United States
Environmental Protection
Agency
Environmental Sciences Research
Laboratory
Research Triangle Park NC 27711
EPA 600 2-79-191
Ociober 1979
ch and Development
Technique for In Situ
Calibration of
Particulate Mass
Monitors
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-79-191
October 1979
TECHNIQUE FOR IN SITU
CALIBRATION OP PARTICULATE MASS MONITORS
Walter John, Susanne Hering and Jerome J. Wesolowski
Air and Industrial Hygiene Laboratory Section
California Department of Health Services
Berkeley, California
Grant Number R 805577010
Project Officer
John Nader
Emissions Measurements and Characterization Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection
Agency, nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
-------
ABSTRACT
Two types of aerosol generators, the Riker Laboratories metered spray
can and the Mistogen ENlU5 ultrasonic nebulizer, were evaluated by laboratory
measurements for application to the in situ calibration of particulate mass
monitors for stationary sources. The metered spray can delivers a fixed
amount of aerosol each time the valve is depressed. The average mass of
propellant and solution in each squirt is 52 mg, and is reproducible within
H$. The total volume of the dried particles per squirt is of the order of
10~5cm3. The volume median diameter was varied from l.U to 3.2 ym by
selection of solute concentration. Because of its simplicity and repro-
ducibility of output, the metered spray may be useful for a variety of
applications requiring a portable aerosol source. For calibrating stack
beta gauges, larger aerosol output of 5-10 mg is needed, requiring a valve
with a metering volume at least ten times larger than present valves.
Contact electrification monitors require a test aerosol of 25 mg/m3 at a
flow of 1 m3/min. Appropriate for this application is the ultrasonic
nebulizer, which has an output of 50 mg/min, constant to within 8$ over a
period of hours.
This report was submitted in fulfillment of Grant No. R805577010 by the
California Department of Health Services, under the sponsorship of the
U.S. Environmental Protection Agency.
111
-------
CONTENTS
Abstract iii
Figures vi
Tables viii
Acknowledgments ix
1. Introduction 1
2. Conclusions and Recommendations 2
3. Requirements of an Aerosol Source for Calibrating
Particulate Mass Monitors U
Beta gauge mass monitors h
Contact electrification mass monitors 8
U. Evaluation of Metered Sprays Cans 10
Preliminary work; commercial metered sprays 10
Prepared metered sprays 16
Discussion 21
5. Evaluation of the Mistogen Ultrasonic Nebulizer 32
Experimental methods 32
Measurements and results 3^
Discussion 36
References Ul
Appendix
A. Data for Individual Squirts of Metered Aerosol Cans ^3
-------
FIGURES
Number
Page
1 Metered Valves from Riker Laboratories and Precision
Valve Company 12
2 Sampling System for Preliminary Testing of Commercial
Metered Sprays 13
3 Time Dependence of the Particle Count from Isuprel
Commercial Metered Spray 1^
k Sampling System to Measure Size Distributions of Prepared
Metered Sprays 17
5 Propellent and Solution Mass Output of 0.1$ Benzoic Acid
Spray as the Can is Emptied 22
6 Particle Volume of 0.1$ Benzoic Acid Spray as the Can is
Emptied 22
1 Particle Number and Volume Distributions for Prepared Metered
Spray Containing 0.1$ Benzoic Acid 23
8 Particle Number and Volume Distributions for 0.7$ Benzoic
Acid Metered Spray 2U
9 Particle Number and Volume Distributions for 3$ Benzoic Acid
Metered Spray 25
10 Cumulative Volume Distributions for Benzoic Acid Metered
nf
Sprays ^°
vi
-------
Number Page
11 Particle Number and Volume Distributions for Isuprel
Commercial Metered Spray 27
12 Particle Number and Volume Distributions for Bronitin
Commercial Metered Spray 28
13 Cumulative Volume Distributions for Isuprel and Bronitin
Sprays 29
1^- Average Volume Per Particle and Mass Median Diameter for
Benzoic Acid Metered Sprays at Three Solute Concentrations,
and for Isuprel Commercial Metered Spray 30
15 Sampling System to Monitor Output and to Measure Size
Distributions from the Ultrasonic Nebulizer 33
16 Sampling System to Measure Aerosol Mass Output from
Ultrasonic Nebulizer 35
IT Particle Number and Volume Distributions from Ultrasonic
Nebulizer 37
18 Cumulative Volume Distribution from Ultrasonic Nebulizer 38
19 Ultrasonic Nebulizer Volume Distribution at End of 3.5 Hour
Run. Measured 1.8 Hours After the Distribution Shown in
Figure 17b 38
20 Nebulizer Solution Consumption Rate 39
vii
-------
TABLES
Number Page
1 Commercial Beta Gauges for Stack Monitoring 6
2 Requirements of Calibration Aerosol Source for Beta Gauge
Monitor 7
3 Commercial Contact Electrification Mass Monitors 8
k Requirements of Calibration Aerosol Source for Contact
Electrification Monitor 9
5 Number of Particles Per Squirt from Commercial Sprays 15
6 Number, Volume Per Squirt for Primatene and Bronitin
Commercial Metered Sprays 15
7 Primatene Particle Number, and Volume Per Squirt 16
8 Metered Spray Can Output Per Squirt 19
9 Ratio of the Aerosol Volume from the Optical Counter to
That Derived from Mass Loss 20
10 Mistogen Ultrasonic Nebulizer Output 36
-------
ACKNOWLEDGMENTS
We thank John Nader, project officer, for his continuing interest and
for suggesting the investigation of spray cans. ¥e appreciate the assistance
of Walter Lim, Aerosol Services Inc., who packaged spray cans for us, and
Marty Rosenbaum and Irving Porush, Riker Laboratories, who provided infor-
mation and spray can supplies. Ve thank L. Robert Cameto, Mistogen
Equipment Co., for the donation of an ultrasonic nebulizer. Discussions
with Brian Mokler and with Vokker Mohnen concerning their experience with
metered spray cans were very helpful.
ix
-------
SECTION 1
INTRODUCTION
The objective of this project was to evaluate portable aerosol
generators suitable for in situ calibration of stationary source particle
mass monitors. There are two factors involved in the validity of stack
measurements, (l) the degree to which the sample is representative of the
stack effluent, and (2) the accuracy of the instrument response. With a
portable aerosol generator of known output, the second question of instrument
response could be answered in the field under typical operating conditions.
This would be useful both for testing the monitor performance, and as part of
the calibration of the monitor against the EPA Reference Method. The EPA
Method 5 involves the collection of a stack sample with a filter and a
series of impingers, which are later analyzed gravimetrically. The lack of
immediate information from this method is a handicap in the calibrations of
a continuous mass monitor. It would be desirable to know prior to the
calibration whether the mass monitor is functioning properly, both to save
time arid to increase the reliability of the calibration.
In this project aerosol generators were evaluated in terms of the
calibration requirements for two types of continuous mass monitors, the
beta absorption gauge, and the contact electrification monitor. For the
beta gauge the particulate matter is accumulated for a few minutes on a
filter followed by a mass determination via the absorption of beta particles
from a radioactive source. The contact electrification monitor senses the
electrical charge transfer taking place when particles impact on a probe;
this charge is proportional to the mass. The calibration requirements for
these two monitors are outlined in Section 3. Aerosol generators appropriate
to each were evaluated, as described in Sections k and S. A third common
type of monitor, the optical transmissometer, measures the light attenuation
produced by particulate matter. This type of monitor presents a different
calibration problem than the beta absorption or the contact electrification
monitor and was not included in the scope of the present project.
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
In this study, two types of aerosol generators, the metered spray can
and the ultrasonic nebulizer were investigated for use in the field as a
test aerosol source.
The metered spray can produces a fixed amount of aerosol each time
the valve is depressed. To "be useful for instrument calibration, the
generated aerosol must he reproducible. The size distribution and total
mass delivered should be appropriate to the instrument being tested. For
metered sprays using a 50 mg valve from Riker Laboratories, the total dried
particle volume is of the order of 10~5 cm3. The volume median diameter
varies from l.U to 3-2 ym for solute concentrations of 0.1 to 3$. The
particle volume per squirt for each of the cans tested has a coefficient
of variation between 6% and 29$. However, the mass of propellant and
solution in each squirt is reproducible within k%. Some of the greater
variability in the volume is attributed to measurement error. At higher
solute concentrations, the measured dried particle volume is significantly
less than expected. The discrepancy is attributed to higher losses for
the larger particles. Variation in these losses would also contribute to
the variation in particle volume.
The best results were obtained with low solution concentration. For
a spray composed of 0.1$ benzoic acid in a 30$ ethanol, 70$ Freon-12
solution, the aerosol produced with each squirt has a volume of 1.7 x 10~5
cm3 +_10$ and a mass median diameter of 1.40 +_0.05 ym. Both the repro-
ducibility and the size distribution of this aerosol are appropriate for
testing stack beta gauges, but a greater mass per squirt is needed. Because
of aerosol losses, increasing the solute concentration is not an effective
means of obtaining a larger mass output. To obtain sufficient mass for the
stack beta gauge a valve with a metering volume at least ten times larger
is required; this appears to be feasible.
For applications which require of the order of 10 ug of aerosol, spray
cans using the present metered valves would be suitable. The can weight
loss measurements show the valve output is reproducible within k%. Further
development would be needed to avoid particle losses. Improvements in
aerosol volume consistency will not come from improvement in the valve, but
probably from reduction of particle losses. Smaller particle sizes would
be helpful in avoiding losses. For some applications, a submicron aerosol
is needed. These smaller particles could be produced by using a more
dilute aerosol solution.
-------
The Mistogen ultrasonic nebulizer produces a large, continuous amount
of aerosol. For a nebulizing solution of 0.033 g/ml of potassium bi-
phthalate in water, an output of 50 mg/min -with a mass median diameter of
l.U pm was measured. Aerosol volume and size distributions are constant,
within experimental error, over a period or several hours. This generator
is appropriate for testing the contact electrification monitor.
This study has shown that the metered spray can and the ultrasonic
nebulizer have sufficiently reproducible output to be used as test aerosol
generators. The present metered spray cans are appropriate for applications
requiring about 10 yg of particles. A valve with a larger metering volume
is needed to supply the 5-10 mg of aerosol required for the stack beta
gauges. The ultrasonic nebulizer, with a large continuous output, could
be used for the contact electrification monitor. The use of these aerosol
sources in the field will require provision of portable dilution systems.
This will be relatively easy to accomplish for the spray can but somewhat
more difficult for the ultrasonic nebulizer because of tLe large volume
required. The metered spray can, because of its simplicity and consistent
output, could find a variety of other uses as a portable particle generator.
-------
SECTION 3
REQUIREMENTS OF AS AEROSOL SOURCE FOR CALIBRATING MASS PARTICULATE MONITORS
This study is concerned with the in situ testing of particulate mass
monitors for stationary sources. A portable aerosol generator is to tie
used to calibrate the sensor. The. question of representative sampling is
not addressed. The test aerosol need not have the properties of the source
particles. Also, it is not intended that the sampling efficiency of the
monitor should "be measured. Basically, a portable aerosol generator is
required to deliver a test aerosol of a known concentration or total mass
to the monitor.
The aerosol size distribution and mass loading must be appropriate for
the monitor to be tested. To formulate the requirements for the aerosol
generator, a survey vas made of commercially available beta gauge and contact
electrification mass monitors.
BETA GAUGE MASS MONITORS
Beta gauges provide continuous or semi-continuous monitoring of air-
borne particulate mass loadings (l-5) and for stack sampling, beta gauges
have been extensively used in Europe. Lear Siegler advertises that'their
instrument has been used in over 90 European installations. To date, use
in this country is limited. However, in a survey of measurement methods for
particulate emissions from combustion sources, Gilmore Sem, et al. (6)
rated the beta gauge as having the greatest immediate potential for continuous
stack particulate mass monitoring.
Principle of Operation
For beta particles of energies less than 1 MeV, the principal inter-
action with matter is inelastic scattering with the orbital electrons. The
attenuation of the beta radiation passing through a material is a measure
of the material's areal electron density. As the ratio of atomic number
to atomic mass is Q.k to 0.5 for most elements, the beta attenuation is a
direct indication of the sample mass.
Empirically, the attenuation is given by the relation:
I » I e-KO
o
where I and I are the measured intensities of the g radiation with and
without the sample; and a is the mass per unit area of the sample. The
-------
calibration constant K depends primarily upon the energy of the g-source.
Most samplers use 1I+C, which has a maximum 6 emission energy of 0.156 MeV,
and a 5700 year half life. For this isotope, the constant K is about 0.26
cm2/mg.
To measure the airborne particulate mass, particles are deposited on a
surface; the intensities of the beta radiation which penetrates the
collection substrate before and after aerosol collection determine the mass
per unit area of the deposited particles. Typically, samples are collected
with a filter, but some beta gauges use an electrostatic precipitator (7)
or an impactor stage. (8) It is not feasible to use beta gauging in situ,
e.g., to observe attenuation directly across a stack, because the attenuation
will be dominated by that of the gas phase mass, which is of the order of
103g/m3, far exceeding particulate loadings.
Commercial Beta Gauge Stack Monitors
In the United States, continuous stack mass monitors using the beta
gauging principle were manufactured by Research Appliance Corporation (R.A.C.)
and Lear Siegler (L.S.). Currently, only R.A.C. is actively marketing the
device.
Table 1 lists the operational and design parameters of these two stack
mass monitors. Both devices collect the aerosol onto a filter tape. Lear
Siegler uses a glass fiber filter, whereas^R.A.C. uses the Whatman No. h paper
filter. Both companies use a carbon-1^ beta source and a Geiger Mueller tube
detector. The intensity of the beta radiation penetrating the filter is
measured on the same spot both before and after the sample collection. This
method eliminates error due to variation in filter thickness along the tape.
The filter mass is of the order of 5 to 10 mg/cm2, whereas the collected
particle mass is in the range of 0.2 to 5 mg/cm2.
-------
TABLE 1. COMMERCIAL BETA GAUGES FOR STACK MONITORING
Research Appliance Co.
Lear Siegler
Beta Source
Detector
Type of Filter
Sample Spot Area
Sampling Rate (including
dilution air)
Sample Time
Count Time
Average Count Rate
Range of Filter Loadings
Operating Conditions for
Beta Gauge
< 100 yCi 14C
Geiger Mueller Tube
Cellulose (Whatman #U)
5 cm2
< 0.1 m3/min
(< k cm)
Variable: limits no
specified
60 sec
50,000 min"1
> 2 mg
Ambient Temp. Low RH
(Dew Pt = -40°C)
(purge with dry air)
< 100 uCi lkC
Geiger Mueller. Tube
Glass Fiber
9 cm2
0.06-0.2 m3/min
(2-6 CFM)
Variable
0.5 to 50 min.
50 sec
20,000 min'1
2 to 50 mg
Ambient Temp. & RH
Sample Line Heated,
1TO°C
For both of these commercial instruments, the sample flow rates and
probe diameters are variable, chosen to obtain approximately isokinetic
sampling. Sample times are adjusted to obtain a mass deposit within the
working range of the instrument. There is a dead time, i.e., no sampling,
while the filter tape is being counted. Differences between the two
instruments in the sample collection, dilution, cooling and flow measurement
are described below.
Research Appliance Company Sampling System
The R.A.C. employs a stationary probe, positioned in the stack at a
point where the effluent velocity equals the average stack velocity, as
determined by a pitot tube survey. Upon the initial setup, the sampler is
calibrated against EPA Method 5 to determine that the R.A.C. monitor is
correctly positioned to obtain a representative sample.
-------
Condensation is avoided by using a sample line which can be heated to
120°C. A boundary layer diluter introduces dry (dew pt = -Uo°C) air along
the walls of the sample tube to reduce the relative humidity and temperature
and to lower wall losses. After the particles have been filtered, the
diluted sample gas stream passes through a condenser to remove water vapor,
and then the dry air volume is monitored. The input dilution air is also
measured, the difference of the two measurements used to determined the
actual sample volume. The company specifies there are no losses of particles
below 7 ym diameter. Their beta gauge is designed to be operated downstream
of a bag house or electrostatic precipitator, to monitor performance of these
devices. The company states that at this point in the stack most of the
larger particles have been removed.
Lear Siegler Sampling System
Lear Siegler uses a swiveling probe, designed to sweep out equal areas
in equal time in the stack. The flow rate is held constant, sampling
velocity is set either equal to, or 10% greater than the average stack
effluent velocity. To avoid condensation in the sample line, heating
collars are used to maintain a constant temperature as high as 170°C. An
optional attachment permits the addition of dilution air to the sample
stream.. The sample air flow is measured by monitoring the differential
pressure across a venturi nozzle downstream of the filter. Ideal gas laws
are used to correct the flow to standard conditions. Unlike the R.A.C.
instrument, this sample volume will include volume due to water vapor.
The requirements for a calibration aerosol for beta gauges are listed
in Table 2.
TABLE 2. REQUIREMENTS OF CALIBRATION AEROSOL SOURCE FOR THE BETA GAUGE MONITOR
Type of aerosol solid
Particle size < 7 pm
Total mass delivered 5-10 mg
Time dependence No restrictions
Flow rate 0.06 m3/min
-------
CONTACT ELECTRIFICATION MASS MONITORS
In monitoring instruments based on contact electrification, a flow of
the aerosol is directed at a probe. The transfer of charge from particle-
probe collisions results in a current which is continuously monitored with a
sensitive meter. The electrical charge correlates accurately with the mass
determined gravimetrically for a given source material. The physical
principles and the operating characteristics of the monitor have been reviewed
by John. (9,10)
Two instruments are currently available commercially, the IKOR.Continuous
Particulate Monitor, IKOR Division, OMNI-WAVE Electronics Corporation,
Gloucester, Massachusetts, and the KONY TEST, Kony Co., Ltd., Tokyo, Japan.
The current number of units in the field is not known, but is probably less
than 100. The characteristics of the two instruments are listed in Table 3.
The concentration range for the IKOR can vary somewhat depending on the probe
configuration, the probe material and the material sampled. Both instruments
are equipped with an in-line filter which affords a gravimetric sample for
calibration purposes.
TABLE 3. COMMERCIAL CONTACT ELECTRIFICATION MASS MONITORS
IKOR KONY TEST
Type of sampling Extractive or in situ Extractive
Sensor configuration Streamlined or turbulent Tube
Sensor material Metal Semiconductor
Sampling rate 0.5-0.9 m3/min 0.15-1.0 m3/min
(17-30 CFM) (5.3-35 CFM)
Concentration range 0.2 mg/m3-230 g/m3 0.1 mg/m3-8 g/m3
The calibration of the contact electrification monitors must take into
account a special problem, namely that the instrument requires equilibration
for a period of from several minutes up to 15 or 20 minutes. During this
time the sensitivity varies, approaching an asymptotic value as the surface
of the probe is conditioned in the aerosol stream. The calibration procedure
would consist of exposing the probe to the calibration aerosol until the
current signal reaches its steady state. Then the current would be integrated
over a convenient period of time, say five minutes, to obtain the charge to
be related to the mass of sampled aerosol. If the test aerosol were aluminum
oxide, for example, approximately 25 mg/m3 will give a reasonable current
signal. Calibration of the contact electrification monitor requires a
non-sticky aerosol. The particle size distribution should have a substantial
fraction above 0.1 via. diameter since the particles are sensed by impaction
on the probe.
-------
The desirable characteristics of a calibration aerosol source are listed
in Table U.
TABLE k. REQUIREMENTS OF CALIBRATION AEROSOL SOURCE FOR THE CONTACT
ELECTRIFICATION MONITOR
Type of aerosol Non-sticky solid
Particle size > 0.1 ym diameter
Concentration > 25 mg/m3
Time dependence Constant within 10% for 30 minutes
Flow rate 1 m3/min
-------
SECTION U
EVALUATION OF METERED SPRAY CANS
Metered spray cans were investigated as a means of producing a test
aerosol for the beta gauge monitors. In reporting this work we use the
term aerosol strictly to refer to a suspension of liquid or solid particles
in a gaseous medium, and not to the spray can itself.
Instead of a continuous spray, the metered cans deliver a fixed quantity
of aerosol each, time the valve is pressed. Metered sprays are used commer-
cially for dispensing "bronchial sprays to relieve symptoms of asthma. The
questions of interest in terms of their applicability as a test aerosol for
particulate monitors are: (l) the reproducibility of the mass per spray,
(2) the particle size distribution, and (3) the total mass obtainable with
each spray. To test the beta gauges used for stack monitoring 5 to 10 mg
of an aerosol less than 7 vm particle diameter is required.
PRELIMINARY WORK; COMMERCIAL METERED SPRAYS
Three commercially available metered bronchial sprays, Isuprel (.Win-
throp Laboratories), Primatene ('Whitehall Laboratories) and Bronitin
(Whitehall Laboratories), have been investigated. These are each equipped
with a 50 mg metered valve, U.S. Patent No. 2968^27, manufactured by Riker
Laboratories (Northridge, CA). The Bronitin differs from the Isuprel and
Primatene in that the concentrate is a suspension rather than a solution.
Each of the cans uses a Freon propellant.
A diagram of the metered valve is shown in Figure la. The small
reservoir between the stem and the surrounding cylinder is filled through a
groove in the bottom of the stem when the valve is closed. Depressing the
valve stem first seals the small reservoir from the bottle, and then opens
it to the atmosphere through a 0.8 mm orifice at the top of the valve stem.
There is a second 0.3 mm orifice in the actuator used to depress the stem
(not shown). The two rubber gaskets which seal against the valve stem
determine the upper and lower boundaries of the reservoir volume.
A similar metered valve which was inspected but not tested is shown in
Figure Ib. This valve is manufactured in England by Metal Box Ltd.,
(Reading, England) and marketed in this country by Precision Valve Co.,
(Yonkers, New York). It is very similar in design to the Riker valve.
The principal differences are (l) the major components of the valve are
plastic, rather than metal, (2) communication with the bottle when the
valve is closed is by means of grooves in the side of the reservoir
volume.
10
-------
'Experimental Methods for Preliminary Testing of Commercial Spray Cans
As shown in Figure 2 the aerosol vas introduced into a U cm diameter
glass pipe. The spray can is held in an inverted position, to insure
proper functioning of the metered valve. A transit time down the pipe of
approximately U seconds allows the propellant to evaporate. Aerosol is
sampled with the Climet 208 optical particle counter, using the Climet
dilution system to avoid coincidence losses. The dilution system operates
by introducing a particle-free sheath air flow at the sampling inlet, equal
to 99% of the total sampling rate of 7 1pm. The maximum count rate of the
optical counter was 101* per sec.
Particle size distributions were obtained from the Climet output pulses
using a Nuclear Data Pulse Shape Amplifier, and Tracor Northern Multi-
channel analyzer. The data were recorded on paper tape and reduced by
computer. The optical particle analyzer was calibrated using polystyrene
latex spheres (Dow Diagnostics) and monodisperse glycercl particles
generated with a Berglund-Liu vibrating orifice. Two amplifier gain
settings were used, covering the particle size ranges from O.UO to 2.0 vim,
and 1.7 to 6.1 um. A total number count was also obtained from an Ortec
single channel analyzer and a counter/timer.
Time Dependence
The time dependence of the particle count following a squirt of Isuprel
was recorded on the multichannel analyzer operated in the multiscaler mode.
An example is shown in Figure 3. The analyzer count rate has been adjusted
for sample dilution. The aerosol number reaches a maximum of approximately
6 x 106/s in about one second, falling rapidly thereafter.
Reproducibility of Particle Number and Volume
Data were taken on Isuprel, Primatene and Bronitin sprays. The number
of particles from each squirt were counted with the single channel analyzer.
Because of the necessity of using two different amplifier gain settings,
only part of the particle size distribution, either 0.^ to 2.0 vim, or 1.7
to 6.1 ym could be measured at one time. Therefore the data taken in these
two size regions were averaged over ten squirts. Statistics were obtained
from the variation between these sets of data.
In these experiments 100 seconds was allowed between successive squirts
of the can. Testing with the Isuprel mist showed that waiting longer, up to
20 minutes since the last squirt, did not significantly change the number of
particles generated. For measurements on the Bronitin mist, which is a
suspension, the can was shaken before each spray, as directed by the manu-
facturer .
The single channel analyzer data and knowledge of the flow rates allows
calculation of the total number of particles per squirt. The results are
listed in Table 5. Of greatest interest here is the reproducibility of the
volume or mass, of aerosol generated. Table 6 shows the total aerosol number
and volume for the Primatene and Bronitin mists, based on the pulse height
11
-------
GROOVE
n r
i i
i i
i .
l_':
^ 2.78
«— 1.67
Una OD
mm ID
RUBBER
SPRIHG
RUBBER
3HOOVE
PLASTIC TUBE
3.2 mm OD
2.It mm ID
STEM
3.2 am OD
r
8.2
(a)
(b)
Figure 1. Conmercial metered vaJ.ves: (a) from Biker Laboratories; (b) from Precision Valve
Company.
-------
AEROSOL CM
OPTICAL PARTICLE
COUNTER
SINGLE
CHANNEL
MALYZER
COUNTER/TIMER
DILUTION
AIR
RETURN
LINEAR
AMPLIFIER .
\
/
MULTICHANNEL
ANALYZER
\
/
PAPER
TAPE
PUNCH
CHART
RECORDER
PRINTER
Figure 2. Sampling system for preliminary testing of commercial metered spray cans.
-------
106 _
o
o
W
CO
PH
I
O
M
EH
1* x 10 _
2 x 106 _
TIME (SECONDS)
Figure 3- Time dependence of the particle count
from Isuprel commercial metered spray.
-------
analyzer data. The data are shown in tvo size ranges corresponding to the
two amplifier gain settings used to span the entire particle size range.
Each set of data is an average of 10 squirts of the aerosol can. The
standard deviations are given for the data sets. The number data in
Table 7 based on the multichannel analyzer counts are slightly smaller than
the data in Table 6 which were derived from the single channel analyzer
counts. The latter covered a slightly wider particle size range.
TABLE 5- NUMBER OF PARTICLES-PER SQUIRT FROM COMMERCIAL SPRAYS
Isuprel
Primatene
Can #1
Can #2
Can #3
All cans
Bronitin
No. of
.Squirts . .
67
70
20
52
Ih2
60
No. of
Data. Sets*
5
7
2
5
lU
6
Average
.Individual
U.U3
5.UU
6.63
1*. 37
5-13
8.97
x 10 7
x 107
x 107
x 107
x 107
x 107
of
Squirts
+ 35$
± 33$
+ 30$
+ 30$
+ 37$
±36
Average of Data
Set Averages
U
5
U
5
8
.26-
.53
.35
.15
.97
x 107
x 107
—
x 107
x 107
x 107
+ 26$
± 17$
+ 8.9$
+_23$
+ 23$
*Each data set consists of 8-20 individual squirts.
TABLE 6. PARTICLE MJMBER AND VOLUME PER SQUIRT FOR PRIMATENE AND
BRONITIN COMMERCIAL METERED SPRAYS
Primatene
No. of
Data Sets
N
O.kO - 2.1 ym
2.1 - 6.1 pm
Total
7
7
U.Ul x 107 +
9.51 x 105 +_ 23%
it. 51 x 107 +
1.11 x 107 +. 22$
7.66 x 106 +_ 28$
1.88 x 107 H- 17$
Bronitin
O.UO - 2.1 ym
2.1 - 6.1 ym
Total
3 7-51 x 107 +_ 21$ 2.U9 x 107 + 11$
3 2.09 x 106 +_ 8$ 1.98 x 107 +_ 7$
7.72 x 107 + 20$ U.U7 x 107 + 7$
15
-------
TABLE 7. PRIMATEHE PARTIGI^'NUMBER AND VOLUME PER squiRT
Amount in Can '...'"' N V (ym3)
Full
Half full
Quarter full
Within 5$ of empty
U.5 x 107 +_ 17$ .
U.U7 x 107
l*.0l* x 107
2.0 x 107
1.88 x 107 + 17$
1.27 x 107
1.13 x 107
No data
The Primatene shows an overall 17$ standard deviation in the number
and in the volume between the data sets. The Bronitin mist, -while showing
a large deviation (20$) in the number of particles per squirt, has a
smaller variation (7$) in the total aerosol volume. Thus the variation
in the particle number is not necessarily a good measure of the variation
in the particle volume. The data also show that the volume per squirt is
about twice as large for Bronitin as for Primatene.
One of the Primatene cans was used until it emptied, which required
about 300 squirts. The number of particles did not drop noticeably until
the bottle was less than half full, as can be seen in Table 7. On the
other hand the volume per squirt dropped by about 30$ when the bottle was
half full.
Discussion
The metered spray can shows promise as a portable aerosol calibration
source. The particle size distribution is appropriate. The coefficients
of variation for sets of 10 squirts were found to be 7$ and 17$ for Bronitin
and Primatene, respectively. The volume dropped about 30$ when the can was
half full. These variations were small enough to justify further exploration
of this approach.
PREPARED METERED SPRAYS
Because of the promise shown by commercial sprays, metered cans were
prepared with our own test solutions. The concentration of the solution
was varied to determine the effect on the total aerosol volume and the size
distribution. Solutions of benzoic acid in ethanol, with concentrations of
3 mg/ml, 20 mg/ml and 100 mg/ml were packaged 30$ w/w in Freon propellant.
Benzoic acid was chosen because it is highly soluble in ethanol and in
Freon; thus a homogeneous solution could be obtained. 50 mg solution valves
were obtained from Riker Laboratories; the solutions were packaged by Aerosol
Services Inc., (U25 S. Ninth Avenue, City of Industry, California 917^5)-
16
-------
92 LPM
CLEM DRY AIR
GLASS WOOL
J
7.8
CLEAN DRY AIR
fF
)U
OPTICAL PARTICLE
COUNTER(CLIMET 201)
1
OPTICAL PARTICLE
•COUNTER (CLIMET 208
LOG AMPLIFIER
MIXER /ROUTER
\
/
PULSE HEIGHT
ANALYZER
\
/
PAPER TAPE
PUNCH
Figure k. Sampling system to measure size distributions of prepared metered sprays.
-------
Experimental Methods for Prepared Sprays
The experimental set up is shown in Figure U. The addition of a Tracer
Northern mixer/router enabled the measurement of the complete size distri-
bution from 0-U to 7-8 ym diameter for a single squirt. Two optical
particle counters were used, one with a linear amplifier to cover the small
particle range, O.U to l.U ijm, and one counter with a logarithmic amplifier
to span the larger sizes. The linear amplifier was necessary for the small
particle sizes as the logarithmic amplifier response to the small pulses
was inadequate. The mixer/router sends the two counter signals to separate
halves of the pulse height analyzer. The optical particle counters were
calibrated with polystyrene latex spheres and dioctylphthalate particles
generated with the Berglund-Liu vibrating orifice.
The dilution system was changed to obtain a steadier sampling rate and
to minimize the transit time in the counter sample line. Aerosol is intro-
duced into a 9-5 cm diameter glass tube with a particle free air flow of
92 1pm. The flow rates of the Climets were adjusted to U 1pm each, through
a 0.95 cm tube to which 7-8 1pm dilution air is added concentrically. The
measured aerosol sampling rate was 0.211 +_ .003 1pm, corresponding to a
1*36:1 dilution ratio. The small sample rate was necessary to maintain the
multichannel analyzer dead time below 20$. The maximum count rate for each
of the squirts is of the order of 20,000 sec"1, which gives a optical
counter coincidence loss rate of 6%.
Me asurement s for Prepare d Sprays
Size distributions were measured for approximately ten squirts of each
of the prepared spray cans. For comparison, the Isuprel and Bronitin size
distributions were remeasured using the improved experimental configuration.
The cans were actuated using the Isuprel cap. A Mettler-microbalance was
used to weigh the cans between each squirt. The measurements on each of
the spray cans are summarized in Table 8. The average mass, number, volume
and mass median diameter are listed. Corresponding data for the individual
squirts are listed in the appendix. In Table 8, cans containing the benzole
acid solutions are labeled with the prefix "AS" (Aerosol Services); the
solution concentrations are listed as the weight percent of the aerosol
material in the Freon-alcohol solution. Three cans were packaged for each
of the different concentrations of benzoic acid. However two of the 0.7$
solutions lost most of their Freon so that no measurements could be made.
For the Bronitin spray the concentration was estimated from the manufacturers
stated dosage and our average measured change in the can weight.
Aerosol Volume
The total mass of propellant, solvent and aerosol for each squirt was
measured by weighing the can, without the actuator. For the benzoic acid
cans the average mass change was kk to 56 mg, which is consistent with the
manufacturers labelling as a "50 mg valve." For an individual can the mass
per squirt is quite consistent; those tested had coefficients of variation
of 1-8$, the average being h.0%. These numbers are indicative of the per-
formance of the valve itself, and do not include any losses in the actuator.
18
-------
TABLE 8. METERED SPRAY CAN OUTPUT PER SQUIRT
Can
A.S.
A.S.
A.S.
A.S.
A.S.
A.S.
A.S.
Total Propellant and Particle Dry Particle
Aerosol Solution Mass Number Volume
Solution (mg/squirt) (per squirt) (cm3/squirt)
#1
#2
#3
#5
#7
#8
#9
Isuprel
0.]
o.:
o.:
0.'
3$
3$
3$
L$ Benzoic Acid
L$ Benzoic Acid
L$ Benzoic Acid
r$ Benzoic Acid
Benzoic Acid
Benzoic Acid
Benzoic Acid
0.25$ Isuprel
55
53
51
kk
51
53
53
56
.9
.5
.2
.5
.U
.5
.7
.7
+_ 2.0$
+ u.u$
+ 1.7$
+ 5.W
±3.7$
+ 2.5$
+ 8.2$
+ 6.7$
3.
3.
2.
2.
1.
1.
2.
1.
37x10 7+_ 10$ 1.96xlO~5+_ 8.6$
51xl07+_ U.8$ 2.01xlO~5+_ 11$
oUx!07+_ 7-7$ 1 . 17xlO~5+_ 6 . 2$
1 x!07+_ 27$ 2.1 xlO~5+_ 29$
75x10 7+_ 8.7$ 2.8xlO*~5 + 19$
9Uxl07+_ 8.1$ U.2xlO~5 +_ 19$
07xl07+_ 5.5$ U.5lixlO~5+_ 11$
87x10 7+_ 9.3$ 1.58xlO~5+_ 13$
Volume Median
Diameter (pm)
1.
1.
1.
2.
3.
3.
3.
2.
U
k
U
2
1
1
2
0
+_ .06
+_ .05
+ .05
± -1
± -2
+ .2
± -1
± -°5
Bronitin
hydrochloride
^0.5$ epinephrine
bitartrate
62.9 + l.<
7.1 x!07+
5.3 xlO~5+
1.6 + 0
-------
The aerosol volume (dried particles.) is calculated from the optical
particle counter size distributions. As shown in Table 8, the volume is of
the order of 10~5 cm3 per squirt. Neglecting losses it shoiiLd also be
possible to calculate the volume from the solution concentration, aerosol
density, and weight loss of the can. However the aerosol volume derived
from the optical counter is considerably smaller than that derived from
the mass loss as shown in Table 9. The discrepancy is greater for the
more concentrated solutions. For 0.1$ benzoic acid the volume is almost
half of the calculated value whereas for the 3% solution it drops to about
3% of the calculated number. The total number of particles detected
decreases for the higher solution concentrations, also indicating more
aerosol losses for the larger particles.
TABLE 9.
Can Ho.
AS1
AS2
AS3
AS5
AST
AS8
AS9
Isuprel
Bronitin
RATIO OF THE AEROSOL VOLUME FROM
COUNTER TO THAT DERIVED FROM MASS
Solution Optical
Concentration Volume
0.1%
0.1%
0.1%
0.1%
3%
3%
3%
0.25%
^0,5%
THE OPTICAL
LOSS
Counter Volume
from Mass Loss
0.50
0.51
0.29
0.075
0.02 it
0.036
0.038
0.17
0.23
Of importance in the application of these spray cans as test aerosol
generators is the reproducibility of the aerosol output. As stated above,
the mass of propellant and solution lost from the can with each squirt is
52 mg with an average coefficient of variation for an individual can of U$.
The total dry particle volume, based on optical counter measurement, was
more variable. For the benzoic acid cans, the pooled coefficient of
variance in the measured volume is 9-8$ for the 0.1% solution, and 16% for
the 3% solutions. The can AS #5, with a 0.7$ solution, exhibited an
anomalously high coefficient of variation of 29$. Unfortunately the other
two cans prepared at this concentration were not functional. Some of the
increased variation can be attributed to the larger errors inherent in
the volume measurement, such as small fluctuations in air flows or sampling
rates. Another source of variation is deposition and resuspension of solute
in the actuator. Especially for the higher concentration cans, deposits
were observed visually at the orifice in the actuator. Significant aerosol
losses are also indicated by the discrepancy between the measured and
calculated dry particle volume. These losses can contribute to variations
in the aerosol volume which are not reflected in the can weight loss meas-
urements .
20
-------
For one of the cans, AS #1, measurements were made at successive stages
as its contents were exhausted. Figure 5 shows the total mass of propellant
and solution per squirt as a function of the number of squirts from the can.
The mass is quite constant until the can is nearly empty after 200 squirts.
The dry particle volume, shown in Figure 6 is not quite as steady, but shows
the same sharp decrease around 200 squirts.
Size Distributions
Typical number and volume distributions for individual squirts of the
benzoic acid solutions are shown in Figures 7» 8 and 9. The volume
distributions have peaks at 1.5, 2.7 and. h vim for benzole acid con-
centrations of 0.1, 0.7 and.3% respectively. The volume distributions
are very reproducible as seen by the small standard deviation in the volume
median diameter (Table 8). The cumulative volume distributions, Figure 10,
are approximately log normal for the two lower concentration solutions, but
not for the 3% solution. Shown in Figures 11 and 12 are the Isuprel and
Bronitin size distributions. The Isuprel is similar to the benzoic acid
sprays in that the active ingredient is dissolved in the alcohol and the
Freon propellant. The Bronitin is a suspension of solid particles with a
.surfactant in the Freon. Both of the commercial sprays have size distri-
butions which are approximately log normal (Figure 13), with volume median
diameters of 2.0 and 1.6 ym for the Isuprel and Bronitin respectively.
The volume distribution for the Bronitin spray is distinctly narrower.
For the benzoic acid and Isuprel sprays the particle volume is expected
to be linearly dependent on solution concentration, with a zero intercept.
In Figure 1.k the average volume per particle is plotted against solution
concentration. The slope of the line on this log-log graph is only about
0.3. Also shown is the increase in volume median diameter with solution
concentration. The reason for particle size not scaling with solution
concentration as expected is not known; however, it may be related to
particle loss, which is size-dependent.
In comparison with these sprays, the Bronitin has a larger aerosol
volume with a relatively small mass median diameter of 1.6 ym. Because it
is a suspension rather than a homogeneous solution, its size distribution is
expected to depend on l) particle size in the suspension and 2) the extent
of coagulation in the spray.
DISCUSSION
The metered spray can holds promise as an inexpensive, portable aerosol
calibration source. The output of the metered valve, as determined by can
weight loss, is reproducible within k%. The volume of aerosol produced by
the lower concentration solution of 0.1$ benzoic acid is reproducible within
10%. The performance of the more highly concentrated solutions is not as
good, probably due to losses of the larger particles.
The size distribution meets the criteria stated in Section 3 for the
beta gauges, namely that the aerosol be less than 7 ym. The aerosol mass
produced by the metered cans is 25 to 60 yg per squirt, depending on the
solution concentration. This is considerably less than the 5-10 mg needed
21
-------
60
6b°
O O Q>
o o
20
100
200
HUMBER OF SQUIRTS
Figure 5. Propellant and solution mass output of 0.1
benzoic acid spray as the can is emptied.
3 x 10~6
m
8
1
B 2 x 10~6
co
i
1
g 1x10-6
O
i
0
1 1 1 1 1
0 0
00
0 0
_ IL-TI-
o
o
1 1 1 1 1
100
200
HUMBER OF SQUIRfS
Figure 6. Particle volume of 0,
can is emptied.
benzoic acid spray as the
22
-------
3. 0
SOM»
*OM*
ro
u>
(—1
a.
a
u
a
1
•MM
a
a
>
2.5
2. 0
1 .5
1 . 0
. 5
n
10
PARTICLE DIAMETER. DP CMICRONS)
10"* 2 5 10U 2
PARTICLE DIAMETER, DP
5 101
CMICRONS)
(a)
(b)
Figure J. Particle size distributions for prepared metered spray containing 0.1% benzole
acid: (a) number distribution; (b) volume distribution. Data from the two
particle counters are -indicated by the symbols Q and V •
-------
3. 0
5S01S
ro
10
PARTICLE DIAMETER, DP (MICRONSD
a
o
o
3. 0
2. 5
2. 0
.1.5-
55015
1 . 0
10
PARTICLE DIAMETER, DP [MICRONS)
(a)
(b)
Figure 8. Particle size distributions for O.J$ benzoic acid metered spray: (a) number
distribution; (b) volume distribution.
-------
ro
o
o
o
2
a
3. 0
2.5-
2.0-
1.5-
5I5S7
3. 0
51587
1.0-
10"1 2 5 10U 2 5 101
PARTICLE DIAMETER, DP CMICRONS)
PARTICLE DIAMETER, DP CMICRONS)
(a)
(b)
Figure 9. Particle size distributions for 3% benzole acid metered spray: (a) number
distribution; (b) volume distribution.
-------
908X8
99.9
ro
99. •
95.
90.
a*
Q
vl 70.
o 5°'
I 30.
u
10.
5.
1.
0.1
T—I—I Mill
55015
51587
I III
0.2 0.5 1.0 2.0 5.0
PARTICLE DIAMETER, DP (MICEONS)
(a)
0.5 1.0 2.0 5.0
PARTICLE DIAMETER, DP (MICROHS)
0.5 1.0 2.0 5.0 10.
PARTICLE DIAMETER, DP (MICRONS)
(b)
(c)
Figure 10. Cumulative volume distributions for prepared metered sprays: (a) O.J
"benzoic acid solution; (b) 0.7$ solution; (c) 3% solution.
-------
ro
a
o
o
3. 0
2. 5
2. 0
47343
10
-1
a.
a
o
o
_J
a
3. 0
2.5
2.0-
1.5-
1 . 0
47343
PARTICLE DIAMETER, DP CMICRONSD
PARTICLE DIAMETER. DP CMICRONS)
(a)
(D)
Figure 11. Particle size distributions for Isuprel commercial metered spray: (a) number
distribution; (b) volume distribution.
-------
3. 5
108879
CO
0
10
1
2 5 10U 2 5 10
PARTICLE DIAMETER, DP (MICRONS)
a
o
o
>
o
3. 5
2. 9
2. 3
1 . 8
101679
> 1.2
0
1
10"1 2 5 10U 2 5 10
PARTICLE DIAMETER, DP (MICRONS)
(a)
(b)
Figure 12. Particle size distributions for Bronitin metered spray: (a) number distribution;
("b) volume distribution.
-------
108879
ro
99.9
99.0 .
95.
90.
70.
50.
30.
10.
5.
1.
0.1
0.1
I i
0.2 0.5 1.0 2.0 5.0
PARTICLE DIAMETER, DP (MICRONS)
10.
(a)
99.9
99.
95.
90.
TO.
50.
30.
10.
5-
1.
0.1
T 1 1—I I I I I
0.1 0.2 0.5 1.0 2.0 5-0 10.
PARTICLE DIAMETER, DP (MICRONS)
(b)
Figure 13. Cumulative volume distributions for commercial metered sprays: (a) Isuprel
spray; (b) Bronitin spray.
-------
TICLE
P
U.o
ro
o
0.5
• VOLUME PER PARTICLE
V VOLUME MEDIAN DIAMETER
i i : i i
0.1
1
0.2
0.5 1-0
SOLUTION CONCENTRATION
2.0
U.O
2.0
0.5
5.0
o
Figure lU. Average volume per particle and volume median diameter for benzoic acid
metered sprays at three solute concentrations, and for Isuprel commercial
metered spray. Solid symbols (•..T) are used for "benzoic acid sprays.
Open symbols (O,V) indicate Isuprel spray.
-------
for the stack beta gauge calibration. Increasing the solution concentration
does, not sufficiently increase the mass output. Furthermore the generator
performance becomes more erratic due to deposits in the actuator. Larger
aerosol mass per squirt requires a metered valve with a larger reservior.
To produce sufficient aerosol for the stack beta gauges, a reservoir at
least ten times larger is needed. The Bronitin spray, which is a suspension
of the solid aerosol particles in Freon, produces a greater aerosol mass
than the solution-type aerosols without increasing the particle size. Higher
concentrations of the suspension-type sprays were not explored.
The metered spray cans may find application for other monitors, such
as a stack transmissometer, or ambient aerosol instruments. In those cases
it may be desirable to have a submicron aerosol. The aerosol size can be
reduced by l) decreasing the solution concentration, 2) increasing the
percentage of Freon in the can to produce a finer mist, or by 3) using a
higher vapor pressure propellant. (ll)
In summary, the simplicity of the metered spray can and the repro-
ducibility of their output makes these cans attractive as a portable test
aerosol source for field use. Further development work is required to
increase the mass output for beta gauge calibration and to obtain smaller
particle size. Importantly, the most stringent requirement, reproduci-
bility of the output within +_ 10$, is met by the 0.1$ benzoic acid solution.
These cans may be quite useful where a large aerosol mass is not required.
31
-------
SECTION 5
EVALUATION OF THE MISTOGEN ULTRASONIC NEBULIZER
Ultrasonic nebulizers have a large output, and this may "be suitable
for testing mass monitors which are sensitive only to large aerosol loadings.
In particular, for the contact electrification monitor, the test aerosol
should have a loading greater than 25 mg/m3, at a flow rate of 1 m3/min.
A study by Mercer, et al (12), indicates that these loadings should be
attainable with the ultrasonic nebulizer. In the present study the
Mistogen ENli;5 (Mistogen, Oakland, Calif.) was tested. The parameters
measured are l) constancy of the output, 2) aerosol size distribution,
3) aerosol concentration.
The ultrasonic nebulizer contains a piezoelectric crystal driven at
high frequency. The crystal is located at the bottom of a cup filled with
nebulizing solution. The solution is maintained at a constant level in the
cup by a float valve. At sufficient vibrational intensities the liquid
forms a fountain in the center of the cup; cavitation in the fountain
creates the aerosol mist. (13) Periodic shock waves generated by the
implosion of cavitation voids excite standing capillary waves on the
surface of the liquid. (l*0 Aerosol is formed by the breakup of the crest
of the capillary waves, with a number mean size equal to approximately
0.3 to 0.3^ of capillary wavelength. (1^,15) The capillary wavelength is
given by the expression:
A =
where a is the surface tension, p the liquid density and f the vibrational
frequency. The Mistogen, operating at 1.5 MHz, is expected to give a
number mean size of 3 ym for water.
EXPERIMENTAL METHODS
The sampling system used for measuring aerosol size distributions and
the constancy of the output of the ultrasonic nebulizer is shown in Figure
15- Because of the very high aerosol concentration, a large dilution is
required before the optical particle counter can be used. This is
accomplished by five stages of dilution-. At port 1, 18 1pm of dry, filtered
air is supplied to the nebulizer; the nebulizer's blower was not used.
Dilution air added at port 2 draws a sample from the lower cone by a venturi
effect; a soap bubble flow meter was used to determine the necessary flow
to obtain a positive sampling. Air added at port 3 dries and dilutes the
aerosol without affecting the sampling rate. The venturi tee used again
at port h to sample from the upper cone; further dilution air is added at
ports 5 and 6.
32
-------
1».25 LPM
U)
u>
6.0 LPM
OPTICAL
PARTICLE
COUNTER
LPM
AMPLIFIER
raANNEL
MULTIC:
ANALYZER
6.0 LPM
18 LPM
ULTRASONIC NEBULIZER
Figure 15. Sampling system to monitor output and to measure size
distributions from the ultrasonic nebulizer.
-------
The aerosol sampling rate of 2 1pm into the optical particle counter is
sufficiently large compared to the overall counter sampling rate of 6 1pm
as to render it insensitive to slight variations in the counter sampling
rate. All flow rates are constantly monitored with rotameters. With this
dilution system the counting rate is 9000 sec l, corresponding to 2% coinci-
dence losses. The single channel analyzer and counter-timer were used to
monitor the aerosol concentration.
Particle size distributions were obtained from the Climet 208 optical
particle counter output pulses using a Nuclear Data pulse shape amplifier,
and Tracer Northern 1705 multichannel analyzer. Data from the multichannel
analyzer is punched on paper tape and reduced "by computer. The optical
particle analyzer was calibrated using polystyrene latex spheres (Dow
Diagnostics) and monodisperse glycerol particles generated with a Berglund-
Liu vibrating orifice. Two amplifier gain settings are used, covering the
particle size ranges from Cs.kQ to 2.0 ym and 1.7 to 6.1 ym.
To measure the total output of the ultrasonic nebulizer the arrangement
shown in Figure l6 was used. All of the aerosol flows through the cone from
which the filter sample is taken. At the sample point the humidity is low
enough to prevent condensation on either the walls of the cone or the
filter holder, although the particles are not dry. Filters were dessicated
prior to weighing. The final weight was determined by dessicating the
filters until a stable weight was attained.
MEASUREMENTS AND RESULTS
The Mistogen ultrasonic nebulizer output was monitored over a period
of 3-1/2 hours, using a nebulizing solution of potassium biphthalate,
0.033 g/ml in water. After drying, the potassium biphthalate produces
solid particles. Size distributions, taken midway through the run, are
shown in Figure 17- The volume distribution for the dried aerosol exhibits
a peak at 2.0 ym, -somewhat higher than the volume median diameter of l.U ym.
The distribution is not quite log normal, (Figure 18) , but may be approximated
by a Gaussian distribution with a standard deviation, a = l.U. For several
other ultrasonic nebulizers Postendorfer, et al (l6) recently measured volume
median diameters in the range of 2.9 to 4.5 urn.
No shift is seen in the size distributions measured at the middle and
end of the run, 1.8 hours apart. The final volume distribution, shown in
Figure 19 can be compared with the earlier distribution shown in Figure 17b.
The total volumes of aerosol sampled in these two measurements were
2.ill x 105 and 2.3^ x 105 pm3/!, a change of only 3%.
The constancy of the output was monitored with the single channel
analyzer and counter-timer. The number of particles per 5 minute interval
was recorded over the entire 3-1/2 hour sample time. The standard deviation
of these 5 min. counts is 7-8$. This low standard deviation was achieved
only after the dilution system was refined. It is suspected that variations
in air flow rates still contribute significantly to the standard deviation;
i.e., the nebulizer itself may be even more stable.
-------
Jf
SIERRA FLOW
CONTROLLER
135 LPM
33 LPM
30 LPM
ULTRASONIC
NEBULIZER
Figure 16. Sampling system to measure aerosol mass output from the
ultrasonic nebulizer.
35
-------
The particle mass output of the nebulizer was determined by filter
measurements with the set up shown in Figure 16. Both Gelman A glass fiber
and 0.8 ym Millipore filters were u&ed. The membrane filters are easier to
handle but have the disadvantage of loading too quickly. The Mistogen was
operated with a 30 1pm dispersion air flow (Figure 16, port l) which is
the amount normally provided by the Mistogen blower. The results are shown
in Table 10, the average output is 50 +_ 8 mg/min. The relative standard
deviation of 16% would be reduced to only 9% if the anomalously low 55 min.
run were excluded.
: TABLE 10. 'MISTOGEH-ULTRASONIC NEBULIZER OUTPUT .
Filter Type Sample Time Mistogen Output
Gelman A Glass Fiber
Millipore , 0.8 ym pore membrane
8.5 min
9.0 min
it-5 min
55 min
1.10 min
1.15 min
56 mg/min
59
50
35
1*8
51
Average 50+8 mg/min
The output of the Mistogen ultrasonic nebulizer depends on the dis-
persion air flow through the nebulizer cap. Figure 20 shows the nebulizer
solution consumption rate, for three different flows. As expected, the
value levels off at higher flows. For calibration work it would be im-
portant to operate in this plateau region.
The measured particle mass output of the nebulizer is only about 30$
of that estimated from the liquid consumption rate. This is not unreasonable
since the fountain action throws a substantial amount of unnebulized solution
on the walls.
DISCUSSION
The criteria listed in Section 3 for an in situ calibration source for
a contact electrification monitor specified a non-sticky solid aerosol,
greater than 0.1 ym particle diameter, with a concentration greater than
25 mg/m3 at a flow of 1 m3/min, and constant within 10$ for 30 minutes.
The Mistogen ultrasonic nebulizer can meet all of these criteria. It is
also portable and easy to operate. The mass output of 50 mg/min for a
0.033 g/ml potassium biphthalate nebulizer solution, if diluted in 1 m3/min
air, would meet the concentration requirement. The solution concentration
could be increased threefold before reaching saturation, giving as much
as 150 mg/min aerosol output. The aerosol has a reasonable size, with a
36
-------
1*12532 . 1056"H»
H12532 - 1056ft
3.0 -
2.0 •
1.0 •
2.0
1.0 -
0.1
0.2 0.5 1.0
0.1
0.2 0.5 1.0 2.0 5.0 10
PARTICLE DIAMETER, DP (MICRONS)
PARTICLE DIAMETER, DP (MICRONS)
(a)
(b)
Figure !?• Particle size distributions for the Mistogen ultrasonic nebulizer using a
nebulizing solution of 0.033 g/ml potassium biphthalate in water; (a) number
distributions (b) volume distribution.
-------
M2532 -
U)
CD
a
vl
99.9
99.
95.
90.
70.
50.
30.
10.
5.
0.1
32S376 - 110308
T 1—T~T
0.1 0.2 0.5 1.0 2.0
PARTICLE DIAMETER, DP (MICRONS)
5.0 10.
3.0
Q
a
2.0
1.0 ,
0.1 0.2
0.5 1.0 2.0 5.0 10
PARTICLE DIAMETER, DP (MICROHS)
Figure 18. Cumulative volume distribution
from, ultrasonic nebulizer.
Figure 19. Ultrasonic nebulizer volume
distribution at.end of 3.5
- hour run. Measured 1.8 hours
after, distribution shown in
Figure ITb.
-------
6.0
o
M
U)
VQ
8
o
§
U.o
I
. 2."0
I
I
10 20
NEBULIZER DISPERSION AIR FLOW (LPM)
30
Figure 2Q. Nebulizer solution consumption rate.
-------
volume median diameter of 1.1* ym. Potassium biphthalate makes a solid,
bouncy particle when dry. The 1 m3/min flow rate of dilution air is easily
attainable and enough to dry the aerosol. Importantly the output is very
steady, with less than Q% standard deviation in the number concentration
over a period of 3-5 hours. Unlike a fluidized bed, the time from startup
until a constant output is attained is only one or two minutes.
To use the ultrasonic nebulizer as a calibration source for the contact
electrification monitor, one needs a dilution system which is built into the
aerosol generator. With such large loadings wall losses can be significant,
and will depend on the dilution system employed. However, once it is
calibrated, one would have a portable aerosol test source of known output.
The main problem is to develop a dilution system which is not too unwieldly
for field use. We conclude that the ultrasonic nebulizer could be used as
the basis for a portable aerosol generator to calibrate the contact electri-
fication monitor.
-------
1. Izmailov, G. A. Measuring the Gravimetric Concentration of Dust in the
Air Using B-Radiation. Industrial Laboratory 2J_, *K)-U3, (January 196l).
2. Nader, J. S. and D. R. Allen, A Mass Loading and Radioactivity Analyzer
for Atmosphereic Particulates. Am. Ind. Hyg. Assoc. J. 21, 300-307
(I960). —
3. Husar, R. B. Atmospheric Particulate Mass Monitoring with a 6 Radiation
Detector. Atmos. Environ. 8_, 183-188 (197*0.
h. Dresia, H. and R. Mucha. Registering Radiometric Measuring Instrument
for Combined Measurement of the Emissions of Dust and Radioactivity in
Air. Staub Reinhalt Luft (English Transl) 3^. 103-106 (197*0.
5. Aurand, K. and J. Bosch. Recording Counter for Continuous Determination
of the Concentration of Pulverulent Air Pollutants. Staub Reinhalt Luft
(Engl) 2J_, No. 10, 21-21*. (1967).
6. Sem, G., J. Borgos, J. Olin, J. Pilney, B. Liu, N. Barsic, K. Whitby and
F. Dorman. State of the Art 1971- Instrumentation for Measurement of
Particulate Emissions from Combustion Sources (ThermoSystems, Inc.,
St. Paul, MN, 1971) PB-202-665, APTD 0733-
7. Horn, W. Process for Continuous Gravimetric Determination of the
Concentration of Dustlike Emissions. Staub Reinhalt Luft (Engl) 28,
No. 9, 20-25 (1968).
8. Lilienfeld, P. Beta Absorption Impactor Aerosol Mass Monitor. Am. Ind.
Hyg. Assoc. J., 31, 722-729 (1970).
9- John, W. Investigation of Particulate Matter Monitoring Using Contact
Electrification. Environmental Protection Agency, Research Triangle
Park, NC, Technology Series Report Number EPA-650/2-75-0^3, February
1975, ^5 PP.
10. John, W. Contact Electrification Applied to Particulate Matter Moni-
toring. In: Fine Particles, Aerosol Generation, Measurement, Sampling
and Analysis, B. Y. H. Liu, ed., Academic Press, NY, 1976, pp. 61*9-667.
11. Sanders, P. A. Principles of Aerosol Technology. Reinhold Book Corp.,
New York, NY, 1970, Chapter 8.
-------
12. Mercer, T. T., R. F- Goddard, R. L. Flores. Output Characteristics of
Three Ultrasonic Nebulizers, Ann. Allergy 26, 18-27 (.1968).
13. Sollner, K. The Mechanism of the Formation of Fogs "by Ultrasonic
Waves. Trans. Faraday Soc. 32, 1532-1536 (1936).
iH. Boguslauskir, Ya. Ya. and 0. K. Eknadiosyants. Physical Mechanism of
the Acoustic Atomization of a Liquid. Akusticheskii Zhurnal (.Engl.
transl) 15_, lU-21 (1969).
15. Lang, R. J. Ultrasonic Atomization of Liquids. J. Acoust. Soc. Amer.
3jb 6-8 (.1962).
16. Postendorfer, J. , J. Gebhart and G. RoMg. Effect of Evaporation on
the Size Distribution of Nebulized Aerosols. J. Aerosol Sci. 8_,
371-380 (1977).
-------
APPENDIX A
DATA FOR INDIVIDUAL SQUIRTS OF METERED AEROSOL: CANS
Squirt
Aerosol Can No.
AS #1 -v 88a
0.1$ benzoic 89
acid 139
lUo
190
191
216
217
231
AS #2 1
0.1$ benzoic 2
acid 3
U
5
6
7
8
9
10
AS #3 2
0.1$ benzoic 3
acid U
Gvv^LvL
5
6
7
8
9
10
11
Change in
Can Mass
(mg/squirt)
52.5
55.7
55.2
5U.9
57-7
55.9
25. u
38.9
2U.1
66.0
55.6
5U.6
5U.7
56.2
55-5
50.1
51.1
52.9
•50.6
53-3
51.1
50.1
50.8
51-5
51.1
50.8
50.9
51.9
50.5
Aerosol
Number x 10 ^
(per squirt)
5.U8
3.82
3.60
3.36
3.02
3.06
1.61
3.00
1.39
U.57
3.59
3.92
3.U5
3.UO
3.52
3.38
3.U7
3.U7
3.UO
1.88
2.U9
2.05
2.06
2.10
2.11
1.93
2.11
2.15
2.08
Aerosol
Volume x 105
(cm3/squirt)
3-25
2.15
2.09
1.92
1.72
1.91
0.692
1.70
0.5U5
3.08
2.13
2.53
1.99
2.02
1.98
1.81
1.87
1.83
1.90
1.13
1.33
1.16
1.13
1.19
1.09
1.08
1.21
1.21
1.15
Mass
Median
Diameter
(ym)
1.5
1.5
l.U
l.U
l.U
1.5
1.3
l.U
1.3
1.6
1.5
1.5
1.5
1-5
l.U
l.U
l.U
l.U
l.U
1.5
*
l.U
1.5
1.5
*
l.U
l.U
l.U
\
l.U
1
l.U
l.U
U3
-------
APPENDIX A
DATA FOR INDIVIDUAL : SQUIRTS OF :METERED AEROSOL CMS
Squirt
Aerosol Can No.
AS #5 5
0.1% benzoic 6
acid 7
8
9
10
11
12
13
ll*
AS #7 M.30a
3% benzoic 131
acid 132
133
131*
135
136
137
138
139
AS #8 6
3% benzoic 7
acid 8
9
10
11
12
13
ll*
15
Change in
Can Mass
(mg /squirt)
U0.8
1*5.7
1*1*. 0
1*3.8
1*1*. 1*
\^C ^ 'J
Ij-ij. _ l^.
1*2.0
1*1*. 3
1*9.8
55-2
k9.9
50.6
50.0
50.6
51-5
1*9.0
53.7
51.2
52.5
51.5
5l*. 3
55.1
52.1
52.9
55-5
53.1*
5l*. 2
52.8
52.8
Aerosol
—•• n
Number x 10 '
(per squirt)
3.11*
2.96
2.1*5
1.98
2.0l*
2.03
1.72
1.70
1.52
1.57
1.79
1.91
1.78
I.Ik
1.83
1.68
1.51
1.56
1.67
2.02
2.03
2.10
1.92
2.15
1.85
2.20
1.89
1.71
1.73
1.8l
Aerosol
Volume x 105
(cm3/squirt)
3.08
2.96
2.1*6
2.01
1.93
1.87
1.76
1.73
1.33
1.1*8
2.60
3.73
3.26
3.10
3.30
2.71
2.11
2.01*
2.56
2.87
1*.32
5.22
3.75
1*.66
3.29
5.75
U.09
3.37
3.67
3.72
Mass
Median
Diameter
(urn)
2.0
2.2
2.1
2.3
2.1
2.1
2.3
2.2
2.1
2.2
2.9
3.3
3.2
3.2
3.2
3.2
2.9
3.0
3.0
2.9
3.1
3.2
3.1
3.2
3.0
3.5
3.0
3.0
3.1
3.0
-------
APPENDIX A
DATA-FOR INDIVIDUAL SQUIRTS OF METERED AEROSOL CMS
Squirt
Aerosol Can No.
AS #9 6
3% benzoic 7
acid 8
9
10
11
12
13
ll*
15
_ _ ,^«a
Isuprel ^120
121
122
123
12 1*
125
126
127
128
129
Bronitin -vlOOa
101
102
103
ioi*
105
106
107
108
109
Change in
Can Mass
(mg/s quirt)
50.5
5**. 3
51.1
51.3
55.1
52.1
61*. 7
50.6
50.9
56.3
55-7
63.1
52.8
52.8
53.5
61.5
55.^
58.9
59 A
53.7
62.1
63.2
62.8
61*. 0
62.7
63.7
62.8
62.7
62.1
62.9
Aerosol
Number x 10 7
(per squirt)
2.01
2.30
2.09
2.09
2.10
1.88
2.09
1.95
2.13
2.01
2.06
1.90
1.95
1.68
1.75
1.85
1.7U
2.23
1.76
1-75
5.69
6.82
8.76
6.U6
7.12
7-71*
7.88
7-38
5.27
8.23
Aerosol
Volume x 105
(cm3 /squirt)
U.81
5.31
5.20
3.98
i*.86
l*-32
l*.6l
i*.37
1*.06
3.90
1.81*
1.55
1.65
1.3!*
1.1*8
1.52
1.1*1*
2.01
1.51
1.1*7
U.38
5.l6
6.81
l*.6l*
5.01
5.U3
5.78
5.60
3.68
6.1*1
Mass
Median
Diameter
(pm)
3.3
3.2
3.2
3.1
3.U
3.1*
3.2
3.2
3.0
3.2
2.0
1.9
2.0
1.9
2.0
1.9
2.0
2.0
2.0
1.9
1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.6
alnitial squirt number is estimated
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1, REPORT NO.
EPA-600/2-79-191
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
TECHNIQUE FOR IN SITU CALIBRATION OF PARTICULATE
HASS MONITORS
5. REPORT DATE
October 1979
6, PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
W. John, S. Hering, and J. J. Wesolowski
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Air and Industrial Hygiene Laboratory Section
California Department of Health Services
Berkeley, California 94704
10. PROGRAM ELEMENT NO.
1AD712B BA-012 (FY-78
11. CONTRACT/GRANT NO.
Grant R805577010
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory - RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park. N. C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Two types of aerosol generators, the Riker Laboratories metered spray can and the
Mistogen EN145 ultrasonic nebulizer, were evaluated by laboratory measurements for
application to the in situ calibration of particulate mass monitors for stationary
sources. The metered spray can delivers a fixed amount of aerosol each time the
valve is depressed. The average mass of propellent and solution in each squirt is
52 mg, and is reproducible within 4%. The total volume of the dried particles per
squirt is of the order of 10~5 cm3. The volume median diameter was varied from 1.4
to 3.2 ym by selection of solute concentration. Because of its simplicity and repro-
ducibility of output, the metered spray may be useful for a variety of applications
requiring a portable aerosol source.
For calibrating stack beta gauges, larger aerosol output of 5-10 mg is needed, requir-
ing a valve with a metering volume at least ten times larger than present valves.
Contact electrification monitors require a test aerosol of 25 mg/m3 at a flow of
1 nr/min. Appropriate for this application is the ultrasonic nebulizer, which has an
output of 50 mg/min, constant to within 8% over a period of hours.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
* Air pollution
* Aerosol generators
* Evaluation
* Calibrating
Monitors
Particles
Weight (mass)
13B
13D
14B
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
21. NO. OF PAGES
56
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE,
C46
------- |