in the test bulb.  Adding 02
to a fixed amount of N02 increases the amplitude AV  because it increases
the rate at which the NO returns to N02 during the dark half of each cycle.
In the upper panel, the plot of V" for PNQ  = 72 torr illustrates the de-
crease in the equilibrium amount of NO witfi increasing 0  when the test
                                   2-7

-------
 0.01
0.008  _
0.006  _
0.004
0.002  _
       0
20
60
80
100
                  Actinic  Irradiance  (Arbitrary Units)
     FIG. 2-4.  Plots of V" and AV" vs relative actinic irradiance.
                The ordinate for each curve is labelled.
                                      2-8

-------
    0.04
    0.03
    0.02
 V"
    0.01
   0.010
                                                                ..i
   0.008
                     72
AV"
   0.006
   0.004
   0.002
   /  // /
{///  *


                                           J_
                                              X
                   50         100         150         200

                         0, partial pressure (torr)
                                                        250
    FIG. 2t5.   Plots of V"  and AV"  vs  0  partial  pressure in  the  test  bulb
               for various  partial  pressures  of NO-,.  The partial pressures  of
               NO. in torr  are indicated  for  each curve.   A uv black light
               without a filter served as the source  of actinic energy.  The
               monitoring beam passed  along   the  center of the 10 cm long test
               bulb.  While measuring  AV", each shutter cycle consisted  of
               periods of 30 sec open  and 30  sec  closed;  V" was measured after
               the shutter  had been open  several  minutes.
                                     2-9

-------
bulb  is  irradiated  continuously.   This  result  is  to be expected because
of  the increase  in  the  rate  of  the reverse  reaction producted by the addi-
tional $2-  The  measurements of V" for  the  lower  values of I^Q  were not
reproducible and were inconclusive.  Variations in temperature and the
amounts  of  impurities in  the test  bulb  probably account for the changes
in  V" while the  actinic irradiance is maintained  constant.  These vari-
ations apparently have  less  influence on AV" than on V".

      Increasing the pressure of either N02 or 0, increases AV"; however,
as  pointed  out above, if  the pressures  are  too high the actinic energy
density  and the  NO  concentration are not uniform  throughout the test bulb,
and the  interference by ^0^ is significant.   The optimum pressures for a
test bulb of a given diameter in a field instrument are determined by the
best compromise  among the spectral response, uniform directional response,
and signal-to-noise ratio.   Therefore,  the optimum pressures depend upon the
detectivity of the  detector  and the radiance  of  the IR source and can best
be determined after an  instrument  is assembled.  Decreasing the 0  pressure
Lncreases the required  period  for  the shutter because of the decrease in the
rate at which the NO disappears during the dark portion of the cycle.

      A few tests were  made on the  influence of temperature  changes  in the
 test  bulb while  maintaining the amount  of  gas  in  it and  the  actinic irradi-
 auce  constant.   An infrared heat  lamp heated  the  test  bulb,  and  a  thermo-
 couple,  inserted into  the interior of the  bulb but  shielded  from the
 direct  radiation of the heat lamp, measured the temperature  of  the  gas  in
 the bulb.   For  a typical gas mixture, AV"  increased by approximately  97
 as  the  temperature changed from 300 K to 320  K;  an  additional 20 K  increase
 produced an additional 5% increase in V".   The value  of AV"  quickly
 stabilized after the temperature  had been  changed.  Steady-state values
 of  V" obtained with the shutter open were  more dependent  on  temperature
 and much less  stable than those of AV".

      We  have  performed other tests on NO detection  in  our laboratory  with
 a  gas cell correlation instrument  employing the same  bandpass filter  and
 correlation cell as those illustrated in Fig.  2-1.  Absorption  by water
 vapor in the  atmospheric  path in  the instrument interferes with  the NO
measurement and  produces  a shift  in the zero  reading  that corresponds to
 the test bulb  being free  of any NO. If  the H»0 concentration is  constant
 it  can  be accounted for by adjusting the zero  setting  when  the  test bulb
 has been shaded  from actinic energy sufficiently  long  for all of the  NO
 to  disappear.  However, variations in the  H?0  concentration  can  cause
 zero  shifts that produce  significant errors in measurements  of  V".  Errors
 in  AV"  produced  by H»0 interference are much  smaller  because this quantity
 represents the decrease or increase in  V"  each time  the  shutter  is  closed
 or  opened.   If  several measurements are repeated, the  errors cancel each
 other out.
                                    2-10

-------
     Results obtained in the sunlight with a test bulb having a boro-
silicate glass body were similar to those obtained with a test bulb made
of quartz, which is known to be transparent in the spectral interval of
the actinic energy.  Although the glass is known to absorb strongly at
the shorter wavelengths of the actinic energy, values of AV" were only
about 10% less for the glass test bulb than for the quartz one.  A greater
difference might be observed where shorter wavelengths comprise a greater
proportion of the actinic energy than that transmitted through the hazy
atmosphere of these tests.


MONITORING OF N02 VS NO

     Nitrogen dioxide has at least three absorption bands  in the infra-
red and visible that are worthy of consideration for a gas cell correlation
instrument designed to monitor this gas.  The strongest infrared band
near 6.2 \i,m is probably not practical for an instrument with a signifi-
cant optical path in the atmosphere because of excessive interference by
the strong, over lapping'lUO absorption.  Furthermore, no convenient de-
tector with adequate detectivity without cryogenic cooling exists for this
band.  Considerable structure exists in the spectrum of the 3.4 u-m band
so that it lends itself to the use of gas cell correlation techniques.
This band is also relatively free of interfering absorption by H»0 or other
atmospheric gases.  However, its band strength is only moderate so that
the  sensitivity of an instrument employing this band is limited.  A
relatively convenient detector with a thermoelectrically cooled PbSe ele-
ment would probably be adequate for the 3.4 pm region.

     As pointed out in Section 1, most of the NO? absorption in the uv is
continuous without enough structure in the spectrum for a gas cell corre-
lation instrument to operate efficiently.  We have investigated the use
of the absorption features in the visible spectrum. These features are
quite distinct and consist of many very closely spaced lines that overlap
except for very low sample pressures.  Fig. 2-6 shows spectral curves of
transmittance obtained with a spectral slitwidth of approximately 2 A.
Absorption is quite strong, even at wavelengths adjacent to the stronger
absorption features.  Thus, the strong absorption features are not sepa-
rated by narrow regions of little or no absorption as are the lines in
the fundamental NO band near 5.4 M>m.  Because of the absorption between
the wavelengths of strongest N0« absorption, a gas cell correlation in-
strument is less efficient than we had originally expected.  Our tests
on these visible NO- bands were limited to the measurement of N02 in a
conventional absorption cell, rather than in a test bulb with NO2 being
dissociated photochemically.

     We have used the actinometer illustrated in Fig. 2-1 to obtain data
such as those represented by Fig. 2-2 as well as to study stable samples
of NO + N2 in the test bulb.  By comparing the results, we have been able
                                  2-11

-------
                                             100%
to
I
u
u
z
s
£
                                07,
                      4300
                                                        4400
                                                      WAVELENGTH
                                                                                             4500
                                             (X)
                    FIG. 2-6.  Transmittance  of  NC>2  from 4300  to 4500 A.   The sample cell length
                               was 3.96 cm, the  pressure of  NO 9  was  25 torr,  and the temperature
                               was 51"C.

-------
 to estimate the NO concentration in the test bulb from the measurement
 of V".  During the period when the shutter was cycling to obtain the data
 shown in Fig. 2-2, the concentration of NO was oscillating from about 4%
 to 6% of that of the N02-  It follows that the N02 concentration was
 varying from approximately 94% to 96% of its original value.  If the acti-
 nometer were made to measure the N02 concentration, rather than NO, it
 would be required to measure the small relative change quite accurately.
 Because the relative change in the NO concentration is much greater, the
 required accuracy for its measurement is much less.  For example, the
 temperature of the test bulb can be expected to oscillate slightly in
 synchronism with the opening and closing of the shutter over the test
 bulb.  The small temperature change can influence slightly the absorption
 characteristics of both the N02 and the NO and produce an error in the
 measurement of AV".  Such an error would be more serious when measuring
 the small relative changes in N02 than when measuring the larger relative
 changes in NO.

     After considering the encouraging results obtained with the actinometer
 measuring NO concentrations and the problems anticipated in measuring the
 changes in NO. concentrations, we decided to build the final instrument
 to measure the changes in NO.


 HOT NO AS AN IR ENERGY SOURCE

     The use of a hot gas cell containing NO as a source of infrared energy
 in place of a conventional continuous source has been considered, and a
 few related experiments have been performed.  Such a source would provide
 the energy for the monitoring beam in an actinometer that monitored the
 formation of NO in the test bulb.  The hot NO in the source cell emits
 strongly near the centers of the absorption lines which, of course, are at
 the wavelengths of strong absorption by the NO in the correlation cell and
 in the test bulb.  The length, L, of the hot source cell and the pressure
 PNO, of NO in it should be adjusted so that the product L PNO is somewhat
 greater than the corresponding quantity for the NO in the correlation cell.
 On the other hand, it should be small enough that there is not appreciable
 emission at wavelengths midway between adjacent lines.

     Little information about NO in the test bulb can be obtained from the
wavelengths well separated from the line centers.  Therefore, the extra
 energy emitted by a continuous source at these wavelengths produces little
 information, but it can contribute to misbalance, drift and interference
 by other gases that may be present in the optical path.  Because of the
 coincidence of the emission lines of the source gas and the absorption
 lines of the absorbing gas in the test bulb, a fixed amount of absorbing
 gas absorbs a larger fraction of energy from the hot gas source than from
 a continuous source.   For a low concentration C of NO in the test bulb of
 optical path length LXB,  the normalized voltage V" is related  to C by
                                    2-13

-------
          V" = Va/Vfa = k  C 4B-                                     (2-1)


It follows from  the above discussion that the constant k can be larger
when using a hot NO source than when using a continuous source.  Thus a
given percentage error in balance of the instrument yields a smaller error
in the measurement of C when using  the hot NO source.

     Hot gas cells generally cannot be operated at temperatures as high as
continuous sources, with  the result that both Va and V,  are lower for hot
gas sources.  In the case that the  instrument accuracy is limited pri-
marily by detector noise, the increased energy available from the con-
tinuous source makes it preferable  to the hot gas source.  The results of
the tests made with the actinometer shown in Fig. 2-1 indicated that the
larger values of k and the better discrimination available with the hot
NO source would not be as important for the actinometer to be built as
would the extra signal available with the continuous source.  The extra
signal is required because it is desirable that the actinometer not re-
quire a cryogenically-cooled detector; therefore, the detector detectivity
is much lower than that of the liquid nitrogen-cooled detector used in the
preliminary tests.


CONCLUSIONS DRAWN FROM PRELIMINARY TESTS

     The results of the tests reported here indicate the feasibility of an
N02 actinometer for field use in which the concentration of photochemically-
produced NO is monitored  by non-dispersive infrared techniques.  Accurate
steady-state values of NO concentration are difficult to determine because
of instabilities in the optical components and variations in the tempera-
ture of the test bulb.  Other slow  chemical reactions not considered here
may also produce slow changes in the NO concentration.  In contrast, re-
sults that are more consistent and  easier to interpret can be obtained by
using alternate  light and dark periods of approximately 30 to 60 seconds
each to produce rapid changes of NO concentration around a mean level.
The alternating method has the problem that the total response is reduced
substantially and the cycle periods must be controlled carefully.  How-
ever, these problems are  more than  offset by the improved stability.  The
relatively slow response  in both methods precludes the measurement of
small, rapid fluctuations of the actinic light level.  Monitoring the con-
centration of NO can apparently provide a more accurate measure of the
actinic energy than is possible by monitoring the N02«

     Although the optimum instrument parameters depend on a variety of
operating conditions, it  is possible to prescribe approximate parameters
for a potentially useful  tool for measuring the relative available actinic
light under most natural  conditions of interest.  A spherical, quartz
test bulb 10 cm  in diameter filled  with a mixture of N02 + 02 can be made
                                    2-14

-------
with its sensitivity to light nearly the same from all directions in the
upper hemisphere.  The amount of NO produced under normal atmospheric con-
ditions is sufficient to be measured by non-dispersive infrared techniques
employing a gas correlation cell.  The addition of an inert gas such as N_
to the mixture in the test bulb is not required and would reduce the rate
of mixing of the gases and increase non-uniformities in the NO concentra-
tion.

     A small, readily available InSb detector with thermoelectric cooling
to about 250 K can probably provide an adequate signal-to-noise ratio
when used with a conventional infrared energy source.  Cooling the de-
tector to lower temperatures with liquid nitrogen is probably not required.
The advantage of improved discrimination that could be achieved by using
hot NO as the Infrared source is more than offset by the increased com-
plexity and decrease in signal.

     Windows on the test bulb for the transmission of the IR monitoring
beam and associated mirrors necessarily decrease the sensitivity of the
system to actinic light arriving from certain directions.  By properly
orienting the test bulb relative to the sun so that the windows and mirrors
do not interfere with the direct sunlight, it is possible to minimize the
error caused by these interfering optical components.
                                    2-15

-------
                               SECTION 3


                       DESCRIPTION OF ACTINOMETER
OPTICAL LAYOUT

     The requirements described above and the results of the preliminary
tests described in Section 2 led to the design and construction of the
instrument illustrated in Figs. 3-1, 3-2 and 3-3.   The upper panel of
Fig. 3-1 shows an optical schematic diagram as viewed from above.   Infra-
red radiant energy is directed by mirror Ml through a correlation cell
with an image of the source formed at the reflecting plane of the  mirr&r
chopper.  When the notched, mirror chopper is in the open position, the
beam continues on and another image of the source  is formed on the small,
flat mirror, M8, which is tilted approximately 22.5° from the vertical so
that the beam is reflected at 45° upward to mirror M9, which returns the
beam displaced slightly sidewise to a CaF2 lens.  A reduced image of the
source is formed on the detector.

     During the other half of each cycle the mirror chopper, the energy
falling on the detector is directed by mirror M3 through an attenuator to
the reflecting surface of the mirror chopper.  The fixed portion of the
attenuator is a sapphire window, and the adjustable portion blocks a
portion of the beam.  Beyond the mirror chopper, the beam travels the same
path during both halves of the cycle.  Mirror M8 is sufficiently short that
it limits the useful length of the source from which radiant energy is
collected.  Spherical mirror M7 images mirror M6 near mirror M9, which
is bonded to the quartz test bulb and forms the aperture stop.  The mask
on mirror M6 reduces scattered light that would not be collected by M9.
A narrow bandpass filter similar to the one shown in Fig. 2-1 is placed
in front of the detector and limits the spectral band to a region approxi-
mately 0.15 microns wide centered near 5.3 microns.
                                    3-1

-------
                         MIRROR
                         CHOPPCR
Fig. 3-1.  Optical diagrams  of  the  optics  assembly of the actinometer.  The upper
           panel is a top view  with the  instrument in its normal operating position.
           A side view of a  portion of the instrument is shown in the lower panel.
                                         3-2

-------
FIG. 3-2.  Two views of the optics assembly of the  actinometer.
           The upper panel shows the instrument completely assembled
           with the shutter open.  The lower panel  shows  a top view
           with the shutter open and the lid removed.   S,  source;
           CC, correlation cell; Ch, mirror chopper; TB,  test bulb;
           Pre, preamp.;  P, power supply.
                              3-3

-------
FIG.  3-3.  Photograph of the control assembly, amplifier and recorder.
            The control assembly contains switches and a timer assembly
            for remote control of the shutter and the recorder.
                                 3-4

-------
     Approximately 10 torr of NOo plus 110 torr of 0_  has  been added  to
the quartz test bulb.  As the test bulb is irradiated  with actinic  energy,
from the sun or from an artificial source, part of the N02 is  photolyzed to
produce NO and 0~ according to Eq. (1-1).   The shutter periodically covers
the quartz test bulb to shade it from actinic energy.   Out of  each  2  min
period of the shutter, it is open approximately 55 sec and closed for 55
sec.  Approximately 5 sec are required for the shutter to  open or close.
The primary purpose of the instrument is to measure the change in the
concentration of NO in the test bulb during the two halves of  the shutter
cycle.

    The spectroscopic principles of operation of this  instrument are  the
same as those of the breadboard instrument shown in Fig. 2-1.   After  the
shutter has been closed sufficiently long that essentially all of the NO
has disappeared from the test bulb, the attenuator is  adjusted so that
the same amount of infrared energy is incident on the  detector during
both halves of cycle of the mirror chopper.  Thus, there is no detector
signal at 350 Hz, the chopper frequency.  The adjustable attenuator is
mounted directly to a micrometer screw and is moved into or out of the
beam.  After the instrument has been balanced, the quantity Vj, is measured
by moving the blocking attenuator to the position indicated by the broken
line in the upper panel of Fig. 3-1.  This blocks all  of the energy that
normally passes through the attenuator with the result that the infrared
energy reaching the detector is completely modulated by the mirror chopper.
After the quantity VD has been measured and recorded,the blocking attenu-
ator is moved to its normal position before measurements are made.

DESCRIPTION OF OPTICAL AND MECHANICAL COMPONENTS

    The source is a ceramic-coated resistance wire that is heated elec-
trically by approximately 10 amperes of alternating current at 2.5 V.
This source is commercially available from Perkin-Elmer Corporation and
receives power from a small transformer.  Its spectral radiance in the
region of interest, near 5.3 microns, is only approximately half that of
a Nernst glower under normal operating conditions.  However, this source
is easier to operate, and.its durability makes it more practical for a
field instrument.

    The mirror chopper is machined from stainless steel and has 6 reflec-
ting blades and 6 openings of equal width.  The reflecting surface has
been coated with a thin  layer of aluminum and polished to produce good
reflectivity.  The chopper is mounted directly to the shaft of a small
induction motor that rotates at approximately 3500 rpm to produce a
chopping frequency near 350 Hz.  This frequency is sufficiently high
that the detectivity of the detector is not seriously limited by the  1/f
noise.  In order to obtain higher chopper frequencies, either more blades
would be required on the chopper or it would need to be rotated at a
higher frequency.  Either of these methods has obvious disadvantages.
                                   3-5

-------
    The  detector  contains  a  2 mm x 2 mm sensitive element of InSb that is
cooled to approximately  -25°C by a small  thermo-electric cooler that re-
ceives its  power  from  a  dc power supply mounted in the position shown in
Fig. 3-2.   A  separate  supply powers the preamplifier and the reference
pick-up  and produces bias  voltage for  the detector.

    Calcium fluoride windows are sealed by epoxy cement to the 1 cm long
by 2.5 cm diameter  correlation  cell.   Approximately 680 torr of NO in
the correlation has an average  transmittance of 72% in the spectral band
of the bandpass filter.

    The  test  bulb is approximately  10  cm  in diameter and is made from the
spherical bulb of a quartz beaker.  Holes approximately 2.2 cm in dia-
meter have  been cut in the walls; CaF« windows are bonded with epoxy
cement over the holes.   The  quartz wall had to be removed from the windows
because  quartz is opaque to  the 5.3 micron infrared energy in the moni-
toring beam.. A piece  of Monel  tubing  has been bonded in a smaller hole
cut in the  bottom of the test bulb.  A small Monel valve welded to the
tubing makes  it possible to  evacuate the  test bulb and refill it with
any desired mixture at a pressure less than 1 atm.  It is anticipated
that the mixture  of NO-  +  02 in the test  bulb can be used for several
months without refilling.

    The  test  bulb is mounted above  the remainder of the instrument in
order that  it can receive  actinic energy  from as much of the upper hemis-
phere as possible.  It seems likely that  a negligible portion of actinic
energy would  be reflected  from  the earth's surface.  Thus, most of it
would be incident on the test bulb from a direction above the horizon.
Ideally, the  bulb should have no obstruction that would shade any part of
it from  any part  of the  upper hemisphere.  However, it is necessary to
obstruct small portions  of the  test bulb  in order to pass the infrared
monitoring  beam across a diameter.  The CaF. window on the lower side of
the test bulb is  not expected to obstruct a significant portion of the
actinic  energy from the  sky.  The window  and mirror M9 on the upper
portion  of  the test bulb probably cause the most obstruction and produce
the biggest deviation  in the sensitivity  from all directions.  On most
days, it is expected that  a  large fraction of the actinic energy will
come from a direction  within several degrees of the sun.  Therefore, the
influence of  the  windows and mirror M9 can be minimized by orienting the
instrument  so that  these parts  produce a  minimum of obstruction of light
coming directly from the sun.   This is best accomplished by locating the
instrument  with the end  containing mirrors Ml and M3 pointed in the
azimuthal direction of the sun.  Unless the sun is very high in the sky,
shading by  mirror M9 should  not be serious.  It seems likely that the
shading of  the test bulb by  the windows and the mirror is less under most
circumstances when oriented in this manner than would be possible if
mirror M9 were located on  the top of the  bulb with the monitoring beam
                                   3-6

-------
traveling along a vertical diameter.   Double passing the test bulb with
the infrared monitoring beam increases the sensitivity of the system
without enlarging the diameter of the test bulb.   Although mirror M9
shades part of the test bulb, removing it so the  beam single-passes the
test bulb would not significantly decrease the shading because another
mirror would need to be placed near the bulb to direct the beam to the
detector.

    In order to further minimize any obstruction  of the test bulb from
light in the upper hemisphere, the legs on the instrument are made so
that it slopes downward at about 5° from the test bulb toward the end
containing mirrors Ml and M3.  The shutter, when  open, also is below
the horizon from any point on the test bulb.

    The cycling shutter is chain driven by a small gear motor that is
turned on and  off and reversed in direction by a timing mechanism mounted
in the control panel described below.  Microswitches actuated by the
shutter turn the motor off as it approaches either the closed or open
position.  A small slip-clutch mounted to the shaft of the motor allows
the shaft of the gear motor to continue turning for a short time after
the power has  been turned off and the shutter has reached the end of its
travel.  It is necessary to allow for this small amount of slippage because
of the angular momentum of the armature of the motor.

    Fig. 3-3 shows the control unit,  the synchronous amplifier, and the
strip chart recorder.  On-off switches on the front panel of  the control
assembly operate the chopper, the source, and the detector.  When the
shutter-control switch is in  the manual position, the shutter can be
either opened  or closed by holding a  spring-loaded switch in  the appropri-
ate position.  The automatic  timer controls  the action of the shutter
when the shutter-control switch  is in the automatic position.  If no
electrical  power is available, the shutter  can be opened or  closed by
forcing  it  slowly and allowing the clutch to  slip.

    When the chaxt recorder-control  switch  is in  the manual  position,  the
pen remains in the writing position  and  the  chart is  controlled by  the
switch on the  front of the recorder.   With  the chart  recorder-control
switch in the  automatic position, both the  chart  drive and  the pen may
be deactivated by the timer.  In the automatic mode,  the chart drives
and the  pen writes only for  two  intervals  of  about  10  seconds each  during
each shutter cycle when V" is near its minimum and maximum  values.   By
recording only during these  intervals,  the  amplitude  that corresponds  to
iiv" can  be  measured easily and  less  chart  paper  is  required.  Recorder
tracings that  illustrate  the use of  the  chart recorder in  the automatic
mode appear in Section 4.  The period of  time that  the pen  writes  on
the strip chart  can be changed by adjusting the  appropriate cams  in the
timer  assembly.
                                     3-7

-------
     A  push-button switch on the  front  panel  of  the control unit actuates
an  "event-marker" pen that records  on the  right-hand side of the strip
chart.   The  electrical cables that  connect the optics assembly to the
control unit,  recorder,  and amplifier are  approximately 8 meters long and
can be  removed for easy  handling.   Detailed wiring diagrams are shown in
Section 5.

SET-UP  AND OPERATION

     The instrument is designed for use in the position shown in the upper
portion of Fig.  3-2.   The bench or  table on which it is placed should be
small so that  it subtends a small solid angle from the test bulb and thus
blocks  a minimum of actinic energy.  Other structures that block actinic
energy  from  the  bulb  should also be avoided.  As pointed out previously,
the amount of  energy  blocked by mirror  M9  on  the test bulb is minimized
by  orienting the instrument so that the right-hand end in Fig. 3-2 is in
the azimuthal  direction  of the sun.  The outside surface is blackened to
minimize reflection of actinic energy to the  test bulb.

     The power supplies  and the source  produce enough heat inside the
instrument cover to increase the  temperature  by  several degrees.  Over-
heating of the power  supply for the detector  coolers is avoided by heat-
sinking it to  one side of the box.   It  is  recommended that a fan be used
to  circulate air over the optics assembly  to  reduce its temperature, parti-
cularly if it  is being used in hot,  sunny  weather or in a hot test chamber.
Increased temperature degrades the  signal-to-noise ratio of the cooled
detector.  Excessive  temperature changes in the  test bulb also change the
response of  the  instrument.

     All of  the  electrical plugs and connectors  are labeled so that the
instrument can easily be connected  in accordance with Fig. 5-2.  A few
minutes  should be allowed for warm-up before  the following adjustments
are made.  The beam in the attenuator side of the alternator is blocked
by  moving the  knob for the blocking attenuator to the BLOCK position.  The
knob is  located  near  the chain-drive for the  shutter.  Adjustments are
then made to the amplifier frequency and phase to produce a maximum posi-
tive output  of the amplifier.  The  amplifier  frequency reading should be
approximately  350 Hz,  the chopper frequency.  Positive output corresponds
to  a reading to  the right of center  on  the meter on the amplifier panel
with the meter-monitor switch in the OUT x 1  or  OUT x 10 position.

     Zero and  span adjustments on the amplifier  and recorder are made in
the following  manner.  Set the meter-monitor  switch to the OUT x 10
position and disconnect  the signal  input cable to the amplifier.  Use the
zero adjustment  on the amplifier to  set the meter reading to zero (center).
Next, adjust the recorder zero potentiometer  on  the front panel of the
recorder to  produce a zero reading  of the  recorder.  Set the meter-monitor
                                    3-8

-------
switch on the amplifier to the OUT x 1 position and rotate the zero-
adjustment potentiometer on the amplifier clockwise until the panel meter
reads full-scale (10) to the right.  Now, use the recorder-adjust poten-
tiometer on the rear of the amplifier to produce a full-scale reading
of the recorder.  Return the zero-adjustment potentiometer of the ampli-
fier to the position that produces a zero, or mid-scale, reading on the
meter.  Reconnect the signal input cable to the amplifier.  The recorder
is now spanned so that a full-scale reading corresponds to full-scale of
the amplifier panel meter and is within the useful dynamic range of the
amplifier.  A 1 K ohm resistor across the input terminals of the recorder
matches the recorder to the amplifier output.  A zero off-set of the
recorder can be obtained by adjusting the zero-potentiometer on the
recorder.

     In normal operation, the chart speed-control switches on the recorder
and on the control assembly are in the LO position.  When the control
switch of the control assembly is in the FAST position, the chart speed
is opposite to that indicated by the switch on the recorder.  The chart
speed-control switch on the control assembly is not needed but was in-
stalled before the recorder was delivered.

     From Section 2, we recall that the instrument output of Interest is
AV", which is proportional to AV/V, > where VL is the output with the
attenuator leg of the alternator blocked, and AV is the change in the
output as the shutter periodically opens and closes.  (See Fig. 2.2.)
Before starting a series of measurements, the blocking attenuator is
moved to the BLOCK position to block the beam in the attenuator leg and
V,  is measured.  At the time the instrument was delivered to EPA, Vj,
corresponded to a recorder-pen of approximately 17 cm with the amplifier
on 20 mv scale.  In order to simplify the conversion of AV to AV", the
pen deflection can be adjusted to produce an even-numbered deflection,
such as 20 cm.  This is done by adjusting the calibration-potentiometer
on the input of the amplifier.  It is best if this is done after the
shutter has been closed sufficiently long that little NO remains in the
test bulb.

     After V,  has been measured and recorded, the blocking attenuator is
removed from the beam and the alternator is balanced with the adjustable
attenuator while the shutter is kept closed to avoid the formation of NO
in the test bulb.  The instrument is then ready for operationin the manner
described previously.  Ordinarily, the system is balanced with the ampli-
fier on a more sensitive scale, either 1 mv or 2 mv full-scale.  Measure-
ments of AV are also made on the most sensitive scale. In relating the
pen deflections to V" or AV", the different sensitivities of the ampli-
fier must be accounted for.  For example, assume that a recorder-pen
deflection of 20 cm was observed when measuring Vy, with the amplifier
on the 20 mv scale.  When the actinic irradiance is being measured, the
amplifier is on the 1 mv scale.  As the shutter opens and closes during
                                   3-9

-------
its cycle, the pen deflection varies from a maximum of 13 cm to a mini-
mum of 9 cm, giving a difference of 4 cm.  If V,  had been measured on this
scale, it would have corresponded to a pen deflection of 20 x 20/1 = 400
cm.  Therefore, AV" = 4/400 = 0.010.

     After several minutes of warm-up, Vu remains nearly constant for
several hours unless the instrument temperature changes significantly.
Thus, the frequency that V. needs  to be checked depends on the weather
conditions i£ the instrument is outside, or upon the temperature in a
test chamber.
                                  3-10

-------
                                SECTION 4
                TEST RESULTS AND INSTRUMENT PERFORMANCE
RESPONSE TO SOLAR ENERGY

     The recorder tracing shown in Fig. 4-1 illustrates the actinometer
output in the afternoon of a clear day in early June in Newport Beach,
California.  The instrument was located on a roof top with essentially
clear vision above the horizon in all directions.  Before time t = 0,
the actinometer had been balanced with the shutter closed so that no NO
was present in the test bulb.  At t = 0, the shutter was opened; thereafter
the chart drive and the pen operated continuously while the shutter opened
and closed with a two-minute period.  At t = 0, 2, 4, 6 	 minutes the
shutter opened; it closed at t = 1, 3, 5, 7 	 minutes.  While the shutter
was open, V" did not increase linearly with time during the full one-minute
period.  This non-linearity contributes significantly to the non-linear
relationship between AV" and actinic irradiance.

     The first peak is not as high as the following ones because the NO
concentration had not reached its maximum value when the shutter closed
at t = 1 minute.  After two cycles of opening and closing the NO concen-
tration has essentially reached the point so that the maximum and minimum
values of V" repeat themselves if the actinic irradiance is constant.
Typically, the maximum value of V" attained during cyclic operation of the
shutter is between 2\> and 3 times AV", the difference between the maximum
and minimum values of V".  Detector noise is the major cause of small,
high-frequency variations in V".  The peak-to-peak noise equivalent V"
is typically less than 0.001 with an amplifier time constant of 1 second.

     Recorder tracings of the actinometer output are shown in the upper
panel of Fig. 4-2 for both the recorder and shutter operating in the
                                    4-1

-------
                                                                                                                                    0.03
                                                                                                                                    0.02
                                                                                                                                      V"
                                                                                                                                    0.01
                                                          TIME (roin)
FIG. 4-1.  Response of actinometer to solar energy with the shutter  in  the  automatic mode.

-------
                                                                                                                              0.03
                                                                                     ^y
                                                                                                                              0.02
                                                                                                                              0.01
                                                                   D    C
                    L
"I	
V
          o.oo
                                                               TIME
                                                                                                                           _ 0.03
                                                                                                                           — 0.02
Fig.  4-2.   Recorder tracings of  actinometer output  when  the sky contains  broken clouds.  L'pper  panel:  The shutter  is cycling
           and the recorder chart .drive and pen are in the automatic mode.   Lower panel.  The shutter is closed and  the
           recorder is running continuously.

-------
automatic mode.  From  point A  to  point  B,  the shutter has been closed
sufficiently  long  that NO  exists  in  the test bulb.  At point B, both the
recorder and  shutter are started  in  the automatic mode.  In this mode, the
chart drives  and the pen is in the writing position only for short periods
when V" is near its minimum and maximum values.  The shutter opens or
closes during these short  periods.   Values of AV" can be determined from
these condensed strips of  chart by measuring between the minimum and maxi-
mum values of V" as illustrated in Fig.  2-2.

     From point B  to point C,  the sky contained several broken clouds with
some thin clouds moving into and  out of the direct path between the sun
and the test  bulb.  At point C, no apparent clouds were in the direct path.
From point D  to point  E, a cloud  formed in the path so that at point E
there was enough scattering by the clouds  that no apparent shadows were
cast by trees or buildings in  the neighborhood of the instrument.   This
type of cloud cover remained from point E  to point F.

     At point F, the shutter was  closed and remained closed while the chart
drove continuously and recorded the  curve  shown in the lower panel of
Fig. 4-2.  From point  F to point  G,  V"  decreased continuously, indicating
a decrease in the  concentration of NO in the test bulb.  It is apparent
that the rate of disappearance of NO decreases as its concentration decreases.
Between points G and H the infrared  monitoring beam was blocked in order
to record the pen  deflection corresponding to the absence of NO in the test
bulb.

RESPONSE VS ACTINIC IRRADIANCE

     The curve in  Fig. 4-3 relates AV"  to  the relative actinic irradiance.
A "black-light" placed approximately 20 cm from the test bulb served as the
source of actinic  energy  to obtain  the  data on which the curve is based.
A large, segmented chopper blade  rotating  at  1 revolution per second
blocked the beam of actinic energy  from the test bulb  for different fractions
•of the time.   The  data point corresponding to a relative irradiance of
100 was obtained with  this chopper  blade stopped in the open position so
that the test bulb was fully irradiated.  By rotating  the chopper with the
blades adjusted so that  the chopper  was half open and  half closed, it was
possible to obtain the data point corresponding to a relative irradiance
of 50.  Similarly, a blade that blocked the beam 25% of the time was used
for a relative irradiance  of 75%, etc.   Because of the 1 rps rate of the
chopper blade the  length  of  time  that  the  beam was transmitted, or blocked,
is short compared  to  the  reaction time  for the formation or decay or NO in
the  test bulb. Therefore, it  is  expected  that decreasing the duty cycle
'in  this manner has the same effect  as  decreasing the actinic  irradiance by
the same factor.

     The non-linearity in  the  relationship between actinic irradiance and
AV" arises  from a  combination  of  two factors.  When  the NO concentration
                                    4-4

-------
      0.012
      0.010
      0.008
AV"
      0.006
      0.004
      0.002
                                                  60
80
100
                               Relative  Actinic  Irradiance
          FIG.  4-3.   Plot  of AV"  vs  relative  actinic  irradiance.
                                         4-5

-------
in the  test bulb  is  sufficiently  high  that the absorption is appreciable
at the  centers  of the  absorption  lines,  the amount of energy absorbed no
longer  increases  linearly  with  the NO  concentration.  Secondly, the net
rate at which NO  is  formed during each 1-minute period that the shutter
is open decreases as the NO concentration increases.  By reducing the
period  of  the cycling  shutter,  the non-linearity resulting from the change
in the  rate of  NO formation could be reduced.  However, this would also
reduce AV", and thus the signal-to-noise ratio.

RESPONSE VS DIRECTION  OF ACTINIC  ENERGY

     As discussed in the first  part of this section, we expect the response
of the actinometer to  be somewhat dependent upon its orientation relative
to the sun, particularly when there are but a few clouds in the sky and
most of the actinic energy arrives from a direction near the sun.  In
order to approximate the effect of direction we obtained the data summarized
in Table 4-1 in the  P.M. of May 1, 1974 when the sun was approximately
45° above  the horizon.  The first measurement was obtained with the acti-
nometer sitting on a horizontal table  so that angle a = 0.  The long side
of the actinometer was  in  the azimuthal direction of the sun with the
end containing  the test bulb farthest  from the sun.  The other data were
obtained by rotating the actinometer clockwise around a vertical axis so
that ot was equal  to  the angles  indicated.  Approximately 10 minutes were
required for the  measurement at each angle.  Three of four cycles of the
shutter were used, and  the data for all of the cycles were averaged.

    As expected,  the maximum response  was observed when a = 0  , and the
minimum occurred  at a  = 180° when  M9 provided the maximum shade on the
Lest bulb.  A few wispy clouds  in the  sky probably caused a few percent
variation  in the  measurements.  During the 50-60 minutes between the two
times that the  measurements were  made  at a = 0, the change in AV" corres-
ponded to  approximately art  8%  decrease in irradiance.  Part of this
decrease probably resulted from the increase in the zenith angle of the
sun and the corresponding  increase in  the atmospheric path.  Some of the
decrease may have resulted from increased thickness of the thin clouds.
Although the actinometer response is somewhat directionalt a relatively
small fraction  of the  energy is usually incident from the portion of the
sky opposite the  sun.   Therefore, it is  expected that errors due to the
direction-dependent response will not  exceed a few percent if  the instru-
ment is oriented  as prescribed  above with 01 = 0°.
                                    4-6

-------
                                TABLE 4-1

       RESPONSE VS ORIENTATION OF ACTINOMETER RELATIVE  TO THE SUN
               Angle Qf                             &V"
              (degrees)

                0                                0.0111
                45                               0.0109
                90                               0.0104
                135                              0.0099
                180                              0.0098
                0

RESPONSE VS NO CONCENTRATION

     Although the quantity of primary interest is AV",  it  is  also of interest
to know the approximate concentration of NO in the test bulb  at  any time.
Figure 4-4 shows a curve with which the NO concentration can  be  related to
the instantaneous velue of V".  The data on which the curve is based were
obtained with mixtures of NO + N, in a sample cell in place of the test bulb.
The absorber thickness of NO is equal to the partial pressure of NO in atmos-
pheres times the optical path length of the sample cell.  Since  the optical
path of the test bulb is approximately 21 cm, the NO concentration is (1/21)
times the value of absorber thickness determined from Fig.  4-4 and the observed
value of V".  As an example, if V" = 0.02, the absorber thickness of NO is
approximately 0.014 atm cm.  This corresponds to a partial pressure of
0.014/21 = 0.67 x 10   atm, or 0.9 torr of NO..


     Because the absorption by a given absorber thickness of  NO  depends on
the widths of the absorption  lines, it depends on the collision  frequency
o£ the absorbing molecules and thus upon the total pressure of  a mixture
of NO with other gases such as N02, N2, or 0- that do not absorb in the
spectral region where the NO is measured.  The NO + N« mixtures  used to
obtain the data in Fig. 4-4 were adjusted to a total pressure of 120 torr,
the same as that in the test bulb.  Differences in the broadening abilities
oC N» and 0_ + N0« are not accounted for; however, they are expected to
produce only slight errors when using Fig. 4-4 in the described  manner to
determine the concentration of NO in the test bulb.
                                    4-7

-------
                   0.1
               V"
                   0.01
03
                   0.001
                                              0.001                  0.01

                                         ABSORBER THICKNESS  OF NO  (atm  cm)
0.1
                   FIG. 4-4.  Response of actinometer to different absorber thicknesses  of
                              NO in a mixture of NO + ^ at a total pressure of 120 torr.
                              By comparing the value of V" observed with the actinometer
                              in normal use to this curve, it is possible to determine
                              the NO concentration in the test bulb.

-------
                               SECTION 5
             ELECTRICAL DIAGRAMS AND DETAILS OF COMPONENTS
ELECTRICAL AND MECHANICAL

     Figures 5-1 through 5-8 show schematic diagrams of the  electrical
components.  These figures should be adequate for any trouble shooting,
modifying or replacing of parts.   The phase-lock amplifier has not  been
modified and is used in its normal manner.   An input plug with a 1  K ohm
resistor connected across the input leads has been added to  the recorder.

     Suppliers of the electrical and mechanical components and their
specifications are listed below.   Manufacturers' manuals are provided
separately for the recorder and amplifier.

     Recorder:                 Hewlett-Packard Model 7123A with
                               remote controls for the pen-lift and
                               chart drive.  Chart speeds 3.75 and
                               15 cm/min.

     Amplifier:                Princeton Applied Research Model 122.

     Power Supply:             Zeltec Model Z15AT100 DP. For the preamp
                               and reference circuits.

     Power Supply:             Powertec Model 2B5-3.  For thermo-electric
                               cooler for the detector.

     Source Transformer:       Triad F-3X,  2.5 V ac @ 10 amp.

     Timer Assembly:           Minarik Electric Company, multi-cam timer
                               6C-120S, recyling.
                                   5-1

-------
      Chopper Motor:

      Shutter Drive Motor:


      Clutch:
Globe Industries, 75A-121-1.

Hurst Model GA, 5 rpm, synchronous,
stock #171-01-043.

On shutter drive.  PIC #R3-3-50.
OPTICAL COMPONENTS

     Listed below are the specifications and the suppliers of the
optical components:
     Source:
     Detector:
     Test Bulb:
     Lens:
     Mirrors Ml  and  M3
Perkin-Elmer Part No. 457-0244.  Operated
in this instrument at approximately
2.5 V ac and 10 amps.

InSb vith 2 mm x 2 mm sensitive element
cooled to approximately -25°C by a
                               thermoelectric cooler.
                               Part Wo. OTC-3M1.
                        Optoelectronics
Quartz body with 1" diam Ca?2 windows.
Windows and the line to the valve are
bonded to the bulb with RTV, a silicone
cement.  Additional strength to the
bonds is provided by epoxy cement.
Because of a chemical interaction between
epoxy cement and NO™» the silicone is
used to separate the epoxy from the NOn
in the test bulb.  A Monel valve connected
to the test bulb makes it possible to
refill the test bulb.  The quartz body
was provided by Quartz General, 12440 Exline
Street, El Monte, Ca., 91732, who origi-
nally made it as a round bottom 500 ml
flask.  Cat. #U 106 Q.

CaF,,  1.37 cm diam, 1.39 cm effective
focal  length at 5.3 n,m.  Unique Optical,
P.O. Box 585, Farmingdale, New York.

5.0 cm diam, 10 cm focal length, spherical.
Ealing Corporation, 2225 Massachusetts Ave.,
Cambr id ge, Mas s., 021400.
                                    5-2

-------
M6:
Mirror M7:
Mirror M9:
Mirrors M2, M4,
M5, M8:

Mirror Chopper:
Fixed Attenuator:
5.0 cm diam, 10 cm focal length,
spherical.  Baling Corporation.

Cut from a spherical mirror of 7.5 cm
focal length provided by Baling Corporation.

Cut to approximately 2.6 cm diam from a
7.5 cm focal length mirror provided by
Baling Corporation.
Flat mirrors cut to the desired dimensions

Machined from stainless steel, chemically
coated with aluminum and polished to pro-
vide a smooth finish.  Contains 6 blades
and revolves at approximately 3500 rpm
producing a chopping frequency of 350 Hz.

Sapphire window, 2.54 cm diam by 1 mm
thick.
                              5-3

-------
                                                                                  _s    t-    US(L
                                                                                  • A/     , g/«je.  ofe*j    F«)o
fig. 5-1.  Simplified  schematic diagram for ac power.  The shutter drive motor and  its  limit
           switches  are  located on the optics assembly; all other switches shown are  on the
           control assembly.  The power supplies,  the  transformer for the source and  the
           chopper are on the optics assembly.  The  cycle-timer motor and the cams  driven by
           it are in the control assembly.

-------
     Optics
                                PAR
          CAbJc
     i.y  j I  $vf~f>L-f  [
Fig. 5-3.  Wiring diagram of electrical components on   the optics assembly.
           The  115 V ac inputs for the power suppliers  are shown; the output
           lines, not  shown, go  to the corresponding components.

                                       •5-5

-------
                                                                -rs-/
Fig. 5-4.  Wiring diagram of  timer and control assembly. One
           revolution of Che cycle timer motor corresponds to one
           period of the shutter.  Two interchangeable motors are
           provided; one runs at 1 rpm and the other at 0.5 rpm
           giving shutter cycles of 1 and 2 minutes, respectively.
           Cams 2 and 3 can be adjusted to change the period of
           time that the recorder chart switch is in the automatic
           position.  Similarly, cams A and 5 control the pen lift.
           When the corresponding switches on the control assembly
           are in the manual position, the recorder chart drive and
           the pen lift are controlled by the switches on the recorder.
                                       5-6

-------
                                                             + ISV
Fig. 5-5.  Schematic diagram of preamplifier used with the InSb detector.
                          loo-n.
                          5/OJT.
                                  -    L
_T^L3§L
<
/dK-t- <
«
<
J
r
v_
?
>
>
>
 Fig. 5-6.  Schematic diagram of the 350 Hz reference-pickup.




                                       -5-7

-------
Fig. 5-7.  Power supply for thermo-electric cooler  for  the detector.
        •^rr1 m^d

        <)•*£!
        —	 DC ."jtutrf
             I  s^^pjy
                           tcfe
Fig. 5-8.  Power supply for preamp and reference circuits,
                                  5-8

-------
                                  SECTION 6
                                 REFERENCES
1.  Hanst, P. L., Environmental Protection Agency (private communication).

2.  Leighton, Philip A., Photochemistry of Air Pollution, Academic Press,
    New York (1961).

3.  Giauque, W. F. and Kemp, J.D.  "The Entropies of Nitrogen Tetroxide and
    Nitrogen Dioxide.  The Heat Capacity from 15 °K to the Boiling Point.
    The Heat of Vaporization and Vapor Pressure.  The Equilibria N20,  =
    2N02 = 2ND + 02."   J. Chem. Phys. 6, 40 (1938).

4.  Burch, D. E. and Pembrook, J. D.  "Instrument to Monitor CH^, CO,  and
    C02 in Auto Exhaust."  Report No. EPA-650/2-73-030.  Prepared by Philco-
    Ford Corp., Aeronutronic Division for the Environmental Protection Agency.
    October 1973.

5.  Hanst, P. L.  "Spectroscopic Methods for Air Pollution Measurements,"
    in Advances in Environmental Science and Technology.  (New York:  John Wiley
    and Sons, 1971).

6.  GuLtman, A.  "Absolute Infrared Intensity Measurements on Nitrogen Dioxide
    and Dinitrogen Tetroxide."  J. Quant. Spectrosc. Radiat. Transfer 2, 1 (1962)
                                      6-1

-------
                                   TECHNICAL REPORT DATA
                            (Please read luamcnons on the reierse before completing)
 I RfcPORT NO

 l-.PA-650/4-74-036
                                                           3 RECIPIENTS ACCESSION NO
4 TITLE AND SUBTITLE
  N02  Actinometer for Field  Use
             5 REPORT DATE
                August  1974
                                                           6 PERFORMING ORGANIZATION CODE
7 AUTHORIS)

 Darrell  E. Burch, Ross C.  Bean and Francis J. Gates
             8 PERFORMING ORGANIZATION REPORT NO


                U-6074
9 PERFORMING ORG '\NIZATION NAME ANO ADDRESS
  Philco-Ford, Aeronutronic  Division, Ford Road
  Newport Beach, California   92663
                                                           1O PROGRAM ELEMENT NO
                1A1003
                                                           11 CONTRACT/GRANT NO
                                                              68-02-0798
 11 SPONSORING AGENCY NAME ANO ADDRESS
 National  Environmental Research  Center
 Research  Triangle Park,  N.  C.    27711
                                                           13 TYPE OF REPORT ANO PERIOD COVLREU
             14 SPONSORING AGENCY CODE
 15 SUPPLEMENTARY NOTES
 16 ABSTRACT
    Solar radiant energy  in  the  ultraviolet and short-wavelength visible dissociate
 NOp  in the atmosphere to produce  NO and 0?.  This photolytic reaction plays  an
 important role in the formation of photochemical smog, and information about the
 amount of actinic energy available in the lower atmosphere is required for the
 development of mathematical  models of the atmospheric processes.  This report
 describes the development  and  testing of an actinometer designed to measure  the
 actinic energy available for the  photolytic dissociation of N0?.  <\ spherical  test
 bulb contains a mixture of NCu and 0~ when it has been in the aark for several
 minutes.   When the bulb is exposed to solar energy NO is formed; its concentration
 is monitored by gas-cell correlation methods involving the infrared absorntion  by
 .iO.   A shutter periodically shades the test bulb from the sun for a one-minute
 period each two minutes.   During  the shaded period, part of the NO recombines with
 0? to form NO,.  The cyclic change in the NO concentration is related to  the
 actinic irradlance.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
   Actinometer
   Nitric Oxide
   liitrogcn Dioxide
                                              I) IDENTIFIERS/OPEN ENDED TERMS  '-  COSATI I U'lci'(.nill|i
 3 DISTRIBUTION STATEMENT

  Unlimited
19 SECURITY CLASS f Till': Hcpnrtl

 Unclassified	
21 NO OF PAGES

    5S
                                              20 SECURITY
                                               Unclassified
                                                                        22 PRICE
EPA Form 2220 1 (9 73)
                                             6-2

-------