Note: Descriptions are shown in the official language in which they were submitted.
1049808
This invention relates to methods and apparatus for deter-
mining the photochemîcal reactivity of organic pollutants in
gaseous maxtures such as air, mobile engine exhausts, vapors
from organic solvents and the like.
Hydrocarbons and similar organic pollutants which enter
into photochemical reactions in the presence of light and air,
yielding ozone and oxidants commonly referred to as photochemical
smog, are one of the major current sources of air pollution.
Thus, in order to mDnitor and/or control air pollution, instru-
mentation that could monitor at spheric air and sources of
these pollutants, particularly motor vehicle exhausts, and give
an indication of the total photochemical reactivity of the
organic pollutants in the mixture being monitored would be
desirable.
One of the m~jor problems in determining the photochemical
reactivity of these mixtures is that the ability or propensity
of organic pollutants to produce smogs varies greatly depending
on the rate coefficients of the many reactions involved. For
example, methane can be considered as unreactive and several
fold higher n-butane than ethylene concentrations are required
to produce similar atmospheric effects. Thus, in order to
determine the overall photochemical reactivity of a mixture
of orga~iC pollutants, i.e., the relative tendency of these
pollutants to participate in oxidant or ozone formation in the
presence of light and air`, the concentration of each pollutant
present should be multiplied by an individual or group reactivity
factor.
A system for classifying organics according to their
photochemical reactivity has been proposed by Basil Dimitriades
,
,.
.
, , : . .
1049808
in a paper entitled "The Concept of Reactivity and Its Possible
Applications in Control", published in the P m ceedings of the
Solvent Reactivity Conference, EPA Report 650/3-74-010, pp. 13-
22, National En~ironmental Research Center, U.S. Environ~ental
Protection Agency, Pesearch Triangle Park, N.C. 27711, (~ove~ber
1974). In this system, organic pollutants which are considered
photochemically non-reactive, such as methane, ethane, propane,
acetylene and benzene, are grouped together in one class and
given a reactivity rating o~ 1Ø Other organics are divided
into four classe~ which are given reactivity ratings, determined
by averaging previously measured reactivities for members of
these classes, of 3.5 to 14.3.
Barbara Krieger, Mazin Malki and Ralph Kummler have suggested,
in their article "Chemiluminescent Reactions of Oxygen Atoms
With Reactive Hydrocarbon-q, I. 7000-9000A", Environmental
Science & Technology, Vol. 6, pp. 742-744 (August 1972~, that
photochemically reactive hydrocarbons can be monitored by their
chemiluminescent reactions with oxygen atoms, either in the
7000-9000 angstrom range investigated by the authors or in the
vicinity of 3000 angstroms. The authors indicate (page 743)
that conditions might be found at which the light intensity
wouId be proportional to the product of the rate constant for
O atom attack timeS the hydrocarbon concentration. However,
the authors do not-identify any such conditions nor do they
consider reactivity ratings.
One problem with utilizing the techni~ues suggested by
Krieger et al is that acetylenej which is generally considered
photochemically non-reactive,- reacts with oxygen atoms and emits
radiation at most, if not all, of the wavelengths at which
~hotochemically reactive hydrocarbons emit. The intensity oe
,
.
,
, . - - ~ , . .~.
1049808
the radiation from acetylene is similar to the intensity of
radiation from some of the reactive hydrocarbons. Thus, it would
be desirable to have a method for determining the photochemical
xeactivity of organic pollutants that either does not respond
to acetylene or compensates for its presence.
It is an object of this invention to provide methods and
apparatus for determining the overall photochemical
reactivit~ of organic pollu~ants in a gaseous mixture via
chemiluminescent reactions between the pollutants and oxygen
atoms.
Another object is to provide methods and apparatus for
determining the overall photochemical reactivity of organic
pollutants in a gaseous mixture that conpensate for the
presence of acetylene.
According to the invention, the gaseous mixture or sample
to be analyzed is mixed with oxygen atoms, which react with
both photochemically reactive organic pollutants and acetylene
in the sample to produce chemiluminescence. The intensit~ - -
of radiation emitted at two separate wavelengths in the
OH(A ~- X2~r) system is monitored, and first and second signals
that are representative of the intensity of radiation emitted
at the fir.st and second wavelengths are produced. The second
. . ..
signal is subtracted from the first to eliminate the effects
of acetyle~e and produce a third signal that is representative
of the overall photochemical reactivity of the or~anic pollutants
in the sample.
The preferred apparatus for conducting these analyses
- includes a tubular reactor, with radiation sensors such as photo-
multiplier tubes positioned beside the reactor at substantially
the same position along the axis of the reactor and defining
,
1049808
an observation zone within the reactor. The sample and the oxygen
atoms are mixed upstream from the obser~ation zone and flow
through the observation zone to an exhaust conduit. The sa~ple
and/or oxygen atoms are introduced into the reactor at a point
from which the sample and the oxygen atoms take about 10 3 to
o-l seconds to flow to the center of the observation zone.
Preferably either the sample or the oxygen atoms are introduced
through an inlet oonauit that is adjustably mounted so that
the distance between the openings in the conduit through which
the sample or oxygen ato~s enter the reactor and the center
of the observation zone can be adjusted. This varies the time
- it takes the sample and oxygen atoms to flow from this introduction
point to the center of the observation zone. Ps will be seen
below, adjusting the introduction point in this manner, allows
one to vary the correlation betwee~ the photochemical reactivity
of the hydrocarbons in the mixture and the resulting signal
in any desired direction.
Other objects and aavantages of this invention will be
apparent from the following detailed description.
Figures 1 and 2 are graphs of the intensity and spectral
distr~bution of OH~A ~ - X21T ) radiation emitted in the atomic
oxygen ethylene and atomic oxygen acetylene reactions. ~ -
Figure 3 is a schematic diagram o~ one embodiment of this -
invention.
Figure~ 1 and 2 illustrate spectral emissions from the
atomic oxygen/ethylene and atomic oxygen/acetylene reactions
in the area of 300 nanometers, or 3000 angstroms~ These spectra
and the mechanisms by which they are produced are described -
by K. H. Becker, D. Kley and R. J. Norstrom in "OH* Chemilumin-
escence in Hydrocarbon Atom Flamesn, 12th Symoposium (International)
,,
' ' :
: :. ~: . .
1049808
on Combustion, The Combustion Institute, Pittsburgh, Pennsylvania
(1969). In order to illustrate the peaks in these curves more
clearly, they have been drawn as if both compounds produced
roughly the same intensity of radiation. Actually, the intensity
of the ra2iation from the atomic oxygen/ethylene reaction, under
typical conditions, is 4 to 6 times as great as the intensity
of radiation produced by the same concentration of acetylene.
As may be seen from these curves, both ethylene and acetylene
emit substantial radiation at 308.9 nanometers. Also, the
emission from the atomic oxygen/acetylene reaction at 306.4
or 312.2 nanometers is not very different from the intensity
of radiation from this reaction at 308.9 nanometers, while
the emission from the a~omic oxygen/ethylene reaction at the
latter wavelengths is significantly less than the intensity
from this reaction at 308.9 nanometers.
If a mixture of acetylene and ethylene is mixed with a
sufficient quantity of oxygen atoms to react with both hydro-
carbons, the intensity of the emitted radiation at any given
wavelength will be the sum of the intensity of the radiation
produced by th~ oxygen atom~ethylene reaction at that wavelength
and the intonsity of radiation at that wavelength from the
oxygen atom/acetylene reaction. Since the intensity of radiation
due to acetylene at 306.4 or 312.2 nanometers is substantially
the sa~e as the intensity of radiation from the acetylene
reaction at 308.9 nanometers, subtracting the total intensity
of radiation at one of the latter wavelengths from the total
intensity of radiation at 308.9 nanometers produces a signal
or measurement largely attributable to the ethylene in the
mixture. The effects of acetylene can be eliminated co~pletely
, . , , :
1049808
from the final signal by electronically adjusting the signals
produced at the first and second wavelengths by acetylene so
that the effect of acetylene is the same at both wavelengths.
~hus, the concentration of ethylene in a mixture containing
both ethylene and acetylene can be determined by measuring
the difference in the intensity of radiation at a first wave- -
length, such as 308.9 nanometers, at which both ethylene and
acetylene produce significant intensity peaks, and a second
wavelength, such as 306.4 or 312.2 nanometers, a~t which acetylene
produces substantially the same intensity of radiation but ~ -
at which the ethylene emission is significantly reduced.
I have discovered that the spectra emitted by typical
photochemically reactive hydrocarbons that are commonly present
in the atmosphere and/or in motor vehicle engine exhausts are
generally similar to the ethylene spectrum illustrated in Figure
1, and that the intensity of radia*ion emitted by these - ;
hydrocarbons can be correlated reasonably well with the reactivity
constants proposed for these hydrocarbons by Dimitriades
in the report referred to above, i.e., the intensity of radiation
producea by the same concentration of different hydrocarbons
can be correlated to the reactivity ratings for the groups in
which ~hey are classified by Dimitriades. As will be seen
below, the relative response from different hydrocarbons can
be varied by changing certain operating oonditions. Thus, the
methods and instruments of this invention may alsQ be used with
other reactivity classifications~
The-intensity o~ radiation produced by individual hydro-
carbons is proportional to the first power of the hydrocarbon
concentration, within a factor of about 2, over the range of
co~centrations that are likely to be encountered in the atomsphere
. ., '
'.
,' " . ; '
.'
. ' ' ' ' . ' ' ~
104980~
motor vehicle engine exhausts or the like. The intensities of
radiation emitted by individual hydrocarbons are additive, i.e.,
the total intensity produced by a mixture of hydrocarbons is
substantially equal to the sum of the intensities of the
radiation emitted by the individual hydrocarbons. Since the
same emissions are obtained from saturated and olefinic hydro-
carbons, it may be expected that similar results will also
be obtained with other compounds having alkyl groups, i.e.,
with m~t organic vapors. Thus, by measuring the intensity
of radiation emitted by organic pollutants in gaseous mixtures
such as air, motor vehicle exhaust or organic solvent vapors,
I am able to produce a signal that is related to the sums of
the percentages of the individuai organic pollutants present
times their reactivity constants, or, in other words, to the
overall photochemical reactivity of the mixture.
Any two wavelengths where the acetylene peaks are reasonably
close and where there is a substantial difference in the intensity
produced by photochemically reactive organics can be used in
' ` the methods and apparatus of this invention. As may be seen
fr~m the spectra illu trated in Figures 1 and 2, the acetylene -
peak at 306.4 nanoD-oters is closer than the 312.2 peak to the
intensity at 308.9. However, the difference in the ethylene
i intensity is increased by utilizing 312.2 as the second wave-
length. In order to obtain the maximum sensi~ivity, I believe
$t is generally preerable to utilize the peaks at 308.9 and
312.2. However, other combinations of wavelengths where the
difference in the intensity of radiation from photochemically
reactive organics is substantia~ly greater than the difference
in the intensity of radiation from the atomic o~ygen/acetylene
reaction may also be utilized.
. . ~ '; :
11)49808
Pigure 3 illustrates one form of apparatus that may be
utilized to perform these analyses. This system includes a
tubular reactor 10 within which the chemiluminescent reaction
is conducted. Tw~ photomultiplier tubes 11, 12, positioned
beside the reactor at substantially the same location along
the axis of the reactor, define an observation zone 13 withi~
the reactor. Preferably, photomultiplier tubss 11, 12 are
matched so that they produce approximately the same response
to a similar intensity of radiation at any given wa~elength.
Interference filters 15; 16, centered respectively at 308.9
and 312.2 nanoneters, are positioned between the photomultiplier
tubes and the observation zoneO Thus, photomultiplier tube
11 monitors radiation at 308.9 nanometers and photomultiplier
tube 12 monitors radiation at 312.2 nanometers.
I prefer to utilize interference filters having a half-
width of approximately one nanometer, i e., filters that only
transmit one-half as much radiation having a wavelength 0.5
nanometers above or below their center or rated wavelength
as they transmit at their center wavelengths. Transmission at
center is typically 20%, hence, 0.5 nanometers away it is
", . . .
typically about 10%. Of course, the selection of the filters
will be influenced by the wavelengths being monitored, the
spectral emissions of hydrocarbons in the vicinity of these
wavelengths and the impurities in the sam2le~ -
The frequency or wavelength of radiatlon transmitted byinterference filters 15, 16 varies with the incident angle
of the radiation upon the surface of the filter. Thus, colli-
- mators 17, 18 are positioned between the interference filters
and the reactor 10 in order to insure that radiation strikes
the filters nearly at right angles. Short lengths of metallic
,'-'
9 :
~ C~49801~
honeycomb have proved to be effective collimators in this
apparatus.
Lenses 19, 20 may be positioned between the collimators
17, 18 and the reactor 10. These lenses increase the signal
produced by the photo~ultiplier tubes by a factor of about
3.
In order to shield the observation zone 13 and the photo-
multiplier tubes from extraneous radiation, the filters 15,
16, collLmators 17, 18 and lenses 19, 20 are enclosed in suitable
housings 21, 22. Preferably, the rem~inder o~ the reactor
is constructed of or covered by an opa~ue material.
The photomultiplier tubes 11, 12 are connected to suitable
electronic means, such as a differential electrometer 27, which
- produces a signal proportional to the difference between the
signals produced by the photomultiplier tubes. The signal
from the electrometer is supplied to suitable readout means,
such as a gauge or recorder 28, which indicate the overall
: '
t / photochemical reactivity of the organic pollutants in the sample
being supplied to the reactor.
.
The sample to be analyzed is supplied through a sample
inlet conduit 31, which cbntains a valve 32 that controls the
..
rate at which the sample is supplied. In order to promote -
good m~xing between the sample and oxygen ato~s in the reactor,
the end of conduit 31 is provided with a plurality of openings
33, through which the sa~ple is admitted to the reactor.
Sample inlet conduit 31 is mounted in a vacuum feed-through
34 in~the end of reactor 10 so that the conduit may be recipro-
cated back and forth along the axis of the reactor in order
to vary the distance between the openings 33 through which
the sample is admitted to the- reactor ana the center of the
. ,~;.,- -
, . 1
, , :
- . - .
.. . . . .
, . . . ' : ' ' ' : , . : .: . ' : ' ' ' ' '
1049808
observation zone 13, and thereby adjust the correlation between
the photochemical- reactivity of the individual organics being
analyzed and the signal produced by the instrument. Moving
the openings 33 away from the center of the observation zone
increases the proportion of highly reactive organics that react
before they reach the observation zone and increases the signal
produced by less reactive organics in relation to the signal
produced by the same concentration of more reactive ones. Con-
versely, if the end of the sample inlet conduit is moved closer
to the center of the observation zone 13, the more reactive
organics are weighed more heavily in the signal produced by
the instrument.
Generally, the sample inlet conduit 31 may be positioned
so that the sample and oxygen atoms take about 10-3 to 10-
seconds to flow from the openings 33 in the end of conduit
31 to the center of the observation zone. Under typical
conditions this can be achieved by positioning the end of
the inlet conduit about 1 to 10 centimeters upstream from the
center of the observation zone. With typical hydrocarbon
mixtures, I prefer to position the end of inlet conduit 31
about 5 centimetcrs upstream from the center of the observation
zone, so that the sample and oxygen atoms will take about
2x10-2 second~ to flow fr~m the openings to the center of
the observation zone.
The weight given to various organics in the signal produced
by the instrument can also be adjusted by Varying the concentra-
tions of oxygen atoms within the reactor. Decreasing the
oxygen ato~ concentration increases the relative response o~
the more reactive organics because they are consumed less rapidly.
The oxygen atoms are supplied to the reactor by passing
an oxygen containing gas through an oxygen inlet conduit 36
1049808
which, like the sample inlet conduit 31, contains a ~alve
37 that controls the flow to the reactor 10. The o~gen inlet
conduit 36 passes through a microwave discharge cavity 38,
or other means of producing oxygen atoms, which converts part
of the molecular oxygen in conduit 36 to atomic oxygen. Bends
- 39 are provided in conduit 36 between the nicrowave discharge
cavity 38 and the reactor 10 in order to prevent radiation
from the microwave discharge from reaching the photomultiplier
tubes.
Of course, the oxygen could be supplied through a centrally
located inlet conduit like conduit 31, instead of the Qample,
or both the sample and oxygen could be supplied through conduits
of this sort. However, the illustrated arrange~ent is believed
to be preferable since it provides a relatively unobstructed
flow path for the oxygen atoms, which minimizes recombination
of these atoms before they reach the observation zone. I
prefer to supply a mixture of oxygen ana an inert gas such
as helium or argon to the oxygen inlet conduit 36 in order
to reduce the concentration of molecular o~ygen, which can
quench the chemiluminescent reactions and~or scavenge reactive
intermediates, within the reactor. A mixture of 9%oxygen
and 91% helium has been found to increase the signal from
the in~trument by a factor of about 5, at 1.2 Torr, over the
signal obtained by supplying pure mDlecular o*ygen to conduit
36. Of course, under so~e circumstances it might be desirable
to utilize undiluted oxygen for simplicity~ Undiluted oxygen --
would also increase the atomic oxygen concentration in the
reactor.
The oxygen containing gas supplied through conduit 36
and the sample supplied through conduit 31 flow through the
observation zone 13 and through an exhaust valve 42, which
12
' : .
1049808
serves to regulate the pressure in the reactor and to seal the
system when it is not in use, and through an exhaust conduit
43 to a vacuum pump 44, which maintains the desired flow rate
through and pressure within the reactor. The pressure in the
reactor may be read via a vacuum gauge 45 or mar.ometer connected
by conduit 46 to the reactor.
In order to avoid spreading the reaction out along the
axis of the reactor, which reduces the percentage of emitted
radiation which strikes the photomultiplier tubes and thus
reduces the signal~ I pxefer to supply ~he sample-at a relatively
low rate, e.g., about 0.5 cc tSTP) per second, and to supply
the oxygen containlng gas at about 1 cc (STP) per second.
Passing a 9 percent oxygen/91 percent helium mixture through
a typLcal microwave discharge cavity at this flow rates produces
an oxygen atom concentration of about 1 percent within the
reactor. As was mentioned above, this concentration may be
varied, e.g., by varying the microwave discharge power, to
achieve the desired correlation between photochemical reactivity
~nd the signal produced.
Under these conditions, this instrument will detect concen-
trat~ns of photochemically re w tive hydrocarbons as low as
0.2-1 ppm, depending on the reactivity of the compound. The
response of the instrument has been found to be proportional
to the first power of the hydrocarbon concentration, within
a factor of about 2, from the limit of sensitivity to at least
4,000 ppm. The relati~e response, i.e., the ratios between
the signals produced by the same concentrations of different
hydrocarbons, also remains substantially constant over this
range. As was mentioned above, the total intensity of the
radiation emitted by a mixture of hydrocarbons, and the signal
produced, are substantially equal to the sum of the intensities
13
, ; . . . .
-- ~
1049808
or signals that would be proauced by the same concentrations
of the individual hydrocarbons by themselves.
- Operation
The first step in the operation of this instrument is
to start the flow of gas through the microwave discharge and
strike the discharge. Zero air can advantaoeously be used in
the zeroing operation, which consists of cancelling out any
difference in the intensity of the radiation from acetylene
at the twv wavelengths being monitored and/or variations in
the signals produced by the pho~omultiplier tubes. This is
accomplished by supplying a stream of air or other inert carrier
gas and zeroing both photomultiplier tube signals, then intro-
ducing acetylene and balancing the instrument, e.g., by adjusting
the gain on one of the phtomultiplier tubes, to produce a zero
signal on the recorder. The instrument i8 then calibrated
for photochemically reactive organics by introducing a carrier
gas containing one or more of these organics anZ adjusting
the instrument to produce a signal on the recorder that corresponds
to the known concentration of photochemically reactive organics
in the calibration sample. Since the response of this instrument
is linear, in most cases this instrument can be calibrated
with one data point. However, in some ~ases it may be de~irable
to utilize calibration samples containing different concentrations
of the photochemically reactive organics in order to test
the response of the instrument over a range of concentrations.
Fol~owing calibration~ the instrument can be utilized
to determine the photochemical reactivit~ of many different
orgànics in gaseous mixtures such as atmospheric air, motor `~
vehicle exhausts, ~apors from organic solvents or the like~
The following examples illustrate some of the results obtainable
by the use of this invention.
.'
1~ .
,
.
1049808
Example 1
Reactions between various individual hydrocarbons and
oxygen atoms were conducted in a sy~tem similar to the
system illustrated in Figure 3. The reactor was a Pyrex (T~)
tube 60 centLmeters long and 22 millimeters in internal diameter.
The samples, which consisted of mixtures of scientific grade
air and of the hydrocarbon being investigated, were supplied
thrDugh a centrally located inlet tube at a flow rate of 0.5
cc (STP) per second. The discharge end of the inlet tube was
located 4 centimeters upstream from the center of the observation
zone. The oxygen atoms were supplied by passing a 9 percent
oxygen/91 percent helium mixture through a 2450 MH3 microwave
discharge in a 13 mm. O.D. ~ycor (TM) at a flow rate of 1 cc
(STP) per second. The oxygen ato~ concentration in the reactor
was estimated to be about 5X1014 cc 1. The pressure inside
the reactor was 1.2 Torr.
The reaction was monitored by two matched Centronic Model
4242 photomultiplier tubes mounted as illustrated in Figure
3. One nanometer half-width Corion light filters centered
at 308.9 and 312.2 nanometers, and 2.5 centimeter long honeycomb
collimators with passages 3 mm wide, were placed between
the photomultiplier tubes and the reactor. The photomultiplier
tube~ were connected~to a differential electrometer, which
in turn was connected to a meter and a chart recorder. The
instrument was zeroed and calibrated~ as described above,
using samples of air containing 1 to 1250 ppm of acetylene
and 0.1 to 1250 ppm of ethylene. After calibration with
ethylene, samples containing ~arious o~her hydrocarbons were
supplied to the reactor at concentrations from 0.1 to 1250 ppm.
The relative responses-produced by 1250 ppm of these hydro-
carbons are set forth in Table I. The relative responses at
....... . ... ~ ... ... .....
1049808
other concentrations, under the same operating conditions,
are substantially the sa~e as those in the table.
TABLE 1
Class V. Reactivitya = 14.3Relative Response
Ethylene 100
Propylene 42
Butene-l 79
Butene-2 53
Isobutene 38
Propadiene 133
Butadiene 56
Class IV, Reactivit~ = 9.7
Toluene 55
Class III, Reactivity = 6.5
n-Butane 23
n-Heptane 57
Iso-octane 43
. Class I ,- Reactivity_~ 1.0
Ethane 1.3
-- Propane 7-9
Benzene 20
- Acetyle~e 0
Methane~ dO 2
- a) Reactivity classeæ and numbers as suggested by B. DLmitriades -
~b) No detectable signals from methane were obtained at ooncentrations
up to 1250 ppm, the highest ooncentration investigated.
- .
Egample 2
Using the same-system and the same general reaction
conditions as in Example 1, tests w OE e conducted to see how
the relative signals produced by ethylene and n-butane varied
- ~ .
' with the inlet nozzle spacing. Mixtures of cientific grade --
,
air and 1250 ppm of either ethylene or n-butene were supplied
- through the centrally located inlet tube at 0.5 cc (STP)/sec.
The ~ariation in the signals from the individual hydrocarbons
- and the ratio between the ethylene signal and the n-butanè
- signal with ~he variation in the distance from the end o~ the
inlet nozzle to the center o~ the observation zone are given
in the following tabie.
, ., , , - . . . .: : :. ~: :
1049~3~8
TABLE 2
-
Nozzle Ethylene n-Butane
Spacing Signal Signal g
cm amps x 10-9 amps x 10- P~atio
2 1850 180 10
3 1100 160 6.9
4 1030 270 3.8
700 320 2.2
EXAMPLE 3
In this test, made under the sa~e general conditions as
Example 1, both the nozzle spacing and the concentration of
oxygen atoms inside the reactor were varied. The oxygen atom
ratio was varied by a factor of 2 by varying the power to the
microwave cavity. The variation in the signals produced by
125 ppm of ethylene and 125 ppm of n-butane and the ratio between
these signals with nozzle spacing (x) and relative oxygen con-
centration t~ is shown in the following Table.
- TABLE 3
.
Ethylene n-Butane
;- X . lO] signal signal
cm M~ MA - Ratio
6 2 23 100 0.23
6 1 34 11 3.1
2 2 20~ 10 20
2 1 135 5 27
As may be seen from the~e examples, the nozzle to obser-
vation zone distance and~or the atomic oxygen concentration
. . ~ .
in the reactor may be ~aried to suit particular mixtures of
organics or paxticular times during a photochemical smog episode.
O course, variou~ other modifications in the methods and
apparatus described above will be readily apparent to those
skilled in the art. For example, a sphericaI reactor or a
short cylindrical reactor with photomultiplier tubes positioned
at the end of the reactor could be used instead of the illustrated
tubular reactor. This invention may also be practiced with one
.,
, , , . . ............................ ...... , .. ...... ,.. .,_ ,__ _
, ,
.. . . .
: I
~ 1049808
photomultiplier tube that alternately views the reaction
zone at each of the two wavelengths, e.g., by using two filters
in a filter wheel or by using one interference filter and varying
the angle of the light passing through it. These and various
other modifications may be made within the scope of this inven-
tion, which is defined by the following claims.
~' . .
" ,,.
I
~: . ' ' ' .':
~.
rl
~ ' ' ' ~ ' .
.
''' ' - ' ~
~i '. .' ~.'~,
f ~
: -: ' . .. .....
1
:
, :., , ' , '
~ :
~, .
.
.,
. ' - .
18
-.,
;
, ~: . . : . ' . ' .
