U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
PB-273 018
Methodology and
Instrumentation to Measure
Gaseous Ammonia
Monsanto Research Corp, Dayton, Ohio Dayton Lab
Prepared for
Environmental Sciences Research Lab, Research Triangle Park, N C
Aug 77
-------
TECHNICAL REPORT DATA
(Please read iHXtuctions on the reverse before completing)
8
1. REPORT NO.
EPA-600/2-77-125
4. TITLE AND SUBTITLE
METHODOLOGY AND INSTRUMENTATION TO MEASURE
GASEOUS AMMONIA
. AUTHOR(S)
). David, D. Ruffin and M. Willson
(-EP'-QRMIfJG ORGANIZATION NAME AND ADDRESS
Monsanto Research Corporation
Dayton Laboratory
515 Nicholas Road
Dayton. Ohio
3. RE
5. REPORTTJATE
August 1977
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO
10. PROGRAM ELEMENT NO.
1AD712 BE^-11 (FY-77)
11. CONTRACT/GRANT NO.
68-02-1793
2. 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
5. SUPPLEMENTARY NOTES
6. ABSTRACT
Methodology for the time integrated collection and analysis of atmospheric ammonii
was developed. Ammonia is primarily measured because it can react with SOX to produce
ammonium sulfate. Since SOX is one of the principle air pollutants, it is important
to determine its atmospheric reactions.
The methodology makes use of optical wave guides which are 1 x 20 mm quartz rods
coated with a chemical that is specific for NH3- As ammonia reacts with the chemical
on the rod, the optical properties of light passing through the rod change and can be
directly related to the NH3 concentration. An optical wave guide analyzer was develop
ed to measure the resulting change in optical properties. Concentrations as low as
ppb can be assayed. The coated quartz rods are placed in the field for 24 hours and
brought back to the laboratory for analysis by the analyzer.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Croup
*Air Pollution
*Ammonia
Measurement
Development
*Analyzers
Ouartz
Optical Properties
ATONAL TECHNICAL
INFORMATION SERVICE
U. S. DEPARTMENT OF COMMERCE
SPRINGFIELD. VA. 2061
13B
07B
14B
08G
20F
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (Thli Report)
UNCLASSIFIED
21. NO. OF PA
20. SECURITY CLASS (Thljpage/
UNCLASSIFIED
22.PRICE
Mb-ftol
EPA Form 2220-1 (9-73)
-------
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
providas 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-600/2-77-125
August 1977
METHODOLOGY AND INSTRUMENTATION TO MEASURE GASEOUS AMMONIA
D.J. David
M.C. Ulllson
D.S. Ruffln
Monsanto Research Corporation
Dayton Laboratory
Dayton, Ohio 45407
Contract Number 68-02-1793
Project Officer
James Mullk
Atmospheric Chemistry and Physics 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 Science
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.
ii
-------
ABSTRACT
Methodology for the time integrated collection and analysis
of atmospheric ammonia was developed. Ammonia is primarily mea-
sured because it can react with SOX to produce ammonium sulfate.
Since SOX is one of the principle air pollutants, it is import-
ant to determine its atmospheric reactions.
The methodology makes use of optical wave guides which are
1 x 20 mm quartz rods coated with a chemical that is specific
for NH3 concentration. An optical wave guide analyzer was de-
veloped to measure the resulting change in optical properties.
Concentrations as low as 1 ppb can be assayed. The coated quartz
rods are placed in the field for approximately 24 hours and brought
back to the laboratory for analysis by the analyzer.
iii
-------
CONTENTS
Abstract iii
Figures vii
Tables x
1. Introduction 1
2. Concept and Theoretical Model 3
2.1 Optical Waveguide Theory 4
2.2 Optics of the Coating 9
3. Instrumentation 10
4. Experimental 19
4.1 Survey and Selection of Ammonia Reactions. . 19
4.2 Analytical Reaction Screening ....... 22
4.3 Coating Studies 25
4.4 Exposure to Ammonia of Polymer/Reagent
Coatings on Waveguides 39
4.5 Exposure to Ammonia of Polymer/Reagent
Coatings on Quartz Waveguides to Determine
Effect of Gas Stream Velocity 42
4.6 Sensitivity and Dynamic Range of Collector/
Sensors and Effect of Relative Humidity. . . 46
4.6.1 Effect on Sensitivity of
Coating Polymer Type 49
4.6.2 Effect of Ammonia Concentration on
Collector/Sensor Response 52
4.7 Stability of Collector/Sensors 57
4.8 Effect of Rod Length on Collector/
Sensor Sensitivity .65
4.9 Effect of Acid Pollutants and Ammonium
Compounds on Collector/Sensor Response ... 68
4.10 Effect of Temperature on Collector/
Sensor Response 72
4.11 Absorption Wavelength 75
Preceding page blank v
-------
5. Calibration 77
6. Results and Discussions 79
6.1 Ambient Air Measurement of Ammonia in the Laboratory
Using an Artificial Humidity Chamber 79
6.2 Field Test Results 83
7. Procedures 95
7.1 Solution Preparation . 95
7.2 Rod Coating 96
7.3 Analytical Results 98
7.3.1 Collector Sensor Exposure and Calculations . . 98
References 100
vi
-------
FIGURES
Number Page
1 Light path in collector/sensor
waveguide and associated optics 5
2 Collector/sensor model 7
3 Block diagram of the gradient
light analytical detector (GLAD). 14
4 Gradient light analytical detector (GLAD I) . . . 16
5 Rotating drum waveguide coating apparatus 27
6 Rod holder for dip coating with
beaker of polymer solution 33
7 The effect of high humidity and/or concentrated
NH^OH on the transmission properties of PVA/
reagent coated rods 41
8 The effect of both low and high humidity on
PVA/ninhydrin coated rods exposed to 0 & 1 ppm
ammonia concentrations in an air stream 43
9 Schematic of equipment used to expose collector/
sensors to known concentrations of ammonia in
air 44
10 Transmittance of 3 cm long collector/sensors as
a function of exposure time and air stream
velocity 45
11 Hygroscopioity of various polymers 47
12 Effects of relative humdity on collector/
sensor sensitivity to ammonia 48
13 Schematic of equipment used to expose
collector/sensors to known concentrations
of ammonia in air 50
14 The effect of polymer coating type on
collector/sensor response using ninhydrin
reagent . 51
vii
-------
Number Page
15 The sensitivity of PVP/ninhydrin collector/
sensors at different ammonia concentrations
in air 53
16 The sensitivity of PVA/ninhydrin collector/
sensors at different ammonia concentration
in air 54
17 The effect of relative humidity on collector/
sensor response at a fixed concentration of
ammonia in air 55
18 The effect of relative humidity on PVP/
ninhydrin collector/sensor response at
two ammonia concentrations in air. .. 56
19 The slope of PVP/ninhydrin collector/
sensor transmittance curves as a function
of ammonia concentration in air „ . 58
20 Effect of rod length on the transmission
properties of coated rods exposed to 1 PPM
ammonia in a saturated air stream 66
21 The effect of high humidity and/or ammonia
concentration on 4 cm long PVA ninhydrin coated
rods exposed to 0 & 1 ppm ammonia in an
air stream 67
22 Schematic of apparatus used to simultaneously
expose collector/sensors to acid pollutants
and ammonia 69
23 Effects of both acid pollutants and ammonia
on the transmittance of PVP/ninhydrin coated
quartz rods 70
24 Schematic of apparatus used to expose
collector/sensors to ammonia at various
temperatures 73
25 Effect of temperature and relative humidity on
collector/sensors at 60 PPB ammonia level 74
26 Absorption characteristics of reacted
collector/sensors <> 76
27 Effect of both ammonia concentration
and exposure time on c/s absorbance ....... 78
viii
-------
Number Page
28 Artificial humidity chamber 80
29 Modified GLAD collector/sensor holder 84
30 Holder in position under roof of
high volume sampler 88
ix
-------
TABLES
Number Page
1 Transmission Measurements of an
Uncoated Quartz Rod Using GLAD 17
2 Colorimetric Ammonia Reactions . 20
3 Analytical Reagent/Reaction Requirements
for EPA Ammonia Collector/Sensors 22
4 Laboratory Evaluation of Ammonia Reactions 24
5 Process Conditions and Coating Properties
of Polymer/Reagent Coatings on Quartz
Waveguides (Rods) .28
6 Variation of Polymer/Reagent
Coating Thickness and Weight.- „ . 30
7 Coating Thickness and Weight Reproducibility
of PVA/Ninhydrin Coatings Made Using Rotating
Drum Method 31
8 Process Conditions and Coating Properties
For Rods Coated Using Vertical Axis Method. „ ... 35
9 Comparison of Coating Methods 36
10 Coating Thickness and Weight Reproducibility
of PVA/Ninhydrin Coatings made Using Vertical
Axis Method 38
11 Process Conditions and Coating Properties for
Rods Coated with a Number of Different Polymers
and Polymer Combinations 40
1.2 Stability of Polymer/Ninhydrin
Coated Collector/Sensors 59
13 Long Term Stability of PVA/Ninhydrin
Coated Collector Sensors 61
14 Long Term Stability of PVA/Ninhydrin
Coated Collector/Sensors ,62
-------
Number Page
15 Stability Comparison of Polymer/Ninhydrin
Coated Collector/Sensors Stored at Ambient
and Sub-Ambient Temperatures 64
16 Exposure Conditions and Results of
Collector/Sensors Exposed to Ambient Air
in the Laboratory 82
17 Analytical Results of 1st MRC Dayton
Laboratory Field Site Testing 87
18 Analytical Results of St. Louis
Field Site Testing at EPA Regional
Air Pollution Site #108 in Vicinity
of Granite City, Illinois 92
19 Analytical Results of 2nd MRC Dayton
Laboratory Field Site Testing Using
Improved Holder and PVA/PVP Coating 92
xi
-------
SECTION 1
INTRODUCTION
One of the goals of the Environmental Protection Agency (EPA)
is to accurately monitor the concentration of ammonia in am-
bient air. The concentrating range of interest is 4 to 400
yg/m3 (fy 5-500 ppb) for ammonia and 0.04 pg/m3 for particulate.
Thus, a sensitive specific method of analysis is required,
especially for the lower end of the concentration ranges.
Measurements in this range are needed to allow detection and
measurement of background concentrations of about 6 ug/m3.1
Man contributes a comparatively small portion of ammonia and
the major increases or decreases in ambient air are probably
due to naturally occurring atmospheric phenomena in which ni-
trogen species are created, converted, or removed.
At the present time, monitoring is generally accomplished by
periodic measurements with devices such as the midget impinger
or sophisticated, expensive instruments such as the long path
infrared or chemiluminescent analyzers. Each approach has its
own inherent limitations. Midget liquid impingers use bulky
^Sittig, Marshall, Pollution Detection and Measuring Handbook,
Park Ridge, N. J., Noyes Data Corp. (1974).
-------
and cumbersome solution techniques for collection and require
long collection times (100-1000 hours) to achieve the desired
sensitivity. The use of real-time instrumentation (which is com-
plex and expensive) in the field is useful since it provides the
peak concentrations but unfortunately does not give time weighted
average concentrations. In addition, this latter approach is not
amenable to large scale surveillance.
The purpose of this program, therefore, was to develop practical
instrumentation and methods for detecting and determining ambient
air concentrations of ammonia in gaseous and particulate form.
-------
SECTION 2
CONCEPT AND THEORETICAL MODEL
The prime prerequisites for any collection/analysis technique
are inherent sensitivity and specificity to fulfill quantita-
tive analysis requirements. Thus, other components of polluted
air must not interfere to the -extent that erroneous measurement
results are obtained. Secondarily, the technique should be
relatively simple and not require expensive and complex ana-
lytical instrumentation. Only if both these criteria are met
can the technique be expected to gain wide acceptance and pro-
vide the large scale analytical results desired by the EPA.
In addition to the direct application of the technique for the
collection and analysis of ammonia it is highly desirable that
the technique have wider and more general applicability in the
determination of other air pollutants.
The concept we chose in order to accomplish the goals of this
program is based on a new technique and instrumentation2 that
utilizes coated optical waveguides for the spectrophotometric
measurement of chemical reactions carried out in situ on the
coated surface of the waveguide. The high sensitivity
2Hardy, E. E., David, D. J., Kapany, N. S. and Unterleitner,
P. C., Nature 257, 666 (1976).
-------
demonstrated, the small size, the simplicity, and the low cost
of the sensitized waveguides and instrumentation appeared to
make this technique ideal for large scale ambient air monitor-
ing o
The basic approach is to sensitize a waveguide to selectively
reaet with ammonia by incorporating a reactive chemical into a
passive coating on the surface of the waveguide and then plac-
ing the sensitized waveguide at desired field locations for re-
action . If the reaction brings about a color change or light
scattering, the change in light transmission (T) through the
coating can be measured spectrophotometrically. A simplified
model of the system is shown in Figure 1, i.e., light source,
waveguide, and detector.
2.1 OPTICAL WAVEGUIDE THEORY
Although the general optical mechanisms have been considered
previously,2 we felt that it was necessary to examine a model
in detail in order to define those physical parameters that
would influence sensitivity dynamic range, and analytical
accuracyo The question of sensitivity is of particular impor-
tance! in the case of the determination of low levels of pollu-
tant:; in ambient air.
-------
LIGHT PATH IN COLLECTOR/SENSOR WAVEGUIDE
SIGNAL
PHOTODETECTOR
COATED WAVEGUIDE
LENSES/APERTURE/FILTER
LIGHT SOURCE
Figure 1. Light path in collector/sensor
waveguide and associated optics
-------
Accordingly/ the model shown in Figure 2 was considered.3 This
mode] consists of a quartz rod of refractive index n\ coated
With a polymer containing reactants for a material of refractive
index n2 and light entering the rod from a medium (air) of re-
fractal ve index I\Q.
Thus,, when the coated quartz rod is functioning as a waveguide,
light impinging on the polymer/air interface at angles greater
than rj2 is internally reflected, and light impinging on the
polyner/air interface at angles less than 4>2 is lost.
for (J>2 = 42°
4/2 - 48°
From Snell's Laws we have
= n2 sin
-------
"c.
Figure 2. Collector/sensor model
"0
2 = critical angle, where sin $2 = — for a critical
situation ?
no = refractive index of air s 1.0
r\\ - refractive index of quartz - 1.48
no = refractive index of polymer matrix » 1.50
sin $2 ~ V VQ f $2 = 42°
«i
4>3 = critical angle, where sin 3 = —- for a critical
situation 2
x = coating thickness
6! = acceptance angle or rod or angle of incidence
i|»i = angle of refraction in rod
n2
<)>! = critical angle where sin 4>i » — for a critical
situation *
e2='j»i = acceptance angle of polymer matrix and reactants or
angle of incidence
= angle of refraction in polymer matrix
-------
Further, this result shows that light at all angles is accepted
and that even light entering at 90° will be internally reflected
in the waveguide, in practace a hollow cone of light will be
instrumentally allowed to enter the end of the rod at steep
angles (=70°-80°) by means of an appropriate aperture in order
that the maximum and consistent path length through the reactive
coating is followed as nearly as possible by all light rays.
However, some of the light entering the coating will be trans-
mitted through the coating as such and will not reenter the rod.
This is dictated by $3 which is the critical angle which de-
termines whether or not internal reflection occurs in the poly-
mer and is defined by
*i
sin* 3 = — = (j»3 = 81°
a n2
Since light will normally impinge on the polymer glass inter-
face at angles less than the critical angle or angles about 42°,
it will not be internally reflected within the polymer coating.
This effect depends on the spread of angles of the hollow cone
of light and whether or not scattered light at very low angles
of incidence is able to enter the rod. This aspect is con-
trolled instrumentally.
The number of reflections through a waveguide is given by
- sin9
m ~ d(n?-sinze)
8
-------
where 0 = acceptance or incidence angle
d * diameter
n\ » refractive index of waveguide
For a 0.1 cm rod 2 cm long, n = 12
m
Assuming a coating thickness of 'v-O.Ol cm, which is borne out by
actual measurement, we have for the path length
sin 48° =
° cm
pin 48°
The total path length = 1 total = 0.0135 x 2 x 12 = 0.32 cm
2.2 OPTICS OF THE COATING
There appear to be three mechanisms which control the inter-
action of a gaseous pollutant with the collector/sensor and its
ability to be measured:
(1) The amount of sample contacting the polymer
matrix layer
(2) The diffusion of the active component through the
polymer matrix layer; and
(3) The light absorption of the reaction products in
accordance with Beer's Law
The change in concentration with respect to time can be formally
-------
expressed as:
dC
where C = concentration of pollutant vapor in polymer
j
CG = concentration of pollutant in gas or air
EG = extraction coefficient of polymer/polymer
reactants.
E,, is determined by the diffusion of the active sample component
G
through the polymer film and the reactants ability to effec-
tively capture it after the component has contacted the polymer
matrix.c Under simplified conditions, where both EG and CG (mass
transfer from the air not limiting) are constant, we have
Cp = CG X EG
From previous experiments, we utilized data at which the re-
sponse to a known concentration of gaseous HCN over a specified
time interval was the same as when a known concentration of
j
NaCN was applied to the rod which allows an estimation of E ,
assuming similar thicknesses of layers and extraction co-
efficients.
The absorption coefficient kj from Beer's Law, T = I/ = e ,
•••0
was estimated in a similar manner. Assuming a similar response
for NH;) and a time span of 48 hours allowed estimation of
10
-------
contained in the volume element, V = nxdl. This con-
centration in g/cm was substituted in Beer's Law equation equa-
tion which showed that a measurable absorbance could be obtained
at the desired lower concentration limit.
The model also illustrated that the C/S operates as a concen-
tration-sensitive detector as opposed to a mass-sensitive de-
tector.1*'5 Since flow does not appear in the equation, the de-
vice should be essentially flow insensitive and it should not
be necessary to monitor flow or attempt to pump air over the
C/S.
The extraction or partition coefficient (E ) addresses the re-
moval of NH3 from the air only on a gross basis since, in
addition to removal, the diffusion of the NH3 through the poly-
mer coating and the rate of reaction of the NHs with the re-
actant in the polymer are involved in producing the reaction
product. The diffusion of the NHa into the polymer coating and
mass transfer to the ninhydrin, which maintains a low or zero
concentration gradient by formation of irreversible colored
reaction products, may well be the reaction limiting steps, as
opposed to the velocity with which the sample is brought to the
C/S.
l+Halasz, I. Anal. Chem. 3£, 1428 (1964).
5David, D. J., "Gas Chromatographic Detectors", Wiley-Inter-
science, New York (1974).
11
-------
These aspects hopfully provide guidance in exercising control
over those factors primarily responsible for sensitivity and
quantahive response and how they might be manipulated to pro-
vide trie desired result.
12
-------
SECTION 3
INSTRUMENTATION
To measure the response of the collector/sensors, suitable instru-
mentation was fabricated. Similar instrumentation has been des-
cribed previously.
A model of the basic elements of this instrumentation and its
use with a reactant coated waveguide is shown in Figure 1.
These elements consist of a light source, collimating lens,
aperture to produce the hollow cone of light, a lens to focus
the cone on the end of the coated waveguide, and a photode-
tector that measures the axially transmitted light. The instru-
ment into which these elements have been incorporated is termed
a Gradient Light Analytical Detector (GLAD). This instrument
will accommodate rods 0.9 - 1.3 mm in diameter and either 10 or
20 mm long. A schematic diagram showing the basic optical de-
sign is shown in Figure 3. Three filters are available which
allow the selection of white, red, blue, and green light. After
multiple reflections within the rod, the light from a tungsten
lamp emerges at the upper face and is scattered by a diffuser,
part of the light going into a silicon photodiode detector. An
13
-------
Detector
Hemisphere
Annular Aperture
f
Figure 3. Block diagram of the gradient
light analytical detector (GLAD)
f3)
(4)
(5)
16)
(8)
(•9)
(10)
a tungsten filament lamp light source
& eondenser system to produce nearly collimated light
a filter for wavelength selection
an annular aperture to block axial light rays
a condenser to produce a hollow cone of light rays
coupling hemispheres and apertures to couple large
angle rays into the rod
a rod mount to accurately position rods with respect
to the aperture while presenting a minimum of surface
contact
a silicon photidiode detector
an operational amplifier operating as "current-to-
voltage" converter
a three-digit voltmeter for relative-transmittance
readout
14
-------
output voltage suitable for the 200 mv full scale digital panel
meter is selected by a decade range switch. The actual instru-
ment is shown in Figure 4.
In order to separate inherent instrument variations from coat-
ing process variations a single uncoated quartz rod was se-
lected and measured over a period of several weeks and at
different times during the day. Two readings were taken at any
one time and an average of these two was used. The procedure
used for all transmittance measurements was to turn the rod end
for end after the first reading. The results of this study are
shown in Table 1. The percent variance is very minimal in the
case of green light and thus ensures meaningful transmission
measurements.
In accordance with the requirements of this contract this in-
strument (GLAD I) will be delivered capable of measuring the
light transmittance before and after exposure to ammonia of
polymer/reagent coated quartz rods (collector/sensors). The
instrument to be delivered is shown in Figure 4. The collector
sensors are inserted into the front of the instrument and form
part of the optical system. A filter wheel allows the trans-
mittance to be measured at a number of wavelengths. A special
interference filter has been installed which will provide
measurements at 580 run ± 10 nm. This is at or near the wave-
length where maximum absorption due to the ninhydrin-ammonia
15
-------
(a) Overall view
(b) Inserting collector/sensor
Figure 4. Gradient light analytical detector
(GLAD I)
16
-------
TABLE 1. TRANSMISSION MEASUREMENTS OF AN
UNCOATED QUARTZ ROD USING GLAD
Date
(1975)
~7/9
7/15
7/17
7/18
7/21
7/22
7/22
7/23
7/23
7/23
7/24
7/24
7/25
7/31
8/1
8/4
Trial (X)
1
2
3
4
5
6
7
8
9
10 '
11
12
13
14
15
16
Average (X)
Variance
Percent Variance
White Light,
millivolts
128.5
122.2
122.8
128.4
134.6
127.6
129.7
128.1
132.5
131.5
128.0
131.4
126.8
127.2
121.9
125.1
127.9
13.2
±5.2
Green Light,
millivolts
.745
.715
.715
.744
.778
.742
.744
.727
.774
.750
.728
.760
.718
.728
.696
.718
.736
.033
±2.2
Variance
- nX2 where n = 16
n - 1
17
-------
reaction takes place.
A more advanced, feasible, version of this instrument was de-
signed and developed in house. This instrument, GLAD II, was
also used during this program as a supplement to GLAD I. This
versatile instrument allows a broader selection of filters and
light entry angles along with greater sensitivitv.
18
-------
SECTION 4
EXPERIMENTAL
4.1 SURVEY AND SELECTION OF AMMONIA REACTIONS
A library literature search was conducted in an effort to find
analytical ammonia reactions with emphasis on colorimetric ones.
Typical sources were books on analytical reagents and spot test
methods. Also, the chemical abstracts were searched from the
present back to 1925. A computer search of the chemical ab-
stracts to 1972 was also made and proved particularly helpful.
Pertinent information sources are listed in literature cited.6~70
Table 2 shows a tabulation of those reactions which looked
promising. An analysis of these reactions according to the
characteristics shown in Table 2 was then made. Out of a total
of thirty-eight potential reactions, eighteen were selected for
further study. The status of each reaction is also shown. In
most cases, a reaction was disqualified prior to screening if
the stability of the reagents or reaction products was judged
to be poor.
At the outset of this work we felt that the reactant coating
formulation must meet a number of system requirements. Consider-
ation of possible reactants, the process of selection, and the
Reference Cited
19
-------
TABLE 2. COLORIMETRIC AKKONIA REACTIONS
Iteaqmt/Beiciant
AlUwantira in Kxtlc Add
0-Senzenesulftraniio-
B™*!^
Calcium salfcte
Ghloxoplatinic acid
Cupric Chloride t
Chranototropic Acid
Qrprous Chlccldi e
Knqaneee Nitrite
CyprouB Chloride &
Tiimic Acid
Diflfnino phenol
Diozotizod BcnzidkJiB
Di
-------
Table 2. (Continued)
Raagent/Reactant
Phenol s Sodiun Nitreprusside
Ninhydrin
Picric Add, Ethyl Malonate,
Alcohol 1 SI Clyoerol
Potassiun Mercuri Iodide
(Nessler's Reagent)
Potassiun Thiocyanate 6
Ferric Alun*
Pyridine-Pyrazolone
Chloride (Riegler's Reagent)
Silver Nitrate * Formaldehyde
Silver Nitrate s Tannin
Sodiun Wpochlorite, Sul-
phanilamide, N-l Napthyl-
ethylenedianlne Dihydro-
chloride
Sodiun Nitroprusside I
Resorcinol
Sodiun Phenolate & Sodiun
typachlorite in presence
Of «3S>4
Sodiun Salicylate I
Nitroprusside
Sodiun Salicylate, Chlorine,
and Sodiun HypxrOarite
Thymol t Alkaline Bromine
Hater
Thymol t Chlozamine T
Thymol in Ethyl Alcohol
Sensitivity
microorms
2
0.9
100
O.OS
N.G.
o.s
0.7
.OS
0.1
20*
N.G.
N.C.
s
N.G.
280
N.G.
100
Specificity
Good
Sensitive
to volatile
amines
Sensitive
to potassium
Sensitive
to Hg, Od
M3, SO,"2
N.G.
Good
Sensitive
to amines
N.G.
N.G.
N.G.
Sensitive
to volatile
amines
N.G.
Good
N.G.
N.G.
N.G.
N.G.
Estimated stability
Color Reagent Reaction Products Rnarks
Blue
Blue
Yellow
(chrystal)
Yellow or
Orange-
Red
N.G.
Orange
Yellow-
Red
Black or
Silver
Silver
Pink
Green or
bluish
green
N.G.
Blue
Blue
N.G.
Blue
N.G.
Good
Good
Poor
Poor
Good
Fair
Poor
Fair
Poor
Poor
Good
Poor
Good
Poor
Fair
Fair
Fair
Good
Good
Poor
Poor
Good
Fair
Poor
Fair
Poor
Poor
Good
Poor
Good
Poor
Good
Fair
Fair
-
PEf jrptjt are
volatile
Iodine oolor
fates rapidly
Reagents are
unstable
Reagents are
unstable
Photosensitive
Photosensitive
Reagents are
unstable
Reagents are
unstable
Reagents are
volatile
Status
Mil be
evaluated
Has been
evaluated
Mil not be
evaluated
Mil not be
evaluated
Mil be
evaluated
Mil not be
evaluated
Mil not be
evaluated
Mil not be
evaluated
Mil not be
evaluated
Mil not be
evaluated
Mil not be
evaluated
Will not be
evaluated
Mil be
evaluated
Mil not be
evaluated
Has been
evaluated
Mil be
evaluated
Mil not be
evaluated
Reference
56
28
4, IS
13.37.S1
52,59
CS
45
22
12,20
2.3,21
46
17
30
29,44
33
31, M
64
18
• ug in 1 liter of air
•• rag in 1 liter of air
N.G. •vitot given
NOT REPRODUCIBLE
21
-------
subsequent qualification testing was carried out with these re-
quirements in mind. They are listed in Table 3.
TABLE 3. ANALYTICAL REAGENT/REACTION REQUIREMENTS
FOR EPA AMMONIA COLLECTOR/SENSORS
1. Provide sufficient sensitivity to respond to the ranges
of interest
2» Provide an analytical accuracy of about ±10%
3. Provide the required dynamic range
4. Provide the desired specificity under the conditions of use
5. Be compatible with the polymer solution matrix
6., Provide good optical properties in the final polymer
reactant coating
7o Provide polymer reaction coating formulation which is not
adversely affected by varying atmospheric conditions of
relative humidity and temperature
80 Provide sufficiently fast response so that short-term, high
concentration levels can be determined
9o Provide immediate reaction that does not require further
analytical workup in the laboratory
10. Provide the required stability of the end products
4o2 ANALYTICAL REACTION SCREENING
On the basis of the above mentioned analysis of potential re-
actions, screening of the most promising ones was carried out
qualitatively on filter paper. The reagents were evaluated by
observing the color change on filter paper when a small amount
of reagent in solution was exposed to ammonium hydroxide in
22
-------
deionized water. A blank was made by applying the reagent and
deionized water only to filter paper. In addition, the com-
patibility of the screened reagents with polyvinyl alcohol poly-
mer was determined. The results of the screening studies are
shown in Table 4.
Tentatively, three reactions were selected for evaluation in coat-
ing form on quartz waveguides. These are:
1. Ferric Sulfate [Fe2(SOu)3]
2. Ninhydrin (1,2,3-Triketohydrindene monohydrate)
3. Manganese Nitrate and Cuprous Chloride [Mn(N03)2+ CuCl]
Upon exposure to ammonia the ferric sulfate is thought to
chemically react to form ferric ammonium sulfate which has a
violet color according to the equation shown below.6
2NH3{aq) +
Fe2(SOu) 3.
Ninhydrin (triketohydrindene hydrate) was considered as a
possible reagent for the detection of low concentrations of
ammonia based on the fact that an intense blue to reddish-
violet color is formed above pH 2-3 when primary amino compounds
are heated with ninhydrin and that this reagent supported on
silica gel was used for the determination of NH3 in air.33
The intense color is due to the reaction of ninhydrin with the
liberated ammonia to form the intensely colored compound shown
23
-------
ro
-_
Reagent/Reactant
Ferric Sulfate
Manganese Nitrate
& Cuprous Chloride
Ninhydrin
Cuprous Chlor-
ide & Tannin
Thymol 6 Alkaline
Bromine Mater
Phenol S Sodium
Hypochlorite
Alloxantine in
Acetic Acid
Calcium Sulfate
Qualitative Spot Teac Color Rssulta1
•OOO ppm
Reagent Blank Concentrated NH3 in C
Concentration (Deionized Ammonia Ammonia
g/ml (H2O) Water) Water Water P
0.1 It . brown dk . brown brown
0.14 & 0.049 colorless dk. brown brown
0.035
(saturated
solution) It. blue brown & blus blue
0.01 6 0.05 It. brown yellow brown
O.I2 & 0.018 colorless colorless colorless
0.04 I 0.05 colorless colorless colorless
<0.04
(saturated
solution) It. brown red It. red
-------
below.
0
gjj * NH,
The reaction is positive for ammonium salts and amines, and
even very dilute solutions of ammonium salts will give a strong
positive test. It was found that the reagent was especially
sensitive to gaseous ammonia during chromatographic studies
when paper chromatograms were sprayed with ninhydrin solution.
Release of small quantities of ammonia into the room would
color the entire chromatogram an intense blue to bluish-purple.
The manganese nitrate and cuprous chloride reagent is an attempt
to replace the light sensitive silver nitrate by a copper I
salt, in the reaction shown below.
2 Ag
40H~
Mn02
2H20
A brown precipitate forms when a Lewis base, such as HU$, con-
tributes an OH" ion in water . 1 6
4 . 3 COATING
Using poly vinyl alcohol (PVA) polymer as a coating matrix,
25
-------
2 cm x 0.1 cm diameter quartz rods (waveguides) were coated with
polymer both with and without reagent. In early studies a rotat-
ing drum method was used which allows for the coating of between
one and twelve rods at any one time. The rods were oriented so
that as the drum rotated the rods, located on the pheriphery of
the drum, entered the solution with their axis in a horizontal
position. (See Figure 4). Upon each pass through the PVA
solution the rods picked up a certain amount of polymer/re-
actant which was then dried by passing the rods under an infra-
red lamp. Drying was accomplished with the rods in position on
the rotating drum holder to insure even distribution of the
coating, "he process conditions used in coating a number of
waveguides and their properties are shown in Table 5. The
use of a wetting agent was discontinued after observing no im-
provement in coating uniformity, especially with higher viscos-
ity solutions. The best coatings were obtained by making
one pass through the solution and air drying. Additional
passes after drying caused the coating to wrinkle. This is
probably due to the absorption of water from the second layer
into the first layer.
The coating thickness was measured, using a micrometer caliper,,
at both ends and in the middle of each rod and an average is
given in Table 5. In most cases, the coatings tended to be
slightly oval in shape. The uniformity around the circumference
of thes rods appears to be good as determined using an optical
26
-------
Figure 5. Rotating drum waveguide coating apparatus
27
-------
TABLE 5 . PROCESS CONDITIONS AND COATING PROPERTIES OF POLY-
HER/REAGEHT COATINGS ON QUARTZ WAVEG0IDES (RODS)
Set No.
185334
185358
185399
185804
185343
185360
185361
I of
Rods
4
6
6
6
6
6
4
Reagent
Ninhydrin
Ninhydrin
Ninhydrin
Ninhydrin
Mn(N03)2
+CuCl
Alloxantin
Fe2(SO,,) 3
Coating
PVA, * wt.
93.5
83.4
83.4
71.4
89
50
83.4
Average
-Ccs^paHlttQEL ; Coating Average Coating
Reagent % wt. Weight, g Thickness microns Color
-•: -*.•*!!__ o.
16.6 ^ 0.
16.6 0.
28.6 0.
9 0.
50 0.
16.6 0.
0024
0014
0021
0031
0027
0015
0030
20
25
24.
36
29
18
28
Lt. Blue
Lt. Blue
Lt. Blue
Lt. Blue
Lt. Tan
Lt. Pink
Lt. Brown
Visual, -i '.
-------
microscope.
A number of different reactants were incorporated into polymer
solutions and applied to quartz waveguides. Coating solutions
were also prepared using ninhydrin and a number of different
polymers in order to determine any differences in coating
characteristics. Summaries of these coating studies are given
in Table 6 and Table 7. The variation in coating thickness and
weight for different rods in the same set also shown in Table 6.
The rods were coated in a similar manner with the exception that
a different polymer was used with each set. Commercially avail-
able polymers were selected for this study on the basis of their
transparency, solubility, and compatibility with ninhydrin. The
uniformity from rod to rod in coating weight is considerably
better than the thickness uniformity. Also the PVA polymer is
considerably better than the other two polymers in coating
weight uniformity.
In order to determine the reproducibility in transmission
properties of 2 centimeter long collector/sensors, twenty-four
rods were coated using the rotating drum method. Three sets
of eight rods each were coated using optimized coating procedures.
The coating thicknesses and weights were also determined and are
tabulated in Table 7.
29
-------
TABLE 6. VARIATION OF POLYMER/REAGENT
COATING THICKNESS AND WEIGHT
Polymer
Percent Polymer In
Coating Solution
Reagent
Set Mo.
'# of Reds
Average Coating Weight ,g
Percent Weight Deviation
Averag;o Thickness, microns
Percent; Thickness Deviation
PVA - polyvinyl alcohol, ICN
Lot 95848
HEC - hydr-oxyethyl cellulose
,?QP~09L, Lot-W 1184-B
PVP - polyvinyl pyrrolidone,
PVA
8.3
Ninhydrin
185399
5
.0021
HEC
14
Ninhydrin
185393
6
.0015
14 26
24 37
26 19
Pharmaceuticals Inc. #2243
, Cellosize, Union Carbide
GAP Corp, Type
NP-K30, Lot
PVP
24 -
Ninhydrin
185396
6
.0011
21
20
46
Corp . ,
40822
_ • _ . . . „ Average Deviation
Percent Deviation - Average (wSight or Thickness)
,x 100 where
deviation is defined as the difference between the actual
value and the average value of coating weight or thickness.
30
-------
TABLE 7. COATING THICKNESS AND WEIGHT REPRODUCIBILITY OF PVA/
NINHYDRIN COATINGS MADE USING ROTATING DRUM METHOD
— , Coating
Spec, No. Thickness, vi
186818-1
186818-2
186818-3
186818-4
186818-5
186818-6
186818-7
186818-8
186819-1
186819-2
186819-3
186819-4
186819-5
136819-6
186819-7
186819-8
186820-1
186820-2
186820-3
186820-4
186820-5
186820-6
186820-7
186820-8
Average Thickness = 39y
Standard Thickness
Deviation (o) = ±lly
Average Weight = 2.2 mg
Standard Weight
Deviation (o) = ±.44 mg
55
36
36
29
38
49
36
25
38
3^
52
45
51
59
51
51
41
18
42
35
38
21
34
34
Coating
Weight, mg
1.8
1.6
2.7
2.0
2.1
2.6
1.7
1.8
2.5
2.0
2.6
2.1
2.3
2.0
2.0
1.8
3.0
2.6
2.4
3.0
2.4
1.9
2.3
1.2
31
-------
Using the rotating drum coating method the concentration of
polymer (polyvinyl alcohol) was varied in an effort to see the
effect of coating solution viscosity on coating uniformity. How-
eve:*, no improvement in coating uniformity was obtained over
that cbserved using the 8.3 wt.% PVA/ninhydrin solution found
best in previous studies. Pour centimeter long by 0.1 centi-
meter in diameter quartz rods were also coated following the
same coating procedures used for 2 centimeter length rods. Coat-
ing ncnuni fortuities were more prevalent with the longer rods.
Apparently, the longer length minimizes end effects due to the
rod holder and promotes the formation of bands as a result of
surface tension forces. The 4 centimeter long rods were used
in exposure studies to investigate the effect of collector/
senoor length on transmission properties when exposed to
ammonia.
In order to improve the coating uniformity a number of quartz
rods both 2 and 4 centimeters in length were coated by holding
at one end and slowly pulling out of the solution with the axis
of the red oriented vertically.(vertical axis method). A rod
holder was fabricated for this purpose which holds the rods at one
end along the periphery of a cylindrical piece of aluminum as shown
ir. Figure 5. The rods are placed into drilled holes and held by
means of set screws. The holder is capable of holding forty rods
for coating at one time. A Fischer-Payne dip coater is used at-this
time to pull the holder and rods in a vertical direction and at a
32
-------
Figure 6. Rod holder for dip coating with beaker of
polymer solution
33
-------
uniform velocity out of a beaker containing the coating solu-
tion. These coatings tend to be much more uniform than coatlnqs
mad© using the* rotating drum method. They are also thinner for
a given viscosity solution. A slight variation in thickness
from top to bottom was eliminated by rotating the rod 180° and
coating a second time. This also allowed the coating to extend
to the very end of the rod. The process conditions and coating
properties of quartz rods coated using the vertical axis
method are shown in Table 8. As would be expected those rods
coated with the higher viscosity PVA solution were the thickest„
The rotating drum and vertical axis coating methods are compared
in Table 9» The average thickness values given in Table 8 were
measured using an optical comparator. Four diameter measure-
ments* we:re taken along the length of each rod. The total num-
ber o;l measurements for all the rods within a set was then
averaged after subtracting out the diameter of the uncoated rod
to determine the average thickness. In order to better define
the deviation in thickness along the length of a rod, the maxi-
mum thickness reading for each rod within a set was then se-
lected and an average maximum thickness determined for the set.
The name was done for the minimum thickness readings to deter-
mine the average minimum thickness. As can be seen in Table 8
the percent deviation is considerably higher for those rods
coated using the rotating drum method. The difference is
large:? in the case of the 4 centimeter length rods. Also, the
34
-------
TABLE 8. PROCESS CONDITIONS AND COATING PROPERTIES FOR
RODS COATED USING VERTICAL AXIS METHOD
Specimen
Number
1852UU-1
1852U «»-2
185241-1
l852ll-'4
185233-1
l85233-<4
Rod Coating
Length, Solution
cm Wt %. PVA
U ^5
M -\,5
4 "V10
A -v.10
2 -x.15
2 -v.15
Coating
PVA
Wt %
70
70
70
70
80
80
Composition
Ninhydrin,
Wt. %
30
30
30
30
20
20
No. of
Coats
1
2
1
1
2
1
Coating
Thick-
ness, u
<5
•v-5
15
10
38
10
Visual
Appearance
smooth and
uniform
smooth -and
uniform
smooth and
uniform
smooth and
uniform
smooth ;
slightly
thicker in
center
smooth and
uniform
Remarks
Coating very
thin
Specimen reversed
for second coat
specimen reversed
for second coat
All specimens pulled with the rod axis in a vertical position at 5.2 cm/min from the coating
solution and air dried for ^2 hours before recoating or placing in sealed containers.
-------
TABLE 9. COMPARISON OP CO^TINR METHODS
Maximum
Thickness Readings
Set No.
185233
186818
185241
186821 .
1 Average
within
Zppr^onl
No. of Coating
Specimens Method
Vertical
6 Axis
Rotating
6 Drum
':-_ Vertical
IJh ;< Axis
Rod
Length
2 cm
2 cm
4 cm
Rotating
5 Drum 4 cm
j Thickness determined
-. rlovlAtMnn - r Max • Or
Average l
Thickness
15 P
41 p
13 v
69 v
by measuring at small
Average
Maximum
Thickness
23 v
58 p
18 p
125 P
Percent
Deviation2
50 .
87
40
81
Minimum
Thickness Readings
Average
Minimum Percent
Thickness Deviation5
8 P
10 p
(•
13 P
Intervals along the length of each
- Avg. Thickness y inn
50
75
60
81
rod
-------
average thickness is less in the case of those coatings made
using the vertical axis method even though the same coating
solution (8.3 wt.% PVA/ninhydrin) was used.
In the vertical axis coating method, the entire length of the
rod is coated, the rod is reversed and a second coat is applied.
The coating on the flat end is then easily removed with
a sharp knife or razor blade. A higher viscosity solution
(°» 15 wt.% PVA) than used with the rotating drum coating method
was found necessary to achieve a comparable coating thickness.
A numerical bubble viscosimeter was used to achieve reproducible
coating solutions.
Table 10 shows the degree of reproducibility obtained in both
coating weight and thickness for rods coated with PVA/ninhydrin
using the vertical axis method. In general, the results show
an improvement in coating reproducibility be a factor of five.
Consequently, this technique was used for all further work
throughout this program.
Additional coating studies were carried out using quartz which
vere dip coated using the vertical axis method from solutions
containing either polyvinyl alcohol (PVA) and/or polyvinyl pyrroli-
dene (PVP polymer and ninhydrin reagent. In several experi-
ments glycerin was added to the PVA solution to improve the
hygroscopic properties of the coating. Since the PVP has better
37
-------
TABLE 10. COATING THICKNESS AND WEIGHT RE-
PRODUCIBILITY OF PVA/NINHYDRIN COAT-
INGS MADE USING VERTICAL AXIS METHOD
Coating
Spec. No. Thickness, u
186885-1 20.3
186885-2 30.5
186885-3 33.0
186885-4 30.5
186885-5 27.9
186885-6 30.5
186885-7 33.0
186894-1 27.9
186894-2 25.4
186894-3 30.5
186894-4 33.0
186894-5 25.4
186894-6 20.3
186897-1 30.5
186897-2 33.0
186897-3 25.4
186897-4 27.9
186897-5 25.4
186897-6 40.6
186897-7 33.0
Average Thickness
Standard Thickness Deviation (a)
Percent Thickness Deviation
Average Weight
Standard Weight Deviation (a)
Percent Weight Deviation
Coating
Weight, mg
3.0
2.3
2.6
3.0
2.2
2.5
1.6
1.9
1.9
1.6
1.8
2.2
1.8
3.6
2.2
2.4
2.5
2.5
2.6
2.4
«• 29. 2y
= ± 4.8
= i!6.4
= 2.33mg
= ± 0.503
= ±21.6
38
-------
hygroscopic properties than the PVA and it should allow better
sensitivity to ammonia in dry air conditions. Table 11 shows
the coating properties for quartz rods coated using various
coating solutions. The variation in properties are due, in
most cases, to viscosity differences in the solutions. The
solution characteristics can be modified by varying the polymer
molecular weight with no adverse effect on polymer hygro-
copicity.
4.4 EXPOSURE TO AMMONIA OF POLYMER/REAGENT COATINGS ON WAVEGUIDES
Individual coated quartz rods containing the reagents shown in
Figure 7 were placed in a desiccator containing ammonia water.
Sister samples were placed in another desiccator containing only
distilled water. The color change observed in the ammonia ex-
posed rods was quantitatively determined by measuring the light
transmission axially down the waveguide as a function of time.
The results are shown in Figure 7. The ninhydrin reagent samples
showed the largest drop in transmission although all but the
alloxantin had acceptable changes in transmission.
In order to determine the characteristics of the collector/
sensors at lower levels of ammonia concentration a precision
gas calibration system was purchased capable of achieving parts/
billion NHa concentrations in a dilute gas stream. The system
uses sealed permeation tubes which contain liquid ammonia for
the 0.5-500 ppm levels. For lower levels an ultra low emission
39
-------
vrr
Ljt3 A£»L« rCI/xnCttt ojriB.LW*v.L-.tCMo Uoxni? Vtivi:.H-j\jj MAJ.S Pu«M'fiGLf
Coating
Composition
Set t
189507
189514
189518
189525
189527
189537
189539
189542
189511
189535
Rod
Length
cm
2
2
2
2
2
2
2
2
2
2
Polymer**
•PVA/
glycerin
PVA
PVA
•PVA/
glycerin
•PVA/
glycerin
PVA
PVA
PVA
PVP K-90
PVP K-90
Coating
Solution
wt.%
11
11
11
11
11
11
11
11
9
9
Polymer
wt.%
79
79
79
79
79
79
79
79
75
75
Nin-
hydrin
wt.%
21
21
21
21
21
21
21
21
25
25
1 of
Coats
2
2
2
2
2
2
2
2
2
2
Coating
Thickness
w
19.2
19.3
11.9
17.2
24.9
13.4
15.1
14.9
5.9
6.8
Visual
Appearance
smooth &
uniform
smooth &
uniform
smooth &
-•uniform
smooth *
uniform
smooth;
slightly
thicker
on ends
smooth &
uniform
smooth &
uniform
smooth &
uniform
smooth 6
uni form
smooth &
uni form
Remarks
Specimen reversed
for second coat
Specimen reversed
for second coat
Specimen reversed
for second coat
Specimen reversed
for second coat
Specimen reversed
for second coat
Specimen reversed
for second coat
Specimen reversed
for second coat
Specimen reversed
for second coat
Specimen reversed
for second coat
Specimen reversed
for second coat
• Ijil glycerin was added per 100 ml coated solution.
All specimens pulled with the rod axis in a vertical position at 5.2 cm/min from the coating
solution and air dried -v.2 hrs. before recoating.
**PVA (Lot 195848) obtained from ICN Pharmaceuticals Inc.; PVP (Type HP-K-90) obtained from GAP Corporation
-------
100
90
* 80
o>"
S 70
g
1 60
c
CO
50
^ 40
o>
o>
5 30
20 •
10 •
0
100% Relative Humidity, No Ammonia
. • Alloxantin
3 4 5
Time, hours
100* Relative Humidity, Concentrated NH4OH
^_-._—.—•Alloxantin
$
o
Figure 7.
•-. Fe2
-------
membrane housed in a stainless steel cartridge is used. It is
capable of achieving low parts/billion levels of NH3 in a diluent
gas stream.
Several PVA/ninhydrin coated collector/sensors were exposed to
^1 ppm NH3 in a diluent air stream of /v-375cc/min. The change in
light transmission was determined after 1.5 hours and 2 hours
as shown in Figure 8. Using dry air essentially no reaction
occurredo However, using nearly saturated air a colorimetric
reaction occurred along with an appreciable reduction in trans-
mission due to the presence of ammonia, whereas, blank runs
evidenced little change in transmission. Additional experiments
were conducted to better determine the relationship between
relative humidity and reaction time.
4.5 EXPOSURE TO AMMONIA OP POLYMER/REAGENT COATINGS ON QUARTZ
WAVEGUIDES TO DETERMINE EFFECT OF GAS STREAM VELOCITY
Experiments were conducted to determine the effect of gas ve-
locity over the surface of the coated quartz rods on trans-
mittance during exposure studies. To accomplish this a modified
experimental arrangement as shown in Figure 9b was required. The
first configuration shown in Figure 9a uses a glass tube spec-
imen chamber. This experimental configuration was used for
most of the exposure studies up to this time. The calculated
gas velocities obtained with a 0.178 cm2 cross sectional area
glass tube were ^2200 cm/min. By splitting the gas flow and
42
-------
*
oT
100
90
80
70
I 60
= 40
c
| 30
20
10
0
1 ( 1 ppm NH3
0 % Relative Humidity
22
Time, hours
100
90
70
50
§ 40
c
o>
I 30
20
10
0
•* | 100* Relative Humidity
0 ppm NH3
100% Relative Humidity
1 ppm NH3
1.5
Time, hours
Figure 8. The effect of both low and high humidity on PVA/
ninhydrin coated jtods exposed to 0 & 1 ppm Ammonia
concentrations in an air stream.
43
-------
Sample Chamber
for 0 ppm NH3
in Saturated Air
Sample Chamber
(0-lppmNH3
in Saturated Air)
Distilled Water
Flowmeter
KinTek Model 570
Gas Calibration
System
H2S04
Bubbler
a. High velocity air stream
Sealed X
Permeation
Tube
Compresseii
Air
KinTek Model 570
Gas Calibration
System
Flowmeter
0 - 2500cc/min
Sample Chamber
0 -1 ppm NH3 in
Constant Humidity
Chamber
Vented
Bubbler
b. Low velocity air stream
Figure 9. Schematic of equipment used to expose
collector/sensors to Known concentrations
of ammonia in air.
44
-------
100
5 80
O
o>
I «
Q)
O
40
E
i/>
c
20
High Velocity Air Stream
Low Velocity Air Stream
_L
10
20
Time, minutes
30
40
Figure 10. Transmittance of 2 cm long collector/sensors as a
function of exposure time and air scream velocity.
45
-------
using a larger diameter desiccator chamber, as shown in Figure
9b. Essentially no difference was observed in collector/sensor
transmittance as a function of time for specimens exposed to the
same ammonia concentration using two different gas stream velo-
cities} as shown in Figure 10. In field sampling the collector/
sensors would never be used where the gas velocities would be
less tlhan these specimens were exposed to.
4.6 SENSITIVITY AND DYNAMIC RANGE OF COLLECTOR/SENSOR AND
EFFECT OF RELATIVE HUMIDITY
Earlior work had shown that little or no reaction between the
ninhydrin and ammonia took place at low humidity conditions.
By mixing known amounts of dry air with the humidified air,
different relative humidities were created which were used to,
test the effect of humidity on various reactant polymer coatings.
The curves in Figure 11 show the comparative hygroscopicity of
various polymers. Both the PVP and PVA polymers have been used
to form coatings on quartz rods. As the relative humidity de-
creases one would expect the PVP (polyvinyl pyrrolidene) to re-
tain irore moisture than the PVA and consequently remove more
ammonia from the surrounding air when used in conjunction with
ninhyclrin as a coating matrix. Therefore, the relationship be-
tween relative humidity and collector/sensor transmittance
should: follow the general shape of the curves shown in Figure 12.
The PVA would exhibit better handling characteristics at the
46
-------
80
o>
E
o
Q_
C5
O
-------
100
80
0>
te
ex
o
c
ro
g
'E
CO
C
ro
60
20
NH3 Concentration -Ipfim
Exposure Time ~lhr
0
20
40 60 80
Percent Relative Humidity
100
12. Kffects of relative humidity on collector/sensor
soniUt 1 v.1ty to ammonia.
48
-------
higher humdities because of the lower solubility in water of
this polymer. However, at the lower humidities the PVP should
provide better sensitivity to ammonia.
Earlier exposure studies were made at a fixed ammonia concen-
tration of 1 ppm and usually at high humidity levels. To in-
vestigate other NH3 concentrations the low emission membrane was
incorporated into the Kin Tek Model 570 gas calibration system.
Using the membrane, concentrations as low as 15 ppb ammonia can
be achieved in a moving air system. Figure 13 shows a schematic
of the experimental apparatus used in exposure studies.
Different relative humidities were obtained in the diluent air
stream by adjusting the flow rate through the distilled water
bubbler.
4.6.1 Effect on Sensitivity of Coating Polymer Type
PVA, PVA-Glycerin and PVP polymers were investigated in rod
coating studies as matrix materials for the ninhydrin reagent.
Coated quartz rods (collector/sensors) had different sensi-
tivities to low concentrations of arononia depending on the poly-
mer system. In general, the PVP/ninhydrin coating provided the
best sensitivity as shown for a particular concentration and
relative humidity in Figure 14. The least sensitive coating
system was the PVA-Glycerin. Therefore, further work emphasized
PVP/ninhydrin coatings.
49
-------
en
Compressed Air
Flowmeter
Sample
Chamber
§
Flowmeter
Low Emission
Membrane
KinTek Model
570 Gas
Calibration System
Bubbler
Distilled Water
Figure 13. Schematic of equipment used to expose collector/sensors to known
concentrations of ammonia in air.
-------
100
189528-2, glycerin
189518-6, PVA
189535-1 PVP
1 1-5
Time, Hours
Figure 14
-------
4.6.2 Effect of Ammonia Concentration on
Collector/Sensor Response
Initial collector/sensor calibration studies were made at
relatively high humidity levels (^80 R. H.) where maximum
sensitivity to ammonia is observed. Three ammonia concentra-
tions below 1 ppm were selected for study. Exponential trans-
mission decay curves were obtained for each concentration as a
function of exposure time. Essentially straight lines were ob-
tained for the log transmission vs. exposure time with the slope
of each line dependent on the concentration of NH3 in air.
Figure 15 and 16 show the percent change in transmission of PVP/
ninhydrin and PVA/ninhydrin coated collector sensors, respefct-
ively, as a function of ammonia concentration and exposure time.
From these experiments, it appears that at high humidity levels
both the PVA and PVP polymers retain sufficient moisture to
collect ammonia from the ambient air and allow measurement of
the ammonia concentration.
Sensitivity is uniformly lower for both polymers at low humidi-
ties,, This is shown in Figure 17 for concentration of 600 ppb
ammonia in air at three humidity levels using PVP/ninhydrin
collector/sensors. As can be seen the sensitivity is greatly
enhanced at high humidities. Even so, a measurable response is
obtained at the 45% relative humidity level with PVP/ninhydrin
coatings. Figure 18 shows the response as a function of
ammonia concentration and exposure time for two humidity levels
52
-------
80%RHPVP-Ninhydrin
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time, minutes
Figure 15. The sensitivity of PVP/ninhydrin collector/sensors
at different ammonia concentration in air
53
-------
80* RH PVA - Ninhydrin
0 10 20 30 40 50 60 70
Time, minutes
90 100 110 120 130 140
Figure 16. The sensitivity of PVA/ninhydrin collector/
sensors at different ammonia concentration
in air
54
-------
using PVP/ninhydrin coated collector sensors. A measurable
difference in response even at low relative humidity was obtain-
ed. By contrast, using PVA/ninhydrin coatings poor response as
a function of ammonia concentration at 45% relative humidity.
The lower hygroscopicity of the PVA polymer results in an
excessive loss of moisture from the coating which in turn im-
pedes the ammonia ninhydrin reaction.
By plotting the slope of the log transmission vs. time curves
shown in Figure 15, at each known concentration, one can de-
termine any unknown concentration. A plot of slope vs. concen-
tration using PVP/ninhydrin coated collector/sensors is shown
for a humidity of ^80% in Figure 19.
4.7 STABILITY OF COLLECTOR/SENSORS
Stability studies were made in which both PVP (polyvinyl
pyrollidone and PVA (polyvinyl alcohol)/ninhydrin coated quartz
rods were stored in a dark area at constant humidity (60%) and
temperature (25°C). The results of these studies after a one-
week period are shown in Table 12. The changes in transmittance
values are negligibly small compared to the changes encounted
after exposure to even low concentrations of ammonia and are
nearly within the experimental error of the GLAD instrument.
In addition the long term stability studies with PVA/ninhydrin
coated quartz rods was continued and the results are shown in
57
-------
80% RH PVP - Ninhydrin
200 300 400 500
Ammonia Concentration, ppb
600
700
Figure 19. The slope of PVP/ninhydrin collector/sensor
transmittance curves as a function of
ammonia concentration in air
58
-------
TABLE 13. LONG TERM STABILITY OF PVA/NINHYDRIN COATED COLLECTOR SENSORS
Spec. Ho.
186885-3
186885-4
186880-1
186880-4
186894-2
186894-3
186885-5
186885-6
History
Unexposed
Unexposed
Exposed to
1 un
ppm NH3
for 4 hrs.
Exposed to
1 ppn NHj
for 4 hrs.
Unexposed
Unexpossd
Exposed to
1 ppn NH3
for 40 ntin.
Exposed to
1 ppn NH3
for 40 ndn.
Elapsed
Time (t) days
39
71
102
39
71
102
39
71
102
39
71
102
38
70
101
38
70
101
38
7n
/u
101
38
70
101
Transmittance Readings
Initial
ttiite
184
184
184
234
234
234
93.6
93.6
93.6
83.6
83.6
83.6
-
-
-
.
-
-
.
-
.
_
-
700nm
13.9
13.9
13.9
15.4
15.4
15.4
15.6
15.6
15.6
13.9
13.9
13.9
15.9
15.9
15.9
15.6
15.6
15.6
12.4
12.4
12.4
11.5
11.5
11.5
550nm
18.9
18.9
18.9
25.4
25.4
25.4
2.67
2.67
2.67
2.52
2.52
2.52
25.2
25.2
25.2
21.9
21.9
21.9
1
1
1
0.94
0.94
0.94
White
184
155
153
232
194
195
98.1
84.5
87.3
91.6
77.9
82.2
-
-
-
-
-
-
.
—
-
-
_
-
Final
700nm
12.6
9.8
10
15
12.1
12.3
15.2
12.4
12.5
13.8
11.3
11.8
12.3
10.7
10.7
13.5
11.4
12.3
12.1
10.2
10.4
10.9
9.3
9.1
550nm
17.5
15
14.2
23.6
19.7
19.0
2.7
2.5
2.6
2.6
2.4
2.5
18.6
17.9
16.7
18.1
16.5
16.4
1
0.96
0.99
0.97
0.89
0.84
Percent Change
Transniittance
White
0
-15.7
-16.7
- 0.74
-17.3
-16.5
+ 4.8
- 9.7
- 6.7
+ 9.5
- 6.8
- 1.6
-
-
-
-
-
-
.
_
-
-
_
-
VOOnm
- 9.3
-29.5
-28
- 2.6
-21.4
-20.1
- 2.5
-20.5
-19.8
- 0.7
-18.7
-15.1
-22.4
-32.5
-32.5
-13.5
-26.9
-21
-2
-17.4
-15.8
-5.2
-19.1
-20.8
in
(ACT)
SSOnm
- 7.4
-20.6
-24.8
- 7.1
-22.4
-25
+ 2.2
- 6.7
- 2.6
+ 4.3
- 4.8
- 0.79
-26.1
-29
-33.7
-17.
-24.7
-25.1
0
- 4
- 1
+ 3
- 5.3
-10.6
Average 1
Change ii
4%T
At
White
0
-0.221
-0.164
-0.019
-0.244
-0.162
+0.123
-0.137
-0.066
+0.244
-0.096
-0.016
-
-
-
-
-
-
-
_
-
-
.
-
Rate of Percent
i Transirittanoe
' days'
700nm
-0.238
-0.415
-0.275
-0.067
-0.301
-0.197
-0.064
-0.289
-0.194
-0.018
-0.263
-0.148
0.589
-0.464
-0.322
-0.355
-0.384
-0.208
-0.053
-0.249
-0.156
-0.137
-0.273
-0.206
SOOim
-0.190
-0.349
-0.243
-0.183
-0.315
-0.245
-0.056
-0.094
-0.025
+0.110
-0.067
-0.008
-0.687
-0.414
-0.333
-0.450
-0.350
-0.249
0
-0.057
0.010
+0.079
-0.076
-0.105
-------
TABLE 14. LONG TERM STABILITY OF PVA/NINHYDRIN COATED COLLECTOR/SENSORS
Sptc. No.
186885-3
186885-4
186880-1
.186880
18ei£94~2
18(1694-3
18KE85-5
18()E85-6
Specimen
History
Unexposed
Unexposed
Exposed to
1 ppmNH3
for 4 hrs.
Exposed to
1 ppmNH3
for 4 hrs.
Unexposed
Unexposed
Exposed to
1 ppmNH3
for 40 min.
Exposed to
1 ppmNH3
for 40 min.
Elapsed
Time, hrs.
936
1704
936
1704
936
1704
936
1704
912
1680
912
1680
912
1680
912
1680
Transmittanoa Readings
Initial
White
184
184
234
234
93.6
93.6
83.6
83.6
_
-
-
-
5500m
18.9
18.9
25.4
25.4
2.67
2.67
2.52
2.52
25.2
25.2
21.9
21.9
1
1
.94
.94
Final
White
184
155
232
194
98.1
84.5
91.6
77.9
_
-
-
-
5500nra
17.5
15
23.6
19.7
2.7
2.5
2.6
2.4
18.6
17.9
18.1
16.5
1
.96
.97
.89
Percent Change
in Transmittance
White
0.05
-15.7
- 0.74
-17.3
+ 4.80
- 9.7
+ 9.5
- 6.8
- ,
-
:
-
5500im
- 7.4
-20.6
- 7.1
-22.4
+ 2.2
- 6.7
+ 4.3
- 4.8
-26.1
-29
-17.1
-24.5
0 •
- 4
+ 3
- 5.3
62
-------
The observed effects of initial decrease in transmission within
several days after rod coating followed by good stability was
also found to occur for the finalized mixed PVP/PVA polymer
coating developed as the program neared completion. In fact,
during the field testing portion of this program, this coating
formulation was observed to provide enhanced stability over the
PVP alone.
We speculate that the initial decrease in transmission, dis-
cussed above, may be due to the introduction of ammonia/
ammonium compounds from handling and various sources of con-
tamination and/or exposure of the water coating solutions to
ambient concentration of ammonia at the time of coating of the
rods.
Additional stability tests addressed the problem of assessing
the effect of temperature upon stability, particularly in de-
termining whether or not storage of the collector/sensors at
sub-ambient temperatures might improve stability.
All specimens were stored in tightly closed one dram vials to
protect them from the air. For this stability study, the
temperature and humidity were kept constant. In addition to
placing collector/sensors in a dark area at room temperature,
two specimens were placed in a freezer at a temperature of
-16°C. The results of this study are shown in Table 15.
63
-------
TABLE 15. STABILITY COMPARISON OF POLYMER/NINHYDRIH COATED COLLECTOR/
SENSORS STORED AT AMBIENT AND SUB-AMBIENT TEMPERATURES
Spec. No.
190101-4
190101-4
190101-4
190101-5
190101-5
190101-5
190103-4
190103-4
190103-4
190103-5
190103-5
190103-5
Polymer/Reagent
PVP/Hinhydrin
PVP/Ninhydrin
PVP/Hinhydrin
PVP/Hinhydrin
PVP/Hinhydrin
PVP/Hinhydrin
PVP/MLnnydrin
PVF/Ninhydrin
PVP/Ninhydrin
PVP/Ninhydrin
PVJ/Ninhydrin
PVP/fcinnydrin
Tenpera-
ture, °C
25
25
-15
25
25
25
25
25
-16
25
25
25
Tune{t)
days
8
35.25
69.25
8
35.25
69.25
8
35.25
69.25
8
35.25
69.25
Transmittance Readings
'Initial
**iite
239.5
239.5
239.5
224.5
224.5
224.5
281
281
281
333.5
333.5
333.5
550nm
28.7
28.7
28.7
26.8
26.8
26.8
33.6
33.6
33.6
39.5
39.5
39.5
VOOnro
12.6
12.6
12.6
11.2
11.2 .
11.2
12.8
12.8
12.8
15.2
15.2
15.2
white
232
205.5
211.5
211.5
190
186.5
269
264
265.5
319.5
308.5
303.5
Final
550nm
27.3
23.3
23.5
24.8
21.3
20.6
31.8
30.2
30.55
38.0
36.2
35
700nm
12.3
11.5
11.65
10.9
10.2
10.25
12.8
12.8
12.3
14.6
14.5
13.9
At, days
8
27.25
34
8
27.25
34
8
27.25
34
8
27.25
34
•Percent Change in
Transmittance (AST)
white
- 3.1
-11.1
+ 2.5
- S.8
- 9.6
- 1.6
- 4.2
- 1.8
+ 0.5
- 4.2
- 3.3
- 1.5
SSOnm
- 4.9
-13.9
* 0.7
- 7.3
-13.2
- 2.5
- 5.5
- 4.6
* 1.1
- 3.8
- 4.6
- 3.0
700m>
- 2.4
- 6.3
+ 1.2
- 2.7
- 6.2
- 0.4
0
0
- 4
- 3.9
- 0.7
- 3.9
Average Rate of Percent
Change in Trananittance
(4%T/it) ,Vdays)
white
- 0.4
- 0.4
-0.17
- 0.7
- 0.44
- 0.24
- O.S
- 0.2
- 0.08
- 0.5
- 0.2
- 0.13
550m
- 0.6
- O.S
- 0.26
- 0.9
- 0.6
- 0.33
- 0.7
- 0.3
- 0.13
- 0.48
- 0.24
- 0.16
700nra
- 0.3
- 0.3
- 0.11
- 0.33
- 0.3
- 0.12
0
0
- 0.06
- 0.49
- 0.13
- 0.12
Note: All specimens were exposed to ammonia, measurements made'using GLAD II.
•Represents percent changed in transmittance over the specified interval of time, At.
-------
The transmittance values of those speciments stored at the lower
temperature stabilized or increased only slightly whereas the
transmittance values of those specimens stored at room temper-
ature reflected a slight decrease. In either case, however,
only minor changes in transmittance were observed. In some
cases it may be desirable to store collector/sensors at low
temperature when there is a long time interval between the
initial measurement and actual exposure. However, additional
studies should be made to verify this. The A% T/At, at both
temperatures, shows a sharp decrease since the previous measure-
ments, which is a good indication that the transmittance values
have been stabilized for these collector/sensors.
4.8 EFFECT OF ROD LENGTH ON COLLECTOR/SENSOR SENSITIVITY
Exposure studies were made at the 1 ppm level using both 2 and
4 cm long quartz rods coated with PVA/ninhydrin. These studies
showed the response was greater using the 4 cm rods. (See
Figure 20). However, the effects on transmittance of coating
degradation, such as surface scratches and humidity and other
effects will also increase, as a rod length increases, thereby
nullifying the beneficial effects of increasing rod length
(Figure 21 shows this). In addition, the sensitivity using the
2 cm rods appears at this time to be adequate for measuring the
low concentrations of ammonia (~30 ppb) in air. Consequently
the 2 cm in length quartz rods were used during the remainder of
65
-------
100
90
80
I 70
o
o
60
I 50
| 40
z 30
-------
100
90
80
70
60
50
5 40
E
c
8
NO
DC
30
20
10
0
186821-4
100% RELATIVE HUMIDITY
_. 0 ppm NH3
186821-5
100% RELATIVE HUMIDITY
• 1 ppm
0
10 15 20 25 30
TIME, minutes
35
40
45
Figure 21.
The effect of high humidity and/or ammonia
concentration on 4 cm long PVA/ninhydrin coated
rods exposed to 0 & 1 PPM ammonia in an air stream
67
-------
the contract period.
4.9 INFECT OP ACID POLLUTANTS AND AMMONIUM
COMPOUNDS ON COLLECTOR/SENSOR RESPONSE
Using the experimental arrangement shown in Figure 22, speci-
mens ware exposed to either S02 or NO in the presence of low con-
«>
centreitions of ammonia (60 ppb) in diluent air. The results of
the study are shown in Figure 23. At high ppm levels of either
pollutant an effect is observed on collector/sensor response.
In the case of SO'2 the presence of the pollutant appeared to
inhibit color formation. On the other hand an increase in
response is observed with NO . However, at lower S02 or NO
X X
levels (<0.5 ppm) little effect on collector/sensor response
was noted.
Duriruj the initial qualitative screening reactions, we observed
that iiircnonium compounds did indeed react with ninhydrin« How-
ever,, differences in the color hue of the reaction product were
observed between ammonia as such, and ammonium compounds. These
observations are related to the total expected effect of inter-
fera:.its such as SO2, NO , etc. since in addition to direct
X
effects we might expect secondary effects dud to the reaction
with ammonia and formation of ammonium salts and subsequent
reactions.
68
-------
vo
-200 ppm
So or NO
' 1
Compressed Air
0-5 cc/ min.
0 - 500 cc / min.
0 - 500 cc / min.
H20
Sample Chamber
Sample Chamber
^—IHH
Low Emission
Membrane
Kin - Tek Model
570 Gas Calibration
System
-------
E
c
o
ir\
o
CD
O
03
O
I
e
«/>
c
10
CD
Q_
PVP/Ninhydrin-85%R.H.
. B 2.5 ppm S0
0 ppm SO
o
2
5 1 1.5 2.0
Exposure Time, hours
Figure 23. Effects of both acid pollutants and
ammonia on the transmittance of PVP/
ninhydrin coated quartz rods.
70
-------
We would normally expect the direct effects of S02 such as
bleaching, inhibition of color formation, etc. along with the
indirect effects of salt formation to lower the response. At
approximately ten times the ambient concentration of S02 this
is precisely what occurs whereas NO has an opposite unexpected
X
effect in that it increases the response. However, at the norm-
ally encountered atmospheric levels of S02 and NO , there appears
to be a negligible effect from either pollutant at the 30 ppb
level of ammonia.
It is interesting to note that at atmospheric levels of S02 and
NO (ppm) and ammonia (ppb) there was no effect on response even
X
though S02/N0 were mixed in the vapor state and there was
sufficient time for reaction with ammonia. This would seem to
indicate that at the above concentration levels the presence of
whatever concentration of ammonium salts/ammonium salt particu-
late that are formed have a negligible effect on the quanti-
tative accuracy of the method.
71
-------
4.10 EFFECT OP TEMPERATURE ON COLLECTOR/SENSOR RESPONSE
In order fco maintain the PVP/ninhydrin collector/sensors at a
specific temperature, a constant temperature water bath was used.
Constant temperature was obtained by using the experimental
arrangement shown in Figure 24 which made it possible for us to
measure the response below room temperature. Specimens were ex-
posed to 60 ppb NH3 at 11°C and a relative humidity of 85%.
Figure 25 shows the average transmittance as a function of ex-
posure time for these studies. Previously, studies were made
using .similar exposure conditions at a temperature of 25°C and
are also shown in Figure 25. The exposure studies at 11°C re-
sulted in a greater response and higher sensitivity than was
expected. The variation in collector/sensor response at these
T
two temperatures is minimal and could very well be attributed to
quarts; or coating differences or to transmittance measurement
errors. On the basis of these results any minor temperature
variations encountered during collector/sensor exposure studies
can be neglected.
The small effect of temperature on collector/sensor response ob-
serve^ here is contrary to that normally expected. The general
effect cf temperature and the anticipated effect here was dis-
cussed in a previous monthly report. This small temperature
effect is probably due to the fact that this is a gaseous re-
action and that the forward reaction rate constant is very
large thus serving to minimize any temperature effects.
72
-------
0 - 500 cc / min.
Compressed Air
0 - 500 cc / min.
Water Pump
Variac
Constant Temperature
Low Emission Water Bath
Membrane
Ice Bath
Kin - Tek Model
570 Gas Calibration
System
Figure 24. Schematic of apparatus used to expose collector/
sensors to ammonia at various temperatures
-------
£
e
&
if\
v\
o
*<
W
s
<
1=
CO
2:
<
cc
h-
o
ct:
10
11VC, 85 % r. h,
25uC,85%r.h.
.5 1 1.5
EXPOSURE TIME, hrs.
Figure 25. Effect of temperature and relative humidity
on collector/sensors at 60 PPB ammonia level,
74
-------
4.11 ABSORPTION WAVELENGTH
Based on transmission measurements made at various wavelengths
throughout the visible spectrum, a maximum absorption wavelength
of 550 nm has been employed for NHa measurements. These measure-
ments were repeated using collector/sensors that had been coated
with the finalized polymer coating which is a 50-50 mixture of
PVP-PVA, ninhydrin which is buffered with potassium phthlate.
Collector/sensors prepared in this manner and exposed to
ammonia gave the curves shown in Figure 26. These curves show
that the optimum wavelength of absorption is centered about 580
nm. Consequently, this wavelength should be used for measure-
ments when this coating composition is used if maximum sensi-
tivity is desired. The appropriate wavelengths of 550 nm or
580 nm were used for field measurements.
There appears to be an absorption peak near 400-450 nm. We
speculate that this may be the result of small particles present
in the coating of the collector/sensors which would scatter
light at the shorter wavelengths and indicate an absorption
peak as observed. We have observed that less sensitivity to
changes in ammonia concentration and poorer reproducibility will
be experienced if the lower wavelength is used. This is as ex-
pected because of the lower signal to noise ratio. High concen-
trations of NH3 (^50 ppm) and exposure to amines will tend to
shift the wavelength of maximum absorption which should be re-
determined under these conditions.
75
-------
IUU'
WAVELENGTH VS TRANSMISSION
90-
e
ro
C
'en
O
"o
•+-'
eu
o
CD
D_
70
60
50
40
3C
2C
1C
C —
4000
5000 6000
WAVELENGTH, ANGSTROMS
7000
Figure 26.
Absorption characteristics of
reacted collector/sensors
76
-------
SECTION 5
CALIBRATION
Calibration curves were obtained by exposing a number of
collector/sensors from each batch to different ammonia levels
for different exposure times. A typical calibration curve is
shown in Figure 27. The transmittance (T) is defined as the
final measured intensity (I) divided by the initial measured
intensity (I ) for a given exposure time. By plotting the
product of ammonia concentration (c) and exposure time (t) only
a single calibration curve is required. Calibration studies
made at 0, 3, 30, 60, and 200 ppb NHs with the collector/sensors
verifi-d the relationship between Absorbance (log 1/T) and c x t.
This method then allows one to measure very low concentrations
by increasing the exposure time. Using the calibration curve
for an exposure time of 15 hours and a 1/T value of 7.3 the
ammonia concentration would be 2 ppb.
This is the technique used for all subsequent ammonia measure-
ments during this program including those made during the field
testing. It should be noted that the blank response is quite
low compared to the NH3 response
77
-------
00
Calibration Curve for
PVA-PVP Polymer/25% Ninhydrin
Collector/Sensors
Blanks
15 30 45 60
Ammonia Concentration x Exposure Time, ppb - hrs
Figure 27. Effect of both ammonia
concentration and exposure
time on c/s absorbahce
-------
SECTION 6
RESULTS AND DISCUSSION
6.1 AMBIENT AIR MEASUREMENT OF NH3 IN THE LABORATORY USING AN
ARTIFICIAL HUMIDITY CHAMBER (HOLDER) AND DESIGN OF HOLDER
The possibility of utilizing an artificial humidity chamber has
been investigated. A plastic perforated collector/sensor hold-
er (humidor) was designed which houses a set of sponges. These
sponges, soaked in distilled water, are placed in close proximity
to the collector/sensors. A picture of the humidor is shown in
Figure 28. Because th collector/sensor response depends, to an
extent, upon relative humidity, we felt it should be possible to
artificially provide adequate moisture in order to obtain suffic-
ient measurement sensitivity. In-house studies at higher ammonia
levels (0-10 ppm) were recently made which substantiate this.
No significant difference in collector/sensor response was ob-
served by varying relative humidity between 60% and 90% at an
ammonia concentration of ^8 ppm. The humidor alleviates the
necessity of a high relative humidity by providing adequate
moisture to the air immediately surrounding the collective-
sensors. With this artificial source of moisture, the relative
humidity of the surrounding air should be reduced to a secondary
effect which would make it possible for exposure studies to be
conducted over a wide range of humidity conditions.
79
-------
w, -^
X.
Figure 28. Artificial humidity chamber
80
-------
In an effort to measure approximate ambient air ammonia concen-
trations in two separate laboratories, several collector/sensors
were placed in a humidor and exposed to laboratory air. One set
of collector/sensors were exposed in a laboratory with a constant
relative humidity of 51% and a constant temperature of 24°C. A
second set of collector/sensors were placed in an adjacent lab-
oratory with a constant relative humidity of 20% and a temper-
ature of 24°C. The results of these exposure studies are shown
in Table 16. Collector/sensors were exposed for 25 hours in one
case and 6 hours in the second. If we assume the relative hu-
midity near the collector/sensor to be 85% in both cases, the
ammonia concentration can be determined directly from the
appropriate calibration curve. The exposure conditions and
measured ammonia concentrations are also shown in Table 16.
Initial field testing was begun using the holder shown in
Figure 28. In use, the holder was mounted in a High-Vol
Sampler so that it was well-protected from direct wind currents.
This protection was intended to prevent wind currents under
changing conditions of humidity, from depleting the necessary
moisture in the vicinity of the collector/sensors.
Under a separate in-house program, at higher concentrations of
ammonia and large variations in wind velocity through the holder,
we found that the effect of large changes in outside relative
humidity (56-90) was to provide different calibration curves
81
-------
TABLE 1C. EXPOSURE CONDITIONS AND RESULTS OF COLLECTOR/
SENSORS EXPOSED TO AMBIENT AIR IN THE LABORATORY
Spec. No.
Polymer/Reagent
Exposure Time, hrs. ' .
Relative Humidity, %
Temperature , °C
Percent Transmittance ,
550run
Concentration x Time
NHi Concentration, j?j?b
190146-7
PVP/Ninhydrin
25.5
51
24
17
41
1.6
190146-15
PVP/Ninhydrin
6.5
20
24
47
17.5
2.7
82
-------
with the holder shown in Figure 28. In order to minimize the
effects of wind and large changes in relative humidity, the
holder was modified to that shown in Figure 29. This design
utilizes a double screen with 5 mil openings to cover the entire
C/S opening which is the only entry into and out of the holder
and thus prevents the free movement of air past the C/S. A
wetted sponge extends across the area of the entire holder about
1/4" away from the C/S in order to insure the presence of
moisture in the vicinity of the C/S. This holder was found to
provide the same calibration curve for wide ranges of outside
relative humidities thus obviating the necessity for any type of
humidity correction.
6.2 FIELD TEST RESULTS
We felt it to be essential to have available a backup method
that could be used to independently check the results the
presnet GLAD technique under development. The problems assoc-
iated with comparing GLAD results against a concurrent time
integrated impinger sample are:
(1) At the ambient concentration of about 4 yg/m3(5.3 ppb)
the GLAD responds in 4-6 hours whereas 500+ hours would
be necessary for collection using a midget impinger at
a nominal sampling rate in order to obtain sufficient
quantity for analysis by classical methods.
(2) As a consequence of (1) different time periods will be in-
volved making a comparison of the results difficult.
83
-------
Figure 29. Modified GLAD C/S holder
84
-------
(3) The effects of the atmospheric dirt, attendent inter-
ferences, recovery efficiencies and stability of the NH3
species under long sampling times have not been classified.
The previous literature search showed that Okita and Kanomori71
had developed analytical methodology that would provide an in-
dependent analytical check. This method utilizes a sulfuric
acid impregnated glass fiber filter through which a large volume
of air can be passed in a relatively short period of time to
provide a sufficient mass of NHa for analysis. Due to time con-
straints ; the effects of various experimental conditions on re-
sults were not rechecked and conditions as reported by Okita
and Kanomori were utilized to provide the 95% - 98% collection
efficiency that they reported. For the final analytical method,
we used a water extraction of the filter which was subsequently
analyzed by NHs specific ion electrode to provide the NHs con-
centration based upon the quality of air sampled.
Initial field testing of the C/S at the MRC Dayton Laboratory
site began about July 9, 1976 using the polyvinyl pyrrolidone-
ninhydrin coating formulation. An intrinsic objective of this
effort was to determine whether or not correspondence between
the backup analytical technique and the GLAD technique could be
achieved.
71 Okita, T. and Kanomori, S., Atmospheric Environment, 5_,
621-627 (1971).
85
-------
During some of the C/S exposure studies air was drawn through
sulfuric acid impregnated glass fiber filters at a rate of 10-30
1/min in order to quantatively scrub the ammonia from the air.
The NH; or. the filter was subsequently extracted with water and
the yg of NH3/ml was determined with a specific ion electrode.
It was not always possible to span the identical time period in
which the C/S units were exposed but the filter sample results
should be indicative of the ambient atmospheric ammonia concen-
tration in the general time frame of C/S exposure. The results
of the first Dayton site field testing are shown in Table 17.
The humidified effluent from the Kin-Tek permeation device which
was used for calibrating the C/S was passed through an acid
impregnated filter at a rate of 500 cc/min for twenty-two hours
in order to serve as an accuracy check on the membrane permeation
device, An analyzed value of 36.7 ppb vs. a calculated concen-
tration of 30 ppb was obtained.
The collector/sensor holder shown previously in Figure 29 was
positioned under the roof of an EPA Hi-Vol Sampler for protect-
ion from wind and rain as shown in Figure 30. In most cases the
collector/sensors were exposed for 16 hours. Ammonia concen-
trations at the Dayton site over a three-week period were found
to vary from <1 to 15 ppb with an average of 4.2 ppb for thirty-
three separate measurements. Simultaneous parallel measurements
were made in six tests using the method of Okita and Kanomori71.
86
-------
TABLE 17 . ANALYTICAL RESULTS OF 1ST MRC DAYTON LABORATORY FIELD SITE TESTING
00
Sampling
Date
7-9-76
7-9-76
7-13-76
7-13-76
7-22-76
7-22-76
7-26-76
7-26-76
7-28-76
7-29-76
7-29-76
8-2-76
8-2-76
8-3-76
8-3-76
8-3-76
8-4-76
8-4-76
8-5-76
8-6-76
8-6-76
8-6-76
8-9-76
8-12-76
8-16-76
8-16-76
8-17-76
8-17-76
8-18-76
8-19-76
8-19-76
8-24-76
8-24-76
Location
H
F
H
F
H
F
F
H
H
H
F
W
F
W
F
PCC
W
F
PCC
H
F
PCC
F
F
F
F
F
F
F
L
L
F
F
R.B.%
53-62
53-62
49-54
49-54
97
97
40-91
40-91
59-78
67-74
67-74
42-55
42-55
47-70
47-70
47-70
63
63
88
-
-
•noo
-
-
-
-
-
-
-
-
-
-
-
Temp.°P
72-78
72-78
71-77
71-77
79
79
85
85
75-85
83
83
69-73
69-73
61-66
61-66
61-66
60
60
65
69
69
69
-
-
-
-
-
-
-
-
-
-
-
Formulation
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVP
PVA-PVP
PVP
PVP
PVA-PVP
PVA-PVP
PVA-PVP
PVA-PVP
PVP
NH3 ppb
Determined
by GLAD
6.2
6.1
5.7
5.3
7.6
8.0
5.0
14.9
7.3
14.8
5.0
4.1
6.1
5.1
2.1
1.3
2.8 •
2.3
<1
S.2
3.7
<1
<1
6.1
<1
<1
<1
<1
1.1
2.2
1.1
•vl.O
<1
NH3, ppb
Determined
From Filter
_
-
-
6.4
-
12.8
-
-
3.9
-
-
-
3.1
-
-
-
-
3.5
-
-
-'
-
<1
-
-
-
-
-
-
-
-
-
-
-------
Figure 30.
Holder in position under
roof at high volume sampler
88
-------
Sources of error in the filter analyses, particularly differ-
ences in the blank filters, appear to contribute approximately
±0.4 ppb uncertainty. Sulfuric acid impregnated filters held
over approximately the same period of time as those used for the
analyses were utilized as blanks. These analytical values ob-
tained on these were used to correct the value obtained on the
sample filters to correct for any NH3 absorption during the per-
iod of transport and analysis.
Due to the humidity problem previously mentioned, and its effect
on coating tackiness, we developed a second coating formulation
during this part of the program which seems to largely overcome
this problem. This, was accomplished by using a 50-50 mixture of
low molecular weight polyvinyl alcohol (PVA) and polyvinyl pyrro-
lidone (PVP). The PVA retains less moisture at the same rela-
tive humidity and provides the dimensional stability needed while
the PVP comes to its equilibrium concentration of higher moisture
level rapidly as before. Therefore, this combination provides
both sensitivity and coating ruggedness. This coating formula-
tion has also been found to provide improved initial stability
as judged from the decrease in transmission (%2% average) during
the first week after preparation. For these reasons this coat-
ing was used for the final field test at Dayton although the
v
next set of field tests were made with both the PVP and 50% PVA/
50% PVP formulations since previous field test results had been
obtained using this formulation only near the completion of
89
-------
tests and it was therefore necessary to test both formulations
concurrently in order to identify any unanticipated problems.
Because of an EPA study being carried out at a St. Louis EPA
Site do. 108) and the apparent availability of one or more in-
strumental techniques at that time, Mr. J. Mulik, the contract
monitor, suggested that our field tests be carried out at this
location. This portion of the field testing effort was con-
sidered to be an important aspect of the program since it
afforded an opportunity for a cross-check with one or more
independent methods such as, chemiluminescent analyzer, laser
diode infrared, or opto-acoustic infrared analyzer.
Field tests were begun at the EPA St. Louis Site No. 108 on
August 18, 1976 using the holder shown in Figure 29. On the
morning of August 21, 1976 at the St. Louis site, simultaneous
analyses with the GLAD (integrated response) and the EPA-NH3
chemiluminescent analyzer (real time response) were obtained.
It appeared that there was generally good correspondence between
the two techniques except for periods of time during the early
morning hours that plumes came through the area which caused
the chemiluminescent analyzer to register high peak concentra-
tions of NH3. An ammonia concentration of ^1 ppb was measured
with the GLAD which checked the corresponding filter sample re-
sult. It was estimated that these results should have been on
the order of 3-4 ppb in order to agree with the time weighted1
f
90
-------
average results (estimated) of the chemi luminescent analyzer.
The plume presence was confirmed by laser diode and filter
correlation spectrometers which monitored the presence of CO and
NO. Additional NH3 measuring instrumentation was not available
at this time for cross-check analyses.
Dr. W. McClenny of the EPA subsequently found that the NHs chem
iluminescent analyzer results were compromised by the presence
of high concentrations of NO (confirmed by the other spectral
methods mentioned) which was found to interfere with the chem-
iluminescent method and cause the chemi luminescent analyzer to
produce high results in its present configuration.
Generally good agreement was observed with ammonia concentra-
tions ranging from 1.1 ppb to 3.4 ppb with an average of 1.9
ppb using our technique. Measurements made simultaneously with
the parallel method gave an average value of 1.1 ppb ammonia.
Measurements at both sites showed the reproducibility to be
"-±50% as opposed to ±10% observed in the lab. These results
are shown in Table 18.
The concentrations of CO and S02 were tabulated in order to
evaluate possible cross effects on the NH3 concentrations
measured. Although more measurements of all three materials
are required before a conclusion can be reached, we can say
that from the limited data available CO and SO2 do not appear
91
-------
vo
to
TABLE 18 . ANALYTICAL RESULTS OF ST. LOUIS FIELD SITE
TESTING AT EPA REGIONAL AIR POLUTION SITE
#108 IN VICINITY OF GRANITE CITY, ILLINOIS
•
Sampling
Date
8-18-76
8-18-76
8-19-76
8-19-76
8-20-76
8-20-76
8-21-76
8-21-76
8-21-76
8-21-76
Avg.
R . H . f %
74
74
74
72
75
75
90
90
90
90
Avg.
Temp. ,°C
21.4
21.4
22.3
22.3
23.4
23.4
21.0
21.0
21.0
21.0
Formulation
PVA-PVP
PVP
PVA-PVP
PVP
PVA-PVP
PVP
PVA-PVP
PVP
PVA-PVP
PVP
NH3 ppb
Determined
by Glad
3.4
<1
1.1
<1
1.1
<1
<1
<1
<1
<1
NH3 ppb
Determined
From Filter
1.2
1.2
•vl
•vl
'Vl
'Vl
<1
<1
<1
<1
CO,
PPb
50
50
439
439
522
522
-
-
-
—
SO 2
PPb
2.5
2.5
4
4
5
5
2.5
2.5
2.5
2.5
-------
to be either positive or negative interferants to the ammonia
measurements.
Additional measurements were made at the. Dayton site using the
new holder mentioned previously and shown in Figure 29 which was
designed to maintain an essentially constant and high relative
humidity. Also, a modified PVA/PVP polymer-ninhydrin coating
formulation was prepared containing a buffer to control the pH
at * 4.5 during exposure. The addition of the buffer insures
uniform color development and absorption at the same wavelength
from batch to batch in spite of pH variation in the polymer,
distilled water, and/or conditions of use.
Measurements made over a 16 hour exposure period using the
improved holder and buffered coating gave a reproducibility
of ±10%, even though the absolute values now being determined
are about an order of magnitude below the original requested
detection limit. Over an eleven-day period the ammonia con-
centration varied from <0.1 to 0.8 ppb. At these ranges the
absolute values appear to be reasonable checks. Simultaneous
measurements made using the filter collection method gave simi-
lar values. These results are shown in Table 19.
93
-------
TABLE 19. ANALYTICAL RESULTS OF 2ND MRC DAYTON
LABORATORY FIELD SITE TESTING USING
IMPROVED HOLDER AND PVA/PVP COATING
NH3ppb NH3ppb
Sampling Loca- Temp. Determined Determined
Date tion R.H. °F Formulation by Glad From Filter
10/4/76 F
10/5/76 F
10/11/78 F
10/13/76 F
0.6
PVA-PVP 0.9
0.6
PVA-PVP 0 . 3
0.3
0.3
PVA-PVP 0.13
0.3
0.3
0.3
PVA-PVP 0 . 3
Oo9
0.9
ri
<0.3
^0.3
94
-------
SECTION 7
PROCEDURES
7.1 SOLUTION PREPARATION
1. Prepare an 11.8% by weight solution of PVA (Polyvinyl
Alcohol, K & K Laboratories, Inc., Plainview, New York),
using the following procedure: in a flask (500 ml or larger),
heat 150 ml distilled H20 to 95°C. Add 20 g PVA and stir
until all PVA particles are wet (^30 seconds). Place stir-
ring bar in solution and cover flask with aluminum foil.
Immerse solution in a container of hot water and stir "v-2.5
hours with H2O maintained at 100°C. Resultant solution is
clear. Store in a tightly closed container.
2. Prepare an 11.8% by weight solution of Type N-K90 PVP (GAP
Corporation, 140 West 51st Street, New York, New York) using
the following procedure: in a flask (500 ml or larger),
heat 150 ml distilled H20 to 95°C. Add 20 g PVP and stir
until all PVP particles are wet (^30 seconds) . Place stir-
ring bar in solution and cover flask with aluminum foil.
Imnerse solution in a container of hot water and stir •v-l.S
hours with H20 maintained at 100°C. Resultant solution is
clear. Store in a tightly closed container.
95
-------
3. Prepare a 24.8% solution of ninhydrin (ICN Pharmaceuticals
Inc., Life Sciences Group, Cleveland, Ohio) by using the
following procedure: Heat 35 ml distilled water to 85°C.
Remove from heat. Add 8.69 g ninhydrin and stir until com-
plete dissolving occurs.
4. To prepare coating solution, add 75 ml PVP solution to 75
ml '?VA solution and stir until combined. Add ninhydrin
solution and stir. Store resultant solution in tightly
closed container.
7.2 ROD COATING
1. Clean rods by immersing in chromerge solution for 5
minutes. If rods are coated, clean by boiling in H^O for
10 minutes or until coatings have dissolved, then immerse
in chromerge solution for 5 minutes. Rinse twice in dis-
tilled water, once in isopropyl alcohol and end with a
boiling isopropyl alcohol rinse. Remove rods from this
solution with teflon coated tweezers and place on holding
tray to dry. Place tray in desiccator.
2. With teflon coated tweezers, place rods in rod holder.
Tighten each rod with screw driver, being careful that each
rod is secure in holder. However, too much pressure on rod
96
-------
end will cause breakage.
3. Suspend rod holder shown in Figure 6 (with rods securely in
place from Fisher Payne Dip Coater. Immerse rods in sol-
ution while suspended, being careful not to let rod holder
touch solution (^3/4 of rod should be immersed).
4. Remove rods from solution at 5 inches/minute using the Fisher
Payne Dip Coater.
5. Suspend rod holder containing c/s from a lab stand. Place
infrared lamp 5" (12.7 cm) from suspended rod holder and
c/s. A 5" (15.2 cm) diameter aluminum foil reflector is
placed 2" (5 cm) from the assembly opposite the heat source.
A thermometer is placed parallel to the rod holder and c/s.
6. Turn infrared switch to "on" position. When thermometer
reaches 130°F, turn lamp off and leave assembly suspended
for 25 minutes.
7. Loosen screws and use tweezers to rotate each c/s 180°.
Tighten screws.
8. Apply coating following steps 3 through 6.
9. Place dried, prepared, c/s in clean, labeled, tightly
97
-------
closed vials.
10. Whan ready to use c/a, remove from vial carefully with
tweeters and place lengthwise on a lint-free paper towel
that is backed by a flat surface. With tweezers in one hand.
hold c/s steady. With other hand, remove coating from
circumferential ends with a sharp razor being careful not
to iamage ends of rod.
7.3 ANALYTICAL RESULTS
7.3.1 Collector/Sensor Exposure and Calculations
lo With teflon coated tweezers carefully remove unexposed c/s
frosa vial and place in GLAD. Record transmission value at
predetermined wavelengths (white light, 58OOA* for ammonia).
2. Remove c/s, rotate 180°, place in GLAD and record trans-
mission value.
3. CcJ.cuJ.ate the initial transmission value by taking the
average of 1 & 2.
4. At the testing site, place unexposed c/s in humidor using
teflon coated tweezers.
5. After exposure, open humidor at testing site and let c/s
set ~7 minutes before handling.
ii
98
-------
6. Remove c/s from humidor and carefully place in vials until
transmission measurements arc made.
7. Repeat steps 1 & 2.
8. To determine final transmission, take the average of the 2
transmission values recorded in step 7.
9. Determine the % transmission by dividing the final trans-
mission by the initial transmission.
10. The concentration x time value is obtained from a cali-
bration curve obtained in a similar manner using the 1/T
value of the transmission change obtained in step 9.
11. Division by the time of exposure gives concentration
directly.
99
-------
REFERENCES
1. Sii/;ig, Marshall, "Pollution Detection and Measuring
Handbook," Park Ridge, N. J., Noyes Data Corp. (1974)„
2. Ha::dy, E. E., David, D. J., Kapany, N. S. and Unterleitner,
F. C., Nature 257, 666 (1975).
3. David, D. J., Willson, M. C., and Ruffin, D. S., Analytical
Letters, 9(4) , 389 (1976).
4. Halasz, I., Anal. Chem. _36, 1429 (1964).
5. David, D. J., "Gas Chromatographic Detectors", Wiley-Inter-
acience, New York (1974).
6. ".'Development of Instrumentation and Methodology to Measure
Gaseous and Particulate Ammonia", MRC Proposal to E.P.A.
No. P75-3D, 8 January 1975-
7. Makris, K. G., Z. Anal. Chem. 81, 212-4 (1930); C.A. 24,
4731 (1930). —
8. Makris, K. G., Z. Anal. Chem. 84, 241-2 (1931); C.A. 25,
4200 (193D; Welcher, "Organic Analytical Reagents/1 Vol.
II, D. Van Nostrand, New York, N. Y. 1974, p. 172.
9. Castellana, V., Gazz Chim. Ital. 36, 232 (1906); Welcher,
"Organic Analytical Reagents," Vol. II, D. Van Nostrand,
Mew York, N. Y. 1947, P- 173-
10. Korenman, I. M., Z. Anal. Chem. 90, 115-18 (1932); C.A. 2£,
'13 (1933); Welcher, "Organic Analytical Reagents," Vol. II,
I). Van Nostrand, New York, N. Y., 1947, p. 323.
11. Ibid., p. 409.
12. ibid., p. 416.
13. ibid., p. 423.
100
-------
14. Koronin;m, I. M. , ?,. Anal. Chum. 90, ll'j-18 (193?); C.A. 2£,
4} (XOl'Oi Welohor, "Orf/wl is Analytical Roagentr," , Vol. II,
J. Van NoaUrand, Now Yorl<, N. Y., 1947, p. '»38.
15. Ihid. , p. '158.
16. Feigl, P., Mikrochemie \3, 132 (1933); Feigl, P. "Spot Test:
in Inorganic Analysis," 5th Edition, Elsevier Publishing,
New York, N. Y. 1958, p. 237-
17. Zonghelis, C., Compt Rend. 173, 153 (1921); Feigl, P.
Spot Tests in Inorganic Analysis", 5th Edition, Elsevier
Publishing, New York, N. Y. 1?58, p. 238.
18- Tananaeff, N. A. and Budkewitsch, A. A., C.A. 3_0, 5905
(1936); Feigl, P., "Spot Tests in Inorganic Analysis"
Elsevier Publishing, New York, N. Y. 1958, p. 238.
19. Melissa, H., Mikrochem. Ver. Miohokim Aata 38, 386 (1951);
Feigl, P. "Spot Tests in Inorganic Analysis71^ Elsevier
Publishing, New York, N. Y. 1958, p. 239-
20. Jonescu, A. arid Harsovescu, C., Bit. Soc. Chim. Romania. 4^
61-5 (1923); Welcher, "Organic Analytical Reagents" Vol. I,
D. Van Nostrand, New York, N. Y., 1947, p. 104.
21. Foxwell, G. E., Gas World 64, Mo. 1654 (coking section) .10
(1916); C.A. 3^0, I'lB'J, 191i5T Welcher, "Organic Analytical
Reagents", Vol. I, D. Van Nostrand, New York, N. Y. 19*47,
p.
22. Caseneuve, A., Bull. Soc. Pharm. Bordeaux 6l, 153-5 (1923);
C.A. .18, 208 (1924); Welcher, "Organic Analytical Reagents,
Vol. I, D. Van Nostrand, New York, N. Y., 19*17, p. l8l.
23. Lapin, L. and Hein, W. , Z. Anal. Chem. 98, 236-lJQ (193*1);
C.A. 29, 78 (1935); Welcher, "Organic Analytical Reagents"
Vol. I, D. Van Nostrand, New York, N. Y. 19'»7, p. 187.
24. Mangent and Marion, Ann. Chim. Anal. Chim. Appli. jj, 83
(1903); Welcher, "Organic Analytical Reagents", Vol. I,
D. Van Nostrand, New York, N. Y., 1947, p. 240.
25. Duval, C., Compt Rend. 211, 588-90 (1940); C.A. ^6, 365
(1942); Welcher, "Organic Analytical Reagents", Vol. I,
D. Van Nostrand, New York,' N. Y. 1947, p. 376.
26. Schoeller, W. R., Analyst. 57, 750-6 (1932); Welcher,
"Organic Analytical ReagentP1", Vol. II, D. Van Nostrand,
New York, N. Y. 1947, P- 172.
101
-------
27. Rsigler, R., Chem-?,tg. 21, Reperborium 307 (1897); Felgl,
P., "Spot Tests in Inorganic Analysis", 5th Edition,
Elsevier Publishing, New York, N. Y. 1958, p. 235-6.
28. Thomas, P., Bull. Soa. Chim. n, 796-9 (1912); C.A. &,
32*11 (1912); Welcher, "Organic Analytical Reagents'0, Vol.
I0 D. Van Nostrand, New York, N. Y. 19'*7, p.
29. Jtosenthaler, L. Microchemie 13_, 317 (1933); Wenger, P. and
Duckert, R., "Reagents- for Qualitative Inorganic Analysis",
JHsevier Publishing Co., New York, N. Y. 1948, p. 245-
30. Monger, P. arid Duckert, R., "Reagents for Qualitative Inor-
ganic Analysis", Elsevier Publishing Co., New York, N. Y.,
1948, p. 248.
31. Srismer, R., Bull. Soc. Chim. Biol. 1£, 1000 (1937).
32. flolleter, W., Bushman, C., and Tidwell, P., Anal. Chem.
33(4), 592-591* (1961).
33. V/illiams, D. and Miller, R. , Anal. Chem. 34(2), 225-227
(1962).
34. Henesch, R. , Helgolaer.der Wiss Meersunters 23(3), 365-75
(1972); C.A. 78, 33725d (1973).
35. Bulusek, Jaroslau, Sb. Ved. PP. Vys. Sk. Chemickotechnol
Paraducbice 30, . 1*11-59 (1973); C.A. 82., 382780 (1975)
36. Steinbrecher, K., J. Aasoo. Off. Anal. Chem. 56(3), 598-601
(1973); C.A. T9_, 4ioo5e (1973).
37. Hegedus-Wein, Lldiko and Szam, Instran, Orv Hetil 314(4'-;),
2614-15 (1973); C.A. 80, l42629n (1974).
38. Vchida, Tadashi, Eisei Kinsa 23(11), 868-70 (1974): C.A. 82,
12155f (1975).
39o Matsunaga, Katsuhiko, Anal. Chim. Aota 73(l)a 204-8 (19/4)-
C.A. ^2, 642l8c (1975).
40. Yamaguchi, Ryoji, Hoshi Yakka Daigaku Kiyo 14, 98-102 O972V
C.A'. 78, I46982d (1973). ~
41. Hegedus-Vfein, Ildiko, Ber. Ges. Inn. Med. D.D.R. 8, 241-2
(1972); C.A. 7£, 29l82c (1973).
42. Letkina, N. P. and Gurevich, T., Sh. Zdravookhr Belaruss
19(1), 32-4 (1973); C.A. 7i, 226lr (1973).
102
-------
43. Collins, P. P., Beitr. Tabdkforsoh 6(4), 167-72 (1972);
C.A. £8, 133602e (1973).
44. Suzuki, Kazuo, Chomi Kagaku 19(10), 23-6 (1972); C.A. 80,
45. Naito, Mariko and Vete, Tetsuo, Kitano Byoin Kiyo, 16(3-4),
102-6 (1971); C.A. 77, 72204g (1972).
46. Hasashiro, Hideto, Rinshokensa 18(6), 655-8 (197*0; C.A. £2
I3307y (1975).
47. Ozkan, Kemal, Ankard Univ. Tip. Fak, Mecm. 25(5) > 1011-22
(1972); C.A. 79, 63318m (1973).
48. Hegedus-Wein, I., Med Welt, 2^ (31-32), 1205-7 (1973);
C.A. I£, 102263m (1973).
49. Nikolin, B., Tab Sci. l8> 102-6 (197*0; C.A. 80, 130645b
(1974).
50. Okita, T. and Kanamori, I., Atmos. Environ. 5(8) , 621-7
(1971); C.A. 75, 112608 (1971).
51. Truesdale, V. , Analyst 9_6 (1145), 584-90 (1971); C.A. 7.5,
(1971).
52. Chapman, B., Cooke, G. H., V/hitehead, R., Water Pollution
,Contr. 66(2), 185-9 (1969); C.A. 71, 9*»633?. (1969).
53. Kramer, D. , Seen, J., U. S. Government Research and Dowloi.'-
ment Dopt., AD No. 713550, 12 pgs (1969); C.A. 75, l'i->S.lb
(1971).
54. Rosenthaler, L. , Z. Anal. Chen. 183, 193-** (1961); C.A. 59,
287'li (I960). "~" " ~^~
55. V/eller, H. , Roentgen-Lab Praxis ^5, L77-L85 (1962); C.A. 5T_,
10135 (1957).
56. Magill, Paul L., Am. Ind. Hyg. Asaoo. Quart. 11, 55-63 (1950);
C.A. iii, 26l2f (1951).
57. Lugg, G. A. and Wright, A.S., Australia, Dept. Supply,
Defence Research Labs., Circ. No. 14, 1-48 (1953); C.A. *»7,
9857e (1953). ~~
58. Vasilchenko, L. A. and Drozdova, V. M. , Fr. Gl. Geofiz.
Observ. No. 141, 99-103 (1963); C.A. 6l, 1260g (1964).
103
-------
59. Whitaker, J, R., 0t al.t Clin. Chirn. Acta 12(4), 466-8
(19<55); C.A. 63, 15212e (1965).
60,, Levitskl, A., Anal Biochem. 33(2), 335-40 (1970); C.A.
87070f (1970)o
61. Muramatsu, K. , Agr. Biol* Chem. 3K3)> 301-8 (1967); C.A.
67, I8468b (1967).
62. Newell, B. S., J. Mar. Biol. Assoc. U. K. 47(2), 271-80
(1!)S7); .'C.A. 67, 85722f (1967).
63, Kudeyarev, U. N., Agrokhimiya 4_, 129-31*, (1966) j C.A. 65S
(1966).
64. Sojecki, Wojclech, Pr. Cent. Inst. Ochr. Pr. 17. (54), 213-
23 (1967); C.A. 68, 599in (1968).
65. Zv«rev, Yus9 et al., Uses. Nauch.-Issled. Inst. Okhr. Tr.
2(54-74 (1967); C.A. 7_p_, 99375f (1969).
66. .Karc^atowa, G., Rocs. Panstw. Zakl. Hig. 22(3), 289-94
(1971); C.A. J5, 9!080p (1971).
67. Someyas Takeshi, Oshawa, Shigeki, Nippon Kagaku Kaishi (4),
7';4~747S 1972; C.A. 77, 428l3n (1972).
68. Matsunaga, Katsuhiko, Bunseki Kagaku 20(8), 993-7, (1971);
Co A. 16, 131258g (1972).
69. Janowski, Tomasz M. , Zychlinskis Jozef, Hed Wet. 25(10",
617-20, (1969); C.A. T2S 124794t (1970).
70. Deckert, W., Z. Anal. Chem. 150 » 421-5, (1956); C.A. 50,
10607 (1956).
71. Ckita, T. and Kanamori, S, "Atmospheric Environment", 5^f
121-627 (1971).
104
------- |