CHAPTER XVII:

LIGHT-SCATTERING AND MOLECULAR

SPECTROPHOTOMETRY
A. Introduction
            Light or electromagnetic radiation can be characterized by wave-like behavior.  The
wavelength, l, is the distance between the crest of one wave and the crest of the next one.  The
frequency, n, is the number of complete waves that pass a reference point in a second.  These
two are related by the velocity of light, c (3 x 10 cm/sec), in accordance with equation 17.1.
10

                                                                                                             (17.1)

Alternatively, light may also exhibit behavior resembling that of a particle.  These particles,
called photons, have a unique energy, E, that is dependent on the frequency (equation 17.2)
through Planck's Constant, h.

                                                                                                            (17.2)

As a result of equations 17.1 and 17.2, the energy of a photon of light is inversely proportional to
its wavelength.

The human eye perceives light of different wavelengths as being comprised of separate colors. 
When a substance absorbs light of a certain wavelength (color), it depletes the transmitted light
in precisely those wavelengths that are absorbed.  For this reason, a solution that appears blue to
the human eye does not absorb blue light.  Rather, it is blue because it absorbed orange light, and
allows all other wavelengths to pass unhindered (giving a bluish color to the transmitted light;
see Table 17.1).
 Table 17.1

Relationship Between Color and Absorbance

Wavelength of absorbance Color Absorbed Color Remaining

maximum (nm)
380-420 Violet Green-yellow
420-440 Violet-blue Yellow
440-470 Blue Orange
470-500 Blue-green Red
500-520 Green Purple
520-550 Yellow-green Violet
550-580 Yellow Violet-blue
580-620 Orange Blue
620-680 Red Blue-green
680-780 Purple Green

            Double bonds are strong absorbers of light, especially when they are arranged in an
alternating pattern with carbon-carbon single bonds.  This alternating pattern of single bond
followed by double bond followed by single bond followed by double bond is called a
conjugated double bond system.  It has a property known as resonance, which makes it absorb
light of high wavelengths (i.e., low energy).  The molar absorptivity and wavelength of maximim
absorbance generally increases with increasing numbers of conjugated double bonds.  This is
illustrated by the series of polynuclear aromatic hydrocarbons, below.
            When a solution is placed in a path of light, a certain amount of that light will pass
through it unaffected.  The fraction of the incident light that reaches the other side of a liquid
sample will be dependent on the pathlength through the sample, x, and the ability of the sample
components to block or otherwise impede the passage of light, g, called the extinction
coefficient.  This relationship is given by Lambert's Law:

                                                                                         
                                               (17.3)

or

                                                                                            
                                               (17.4)

where:

            Io = the intensity or radiant power of the incident light

I = the intensity or radiant power of the light after passing through a though a thickness,
x, of the liquid sample

The blockage or impedance of light may be due to scattering or absorption.  Light scattering
occurs when small particles deflect the light so that it does not reach the other side of the
sample.  Absorbance is the process by which a constituent (usually dissolved) absorbs the light
energy and releases it as heat or stores it as bond energy.  Thus, the extinction coefficient in
equation 17.3 and 17.4 is composed of two components, the scattering coefficient (t) and the
absorption coefficient (k).

                                                                                                           (17.5)

B. Light Scattering
            When light scattering is large compared to absorption, (i.e., t >> k), Lambert's Law
reduces to:

                                                                                          
                                                    (17.6)

The scattering coefficient may be expressed as the product of the concentration of particles, N,
and the scattering cross section of those particles, s.

                                                                                                             (17.7)

The scattering cross section is a measure of the amount of light scattered per particle.  It is a
function of the size and shape of the particle.  For particles of extremely small size (i.e., a
diameter 10% of the incident light's wavelength), Rayleigh Scattering predominates.  The
Rayleigh Scattering cross section for an ideal spherical "particle" is given by equation 17.8.

                                                                                               (17.8)

Thus, the Rayleigh Scattering depends only on the pathlength, the concentration of "particles",
the wavelength of the light, l, and the polarizability of the light, a.  Equation 17.8 predicts that
the smaller wavelength light will be scattered the most.  Therefore, blue light should be scattered
more than the other visible wavelengths.  As a result, a clear daytime sky appears blue when
looking at any direction except directly into the sun.

            Larger particles are subject to Mie Scattering.  With these large particles, light may
scatter all over their surfaces at different angles.  Mie Scattering is the type of light scattering
that is responsible for turbidity in natural waters.  It is far more complex than Rayleigh
Scattering, and cannot be described by a simple equation.  The Mie Scattering cross section is a
function of the particle size and shape as well as the wavelength of light.  Maximum cross
section for a given wavelength occurs when the particles present are of a size that is similar to
the wavelength.  Less scattering occurs as the difference between the particle size and
wavelength increases.

1. TURBIDITY

a. Environmental Significance

            Historically, turbidity is considered a measure of the attenuation of light through a sample
of water.  However, modern turbidimeters are based on nephelometry, which is the measurement
of true light scattering.  For this reason, we will use the terms turbidimetry and nephelometry
interchangeably to indicate the measurement of light scattering by small particles in a water. 
These particles may be composed of clays, bacteria, algae, or colloidal organic molecules. 
Accordingly, the sizes of these particles may range over several orders of magnitude.  Turbidity-
producing particles are of concern in drinking water supplies for a variety of reasons.  Cloudy or
turbid waters are aesthetically unpleasing.  Thus, consumers may be tempted to forgo an
otherwise acceptable disinfected drinking water for one that is not disinfected, but clear.  Turbid
waters may be less filtrable as a result of the high load of particles applied to the filters.  This
would obligate shorter filter runs and add to the expense of the finished water.   Finally,
turbidity-producing particles may serve to protect micro-organisms (e.g., pathogens) and render
disinfection processes less effective.

            Turbidity measurements are used to assess the quality of water supplies.  Knowing raw
water turbidities, an engineer can suggest a treatment sequence that is likely to be economically
and technologically feasible for any particular source (e.g., conventional treatment vs direct
filtration vs in-line filtration; see Weisner et al., 1988).  A knowledge of turbidity may also help
the engineer to estimate unit sizes and chemical costs (e.g., for coagulation and sludge
handling).  However, pilot studies must often be performed before a treatment plant is designed;
and here again, turbidity measurements are used to assess the effectiveness of most treatments
processes one would test.  Finally, turbidities are monitored in full-scale treatment installations
both for the purpose of process control and for the purpose of establishing compliance with
finished water turbidity standards.

b. Measurement: Turbidimetry

            Modern methods for determining turbidity employ nephelometric turbidimeters which
measure the intensity of light scattered at 90°±30° with respect to the incident light.  The
detector should be sensitive to light in the range of 400-600 nm wavelength.  This measurement
is then compared to an identical measurement of an accepted turbidity standard.  The standards
have pre-designated turbidities expressed in nephelometric turbidity units (NTU).  Accepted
standard substances include formazin, a co-polymer of hydrazine sulfate and
hexamethylenetetramine, and commercially-available polystyrene divinylbenzene beads (AEPA-
1).

            In the UMass Environmental Engineering Teaching Laboratory we have a Hach Model
2100A laboratory turbidimeter.  Light from a tungsten filament lamp is projected up through a
flat-bottom sample cell.  A 90° slit allows scattered light to reach the photomultiplier tube.  The
photomultiplier is especially sensitive to the ultraviolet light scattered by small particles (i.e., 0.1
to 0.5 microns). The 2100A has a rated bias and precision of -2% of full scale on all ranges. 
Thus, it is always advisable to use the highest range while still being able to remain on scale. 
The ranges are 0-0.2, 0-1.0, 0-10, 0-100 and 0-1000 NTU; and the detection limit is 0.04 NTU
on the lowest range.

            It is not uncommon that the same sample, measured on two different nephelometric
instruments may give different turbidity readings.  This can occur even though the two
instruments may be calibrated against the same standards.  Differences in instrument design,
bulb age and output voltage may account for this.  For example, if the voltage applied to the bulb
drops with time, the tungsten filament will become cooler and the output spectrum will shift
toward longer wavelengths.  Since larger particles are better at scattering light of longer
wavelength, this instrument will become more sensitive to larger particles and less sensitive to
smaller ones.  A formazin standard which has relatively large particles allows one to calibrate the
instrument to the larger particles.  However, the instrument will no longer be sufficiently
responsive to the smaller particles in a water sample.  One way of avoiding this problem is
calibrate against two different types of standard materials (e.g., formazin and AEPA-1; these
have a diameter range of 1.75-20mm and 0.2-0.8mm, respectively) and adjust the bulb voltage (or
replace the bulb, if necessary) until the instument can be calibrated so they both give correct
readings.  Another source of error is the presence of stray light reaching the photo sensor.  Stray
light may be determined using a methanol standard.

C. Molecular Spectrophotometry
1. THEORY

            When absorption is large compared to light scatter, (i.e., k >> t), Lambert's Law reduces
to Beer's Law:

                                                                                   
                                                      (17.9)

The absorption coefficient may be expressed as the product of the concentration of absorbing
substances, c, times their molar absorptivity, a.

                                                                                                      (17.10)

Therefore, the transmittance (T = I/Io) of a liquid sample is given by equation 17.11.

                                                                             
                                               (17.11)

A more convenient measure of the absorption of a liquid sample is the logarithm of the
reciprocal of the transmittance.  This value, called the absorbance (A), is linearly related to the
concentration of the absorbing substance.

                                                 A  =  -log(T)                                                   (17.12)

                                                 A  =  acx                                                        (17.13)

Absorbances may also be presented as absorbance units per pathlength, or simply reciprocal
pathlength (e.g., as cm 1).  We will symbolize this representation of absorbance with Abs.

                                               Abs  =  A/x  =  ac                                             (17.14)

Equation 17.14 predicts that a plot of Abs versus concentration will give a straight line with a
slope of "a".  The absorptivity is commonly expressed in terms of absorbance units per mole/L. 
This value is called the molar absorptivity.

2. MEASUREMENT

a. Spectrophotometers

            Sample absorbances are measured with a spectrophotometer.  This is an instrument

composed of a light source, a wavelength selector (monochromator), a sample compartment, and
a detector.  The type of light source will depend on the desired wavelength.  Light in the visible
range (350nm-800nm) is best supplied by a tungsten filament lamp.  Ultraviolet (UV) light
(190nm-350nm) requires a deuterium lamp.  The wavelength selector splits the light spectrum
into its component colors and selects a narrow band of this spectrum.  Either a prism or a grating
may be used as a dispersing element to split the light.  The spectrum is generally projected on an
opaque wall containing a slit.  The physical width of the slit determines the band width of the
selected light.  The wavelength of the selected light may be adjusted by changing the angle of the
dispersing element, so that the desired wavelength passes through the slit.  The selected light
passes into the sample compartment, through the sample (and cell) and to the detector
(photomultiplier).  By measuring transmittance with (I) and without (Io) the sample present,
absorbance can be determined.  This is how a single beam spectrophotometer operates.

            A more accurate and convenient scheme is embodied in the double beam
spectrophotometer.  This instrument has a chopper motor which alternately deflects the light
beam through a reference cell and the sample cell.  This is done many times per second, and the
average ratio between the two readings gives the transmittance.  This is a more accurate method,
because it minimizes variabilities due to rapidly changing lamp output or momentary stray light
in the detector compartment.  In effect, the analyst is continually monitoring and adjusting for
changes in the lamp output as measured by the detector.

            In the Environmental Engineering teaching laboratory we have several

spectrophotometers.  One is a Perkin-Elmer Model 111 Ultraviolet-Visible Spectrophotometer. 
This is a typical single beam instrument with a spectral range of 200-900 nm.  It uses a
diffraction grating with fixed bandpass of 2.0 nm.  The wavelength accuracy is ±0.5 nm and the
precision is ±0.1 nm.  The photometric linearity is -0.005 absorbance units at a reading of 0.400
absorbance units.  The photometric reproducibility is better than 0.01% Transmittance, and stray
light is less than 0.5% T at 220 nm.  This instrument also has a stability characterized by a drift
of less than 0.5% T per 5 min after a 15 min warm-up period.  As with most UV-Vis
spectrophotometers, the PE 111 uses a Tungsten lamp in the visible region (900-340 nm) and a
Deuterium lamp in the ultraviolet region (340-200 nm).

            For best operation, spectrophotometers should be installed in air conditioned rooms, free
from dust, corrosive fumes, vibrations, and large changes in temperature and humidity. 
Instructions for calibrating and operating the PE 111 are as follows:
 1. With the operation switch in the off position, verify that the meter mechanical zero is
correct.

2. Select the proper lamp for the desired wavelength by adjusting the selector lever to
either VISIBLE or ULTRAVIOLET.

            3. Adjust the wavelength knob to the correct value.

4. Turn on the appropriate lamp.  If the deuterium lamp is chosen, the #1 switch must be
turned on first, then 20-30 sec later the #2 switch is turned on.

5. Turn the operation switch to "on".

6. Open the cell compartment and insert both filled sample and reference cells in the
cell holder.  The reference cell belongs in the position #1.

7. Place the operation switch to the "meter" position.

8. With the cell compartment cover open (this automatically closes the shutter), adjust
the meter to infinite absorbance (0 transmittance) with the zero adjusting knob.

9. Close the cell compartment cover, and place the reference cell in the light path by
adjusting the cell positioning knob.  Adjust the meter to 0 absorbance (100%
transmittance) with the zero adjusting knob.

10. Pull out the cell positioning knob and read the absorbance of the sample.

Also in the environmental engineering laboratory we have a Spectronic 21 capable of

measurement in the UV as well as visible range.  This is also a single beam instrument that
requires adjustment to 100% transmittance (or zero absorbance) prior to use.
 Figure 17.1  Controls on the Spectronic 21D UV-Vis Spectrophotometer

b. Spectrophotometric cells

            Cells (or cuvettes) are supplied with varying light paths and different qualities of glass. 
Most spectrophotometric work is conducted with standard rectangular cells of 1 cm path length. 
These are square in cross-section and about 4 times as high as they are wide.  They generally
have a pathlength tolerance of -0.01 mm.  In addition, rectangular cells of 0.1, 0.5 and 4 cm are
commercially available.  For more dilute solutions, cylindrical cells of 5 and 10 cm are
available.  These have two filler necks into which fit fluoropolymer stoppers.  The ability to seal
these cells is an attractive feature for volatile or hazardous samples.  The cylindrical cells may be
supplied with fluoropolymer covers, but these do not make a very good seal.  Rectangular cells
with extra thick side walls are also available for the analysis of small volumes of liquid.  In
addition, various types of flow through and jacketed cells are available for kinetic studies,
continuous monitoring, and temperature sensitive work.

            Cells for absorption spectrophotometry have opposing optically polished windows with
frosted glass side walls.  Be sure to orient the cells so that the optic windows are perpendicular to
the light path.  Fluorescence cells are made with all (or adjacent) optically polished windows.

            Cells should be cleaned like other laboratory glassware.  If detergents prove ineffective,
they may be soaked in chromic acid cleaning solution.  Note that the windows and side walls are
fused forming an acid-proof seal.  In addition, quartz cells may also be cleaned by immersing in
concentrated nitric acid followed by ultrasonic agitation or 10-15 minutes of boiling in water.
 Table 17.2

Spectrophotometric Cells

Window Wavelength Letter Lot #s Color

Range (nm) Code Code
Optical Glass 360 - 1000 OG yellow
Near-UV Glass or 300 - 1000 OS or 180's green
Special Optical Glass SG
Standard Silica 220 - 2500
Supracil Quartz or 165 - 2600, QS or 280's blue
Quartz UV 2850 - 3600 UV
Infracil Quartz or 220 - 3600 QI or 300's red
Quartz IR IR

            Most spectrophotometric cells sold in the US are marked with a series of letters and
numbers.  Many are marked with and upper case "SCC" (Scientific Cell Company) and number
indicating the path length (e.g., 1.000 for 1 cm) followed by a two letter code, a colored dot or a
lot number.  These markings will identify the type of glass used in the optical windows (see table
below).  Cells should always be matched with the same markings including lot number.  This is
especially important when running absorbance scans.  It is of highest importance to choose a cell
which is useable for the wavelength chosen.  Although any of the cells listed below may be used
in the visible range, the optical glass cells are preferred because of their lower cost.  For near-UV
and UV work, near-UV glass, Supracil quartz or Infracil quartz should be employed depending
on the wavelength chosen.  In general, the lower the minimum wavelength, the more expensive
the cell.  For work in the infrared region, Infracil quartz is required.
3. ANALYTICAL MOLECULAR SPECTROPHOTOMETRY

a. Direct Methods

            Most specific chemical analytes do not have large enough absorptivity to make their
determinations by direct spectrophotometry practical.  However, their molar absorptivities may
be substantially increased by reaction with color-forming reagents.  This is the basis for most
colorimetric or spectrophotometric methods.  The two exceptions common to environmental
engineering are the determination of "color" and "UV absorbing substances".  However, color
and UV absorbance are somewhat unique in that they are gross parameters which are
operationally defined.  They are both commonly used to assess the concentration of natural
organic matter (e.g., humic substances) in a water.  Humic substances absorb strongly over a
wide range of wavelengths (see section b).  Absorbance measurements will depend on the
concentration of the humic substances, the flora and fauna from which the humics were derived,
and the solution pH.  This last factor is particularly important with respect to the analysis of UV
absorbance and color.  As pH increases, UV absorbance and particularly, color increase.  This is
likely due to deprotonation of acidic sites on the humic molecules, causing an unfolding of the
molecules and increasing the multiplicity of resonance structures.  Regardless of the mechanism,
it is important that a water be buffered at a fixed pH (usually 7) prior to analysis for UV
absorbance and color.

            Color.  Natural waters derive color from humic materials and sometimes from metals
such as iron and manganese.  This is a common means of characterizing a natural water's organic
content.  Color is very easy to measure, requires only the simplest glassware, and correlates well
with DOC or TOC in many water treatment systems.  Color is also important because it is a
direct measurement of a water's disagreeable visual properties.  As a result many treatment
systems are designed to remove color whereas the specific goal of removing organic carbon is
secondary.

            Two types of color are often reported, "true color" and "apparent color".  True color is
that which is attributed to dissolved species, and it is therefore, the color remaining after a
sample is filtered.  Apparent color is derived from the absorbance of dissolved species and the
light scattering of particles.  It is, therefore, measured without prior filtration.  The actual
determination of color is done by visual comparison or with the aid of a spectrophotometer.

            The platinum-cobalt method of visual comparison is the classic procedure used for
decades.  Color standards are prepared by mixing predetermined amounts of potassium
chloroplatinate and cobaltous chloride in water, and diluting this stock 10 to 100 fold.  The
prescribed ratio of cobalt to platinum gives a yellowish hue characteristic of natural waters. 
Filtered or unfiltered samples are pH-buffered and placed in 50 mL Nessler tubes.  Platinum-
cobalt standards are added to similar tubes.  Concentrations are in Pt-Co color units.  These
correspond numerically to the concentration of platinum in mg/L in the standard which most
closely resembles the water.  Color analysis in the field may be more conveniently conducted
with pre-calibrated glass disks.
            Color may also be conveniently and precisely measured using a spectrophotometer.  The
range of wavelengths chosen run from 400 nm to 700 nm, with 400 nm being most common.  As
before, the samples are buffered at pH 7 and pre-filtered if desired.  The color value is simply the
absorbance of the sample at the designated wavelength.

            UV Absorbance.  Most natural organic matter will absorb sufficient ultraviolet (UV) light
to be easily detected by a standard UV-Vis spectrophotometer.  By convention, we have chosen
254 nm as the wavelength to measure UV absorbance.  This parameter is quite important
because: (1) it is inexpensive, rapidly measured, and requires a minimum of training; and (2) it
has been found to correlate with certain water quality characteristics, such as DOC and THMFP.

            UV absorbance has been successfully used as a means of estimating DOC and THM
precursor levels in raw waters.  However, its most important contribution is to process
monitoring.  For a single raw water source, coagulation effectiveness can be effectively
monitored by UV absorbance.  One can generally develop good linear correlations between UV
abs and DOC for raw and treated waters from the same plant (e.g., Edzwald et al., 1985).  The
interpretation changes, however, when a disinfection or oxidation step is encountered.  When
monitored across oxidation/disinfection, UV absorbance provides information on the degree of
oxidation of the natural organic matter in the water.

            The specific absorbance, which corresponds to the absorbance per mg/L of DOC, is a
useful tool for rapidly assessing the "humic/non-humic nature" of a water.  Some specific
absorbances for extracted humic and non-humic fractions are shown in Figure 17.2.  Note that
humic acid and fulvic acid show the highest SUVA (6.3 and 4.4, respectively).  In another study,
averages of 10 aquatic humic substances showed SUVA values of 5.8 for the humic acids and 3.6
for the fulvics (Reckhow et al., 1990).  Other fractions, especially the hydrophilic acids, show
lower SUVA values.  For this reason, waters with a high SUVA generally have higher humic
contents, and are more amenable to DOC removal by coagulation.

            Table 17.3 summarizes some attempts to correlate UV absorbance (254 nm) to DOC for
raw waters.  Note that the reciprocal of the slopes in Table 17.3 correspond to the specific
absorbances in Figure 17.2, and that the average slope (~25) gives a specific absorbance of
0.004/cm or 4/m which is similar to those reported for fulvic acids.
 Figure 17.2

Specific UV Absorbance of Eight NOM Fractions from Forge Pond

(After Reckhow et al., 1993)

Table 17.3

UV Absorbance (254 nm) to DOC Regression Equations+

Water Slope Intercept r Reference

Tjeukemeer Lake 24.7* 2.7 0.925 De Haan et al., 1982
Grasse River 19.8 0.65 0.93 Edzwald et al., 1985
Glenmore Reservoir 28.1 -0.3 0.71 Edzwald et al., 1985
24.1 2.2 0.96 Smart et al., 1976
N.C. Surface Wat. 29.1 1.44 0.83 Singer et al., 1982
Norwegian Lakes 19.1 2.3 0.93 Vik et al., 1985

+
Model:  DOC (mg/l) = (slope)*(UV absorbance, cm-1 ) + intercept

* Based on absorbance at 250 nm, neutral pH.

b. Methods requiring the Formation of a Chromophore

            As discussed obove, most specific chemical analytes do not have large enough
absorptivities to make their determinations by direct spectrophotometry practical.  However,
through a variety of means, one can increase their molar absorptivities.  This is most often done
through the formation of a chromophore, a chemical structure that imparts a visible color to a
solution.  By far the most common group of reactions that is used to form chromophores for
spectrophotometric analysis are the complexation reactions (see Table 17.4).  Often extractions
into organic solvents or redox reactions are needed to improve selectivity (i.e., remove
interferences) or sensitivity (i.e., lower detection limit).  Other colorimetric methods are based
largely on redox reactions (see Table 17.5).  This is especially true of analytes that are,
themselves, reductants or oxidants (e.g., chlorine).  Another class of colorimetric methods relies
on the catalysis of an observeable reaction by the analyte (Table 17.6).  Here the rate of reaction
must be directly proportional to the analyte (catalyst) concentration.  These are potentially very
sensitive methods.  Finally, Table 17.7 lists a number of methods that employ coupling reactions,
simple substitutions, or incompletely characterized reactions.
 Table 17.4

Colorimetric Methods Based on the Formation of a Complex

Analyte Complexing Metal or Additional Treatment Wavelength

Ligand
Al Eriochrome cyanine R dye 535 nm
Be Aluminon (aurintricarboxylic 515 nm
acid ammonium salt)
Cd dithizone post-extraction (CHCl3) 518 nm
(diphenylthiocarbazone)
Cr(VI) diphenylcarbazide 540 nm
Cu neocuproine (2,9,-dimethyl-1,10- post-extraction(CHCl3& 457 nm
phenanthroline) CH3OH)
Cu bathocuproine disulfonate 484 nm
Fe 1,10-phenanthroline pre-reduction
(hydroxylamine)
Pb dithizone pre-extraction (CCl4) 520 nm
Hg dithizone pre-extraction (CHCl3) 490 nm
Ni dimethylglyoxime pre-extraction (CHCl3,then 445 nm
H2O)
Se diaminobenzidine pre-oxidation, pre-reduction 420 nm
post-extraction (C6H5CH3)
Zn dithizone pre-extraction (CCl4) 535/620 nm
Zn zincon 620 nm
As silver diethyldithiocarbamate pre-reduction (Zn) 535 nm
B curcumin 540 nm
B carmine 585 nm
SCN Fe(+III) 460 nm
F SPADNS reaction (zirconium) 570 nm
PO4 Mo post-reduction (SnCl2) 690 nm
PO4 Mo post-reduction (ascorbic 880 nm
acid)
Silica Mo 410 nm
SO4 methylythymol Blue pre-precipitation (Ba), 460 nm
chelation of excess Ba &
detn of excess ligand
 Table 17.5

Colorimetric Methods Based on Redox Processes

Analyte Oxidant or Reductant Additional Treatment Wavelength

Mn persulfate or periodate 525 nm
(to Mn(+VII))
K dichromate pre-precipitation (sodium 425 nm
cobaltinitrite) & oxdn of
excess
Br(-I) chloramine-T (to Br(+1)) post-reaction (bromination 590 nm
of phenol red)
Cl(+I) & ClO2 DPD 515 nm
CN chloramine-T (to CNCl) post-reaction (pyridine- 578 nm
barbituric acid)
I(-I,+I) Leuco Crystal Violet pre-oxidation (potassium 592 nm
peroxymonosulfate)
NH3 HOCl post-reaction (phenol) (to 630 nm
indophenol, phenate
method)
O3 Indigo Trisulfonate 600 nm
Tannins & tungstophosphoric & 700 nm
Lignins molybdophosphoric
acids

Table 17.6

Colorimetric Methods Based on Catalysis

Analyte Reaction Additional Treatment Wavelength

V oxidation of gallic acid 415 nm
by persulfate
I(-I) reduction of Ce(+IV) remaining Ce(+IV) reduced by 410 nm
by arsenious acid Fe(+II), residual measured at

Table 17.7

Colorimetric Methods Based on Miscellaneous Reactions

Analyte Reactions/Reactants Additional Treatment Wavelength

NH3 KI + HgI2 410 nm
NO3/NO2 diazotization with pre-reduction (Cd) (to 543 nm
 sulfanilamide then coupling nitrite, for nitrate
with N-(1-naphthyl) analysis only)
-ethylenediamine)
NO3 Chromotripic Acid 410 nm
S(-II) S(-II), FeCl3 & dimethyl-p- 664 nm
phenylenedeamine (forms
methylene blue)
Phenols 4-aminoantipyrine, potassium post-extraction (CHCl3) 460 nm
ferricyanide

b. Measurement of Al and Fe

            Iron and aluminum are polyvalent metals that are used as coagulants in drinking water
treatment.  They are also present in natural waters at usually low concentrations.  In properly
operating drinking water treatment plants, metal coagulants are well removed by sedimentation
or filtration.  However, extremes of pH, incorrect dosing, or poor hydraulics may result in the
breakthrough of high levels of residual iron or aluminum in the distribution system.  This is
undesirable for several reasons.  First these hydrolysing metals may continue to slowly
precipitate far downstream of filtration.  This will cause hydraulic and water quality problems in
any engineered systems that does not have provisions for sludge removal (i.e., clearwells, water
mains, etc.).  Iron may actually impart a metallic taste at high concentrations, and it is well
known to stain plumbing fixtures and laundry.  Under certain circumstances iron residuals can
support the growth of unwanted bacteria.  In the U.S. a secondary maximum contaminant level
has been set for iron at 300 mg/L (based on aesthetics).  The corresponding European
recommended standard is 50 mg/L, with an MCL of 200 mg/L.  High concentrations of aluminum
have been linked to encepalopathy in kidney dialysis patients.  There is also some weak evidence
that it may be associated with Alzheimer's Disease as well.  In Europe, residual aluminum
concentrations in drinking water are limited to 200 mg/L, and the recommended limit is 50 mg/L. 
A primary drinking water standard (most commonly 50 mg/L) is currently being discussed for the
U.S.

            Aluminum and iron may be determined in aqueous solution together by the simultaneous
formation of of two separate colored complexes; Al-ferron, and Fe-phenanthroline (Davenport,
1949).  The formation of an Fe-phenanthroline complex must be preceeded by two steps.  First
iron hydroxides must be dissolved with acid (equation 17.15a), then the iron must be reduced to
the ferrous state with hydroxylamine (equation 17.15b).

                     Fe(OH)3  +  3HCl   ------>    Fe+3    +  3H2O  +  3Cl-                                       (17.15a)

                 4 Fe+3 +  2 NH2OH  ------>  4 Fe+2 +  N2O  +  H2O  +  4H+                            (17.15b)

The ferrous-iron then forms a strongly-colored complex with three molecules of 1,10-
phenanthroline.  Binding occurs at the heterocyclic nitrogens.  This complex is reported to have a
molar absorptivity ( ) of 11,100 M-1cm-1 at lmax = 508 nm (i.e., orange-red).  The intense color is attributed to charge transfer of an
electron from Fe(II) to a vacant p* orbital on the phenanthroline.  Such a transition cannot occur with Fe(III).  As a result, this latter complex only possesses a slight
bluish color due to the weaker ligand-field transition phenomena.
            Aluminum forms a colored complex with several hydroxyquinoline derivatives.  One of
these ligands, 8-hydroxy-7-iodo-5-quinolinesulfonic acid (or Ferron; abbreviated as "F" below),
gives an absorbance maximum at about 370 nm.  This reaction is most sensitive at pH 5.5, and it
is subject to interference from iron.  However iron interference can be conveniently removed by
simultaneous complexation with 1,10-phenanthroline (abbreviated as "P" below).  In this way, a
combined ferron/phenanthroline reagent allows the simultaneous determination of iron and
aluminum when absorbances at 370 nm and 520 nm are recorded.  For example, the overall
absorbance measured at either of these two wavelengths can be broken down as follows:

                                           (17.16)

where:

              =  the Total absorbance of a sample or standard at wavelength "x"; i.e. the
absorbance that you directly measure with a spectrophotometer.

              =  the absorbance of the sample or standard prior to addition of complexing
reagents (i.e., initial sample absorbance) at wavelength "x"

              =  the fraction of the total absorbance that is due to free, un-complexed Ferron at
wavelength "x"

              =  the fraction of the total absorbance that is due to free, un-complexed
Phenanthroline at wavelength "x"

              = the fraction of the total absorbance that is due to Al-Ferron complex

              = the fraction of the total absorbance that is due to Fe-Phenanthroline

complex

Invoking Beer's Law and assuming that all aluminum and iron are completely complexed, we
presume that:

                                                          (17.17)

                                                            (17.18)

                                                                                             (17.19)

                                                                                             (17.20)

where  is the absorptivity (usually in: abs units/mg/L) of substance "y" at wavelength "x". 
The free ligand concentrations are:

                                                                                 (17.21a)

                                                                                 (17.21b)

where "n" and "m" are stoichiometric factors that relate the amount of ligand bound per unit of
metal.  Now combining equations 17.16 through 17.21b one gets:

        
(17.22)

and this reduces to:

         
(17.23)

And if we define the absorbance of the added ligands in their free state (i.e. initial ligands) as
,

                                                                        (17.24)

we get:

                    (17.25)

Now we know that aluminum can be best measured near the absorbance maximum for the
aluminum-ferron complex (370nm), and iron is best quantified at a wavelength near the iron-
phenanthroline maximum (520nm).  From this we can used equation 17.25 to formulate an
expression for Fe and Al concentration:

                               (17.26)

                               (17.27)

The first term of both equations 17.26 and 17.27 (Abs T) is simply the final absorbance of the
samples or standards after treatment with the ferron-phenanthroline reagent.  The second term
(Abssi) is just the absorbance of the samples before anything is done to them.  The third term
(Absli) is obtained from the intercept of the standard curves, and the constant parts of the last two
terms are determined from the slopes of these standard curves.

c. Ammonia-Nitrogen: The Phenate Method

            Ammonia nitrogen may be determined by reaction with phenol and chlorine which gives
a new highly colored condensation product, indophenol (Bolleter et al., 1961).  The reaction is
thought to occur through the formation of monochloramine (equation 17.28).  For quantitative
formation of monochloramine, the pH must be around 6.5 to 7.0.

                                      NH3  +  OCl- ®  NH2Cl  +  OH-                                                           (17.28)

Then this powerful electrophile attacks the anionic phenate to give a quinonechloramine (17.29). 
This intermediate continues to react with phenate forming the highly-conjugated indophenol
(17.30).  Manganous sulfate is added as a catalyst.

                                           (17.29)

                      (17.30)

The color of this final product is quite pH-dependent, yellow at low pH, and blue at high.  The
blue color is most intense at pH 9.9-10.0.

            Procedure

1. To 10 mL of sample in a 100-mL beaker, add 10 mL of high-purity water and

2 drops (~0.1 mL) of manganous sulfate solution.

2. Place on a magnetic stirrer and add 1 mL conc. HOCl solution, followed

immediately by the dropwise addition of 1.2 mL phenate reagent.  Stir
vigorously during the addition of the reagents.

3. Wait 10 min after addition of the reagents, then measure the absorbance at
630 nm.

4. Repeat steps 1-3 for all samples along with a set of 4 standard solutions
prepared from the 25 mg/L ammonia solution.  The concentrations you
choose for the standard solutions should be properly spaced such that the
unknowns all fall between the highest and lowest standard.

            Reagents
 1.  Conc. HOCl solution:  Dilute 20 ml of the 5% NaOCl stock to 100 mL. 
Adjust pH to 6.5 - 7.0 with HCl.

2.  Manganous sulfate solution (0.006 N):  Dissolve 50 mg monohydrate

manganous sulfate in 100 mL distilled water.

3.  Phenate reagent:  Dissolve 2.5 g NaOH and 10 g phenol in 100 mL distilled
water.

d. Nitrite-Nitrogen

            Nitrite can be measured easily and with good sensitivity by a coupling reaction known as
diazotization.  Under acidic conditions, nitrite ions and aromatic amines can form reactive
diazonium salts.  In the test commonly employed for nitrite, the aromatic amine used is
sulfanilamide.  This forms p-diazobenzenesulfonamide (equation 17.31).

   
(17.31)

To improve the intensity of the absorbance band, this is reacted with N-(1-
naphthyl)ethylenediamine to form a reddish purple azo dye, p-benzenesulfonamide
azonaphthylethylenediamine (equation 17.32).

       
(17.32)

Absorbance of this azo dye is measured at 543 nm.  It obeys Beer's law up to 180 mg-N/L.  This
method requires the use of nitrite-free dilution water and accurate nitrite standards.  If its not
certain that distilled water is nitrite-free, it can be treated with a small amound of potassium
permanganate.  This common oxidizing agent converts nitrite to nitrate.  Residual permanganate
can be removed by fractional distillation of the water, discarding any pink distillate.  Because
nitrite is so easily oxidized, even fresh solutions must be standardized.  This is done by oxidation
with a standard permanganate solution and back titration with a standard reducing agent (oxalate
or ferrous ammonium sulfate).

            Procedure
 1. To 50 mL of sample in a 100 mL beaker, add 2 mL of the color reagent and
mix..

2. Wait 10 min after addition of the reagents, then measure the absorbance at
543 nm.  If the absorbance is greater than that measured for the 0.025
mg/L standard, dilute the original sample and repeat steps 1-2.

3. Repeat steps 1-2 for all samples along with three standard solutions prepared
from the 250 mg/L nitrite-N solution.

            Reagents

1.  Color Reagent:  Place about 800 mL of distilled water to a 1-liter volumetric
flask.  To this slowly add 100 mL of 85% phosphoric acid and then 10 g
sulfanilamide.  Once the sulfanilamide is completely dissolved, add 1 g
N-(1-naphthyl)-ethylenediamine dihydrochloride and dissolve.  Dilute to
1 liter.  This reagent must be stored in a dark bottle in a refrigerator.  It
can be used for 1 month.

2.  Nitrite Standard (250 mg/L as N) Dissolve 1.232 g NaNO2 in 1 liter of

distilled water.  Add 1 mL chloroform as a preservative.  Standardize
with permanganate and oxalate or ferrous ammonium sulfate.

Figure 17.3
 Nitrite Standard Curve

Figure 17.4

Determination of Nitrite in an Unknown from Several Dilutions

e. Ozone: Gas phase

            Ozone is powerful oxidant that is widely used in Europe for the treatment of drinking
water.  Its use in the U.S. is less common, but it is a rapidly growing technology.  Principal
applications of ozone include disinfection, oxidation of Fe and Mn, oxidation of industrial
pollutants, bleaching of color, and improvement of coagulation/filtration.

            1. Direct UV Absorption

            Commercial gas phase ozone monitors are based on the direct measurement of ultraviolet
absorbance.  With bench-scale studies, it is often convenient to use a laboratory UV-Vis
spectrophotometer equiped with a flow-through quartz cell (0.1-0.2 cm pathlength) as a
substitute for a dedicated ozone gas monitor.  The ozone concentration may be calculated based
on Beer's Law and the Ideal Gas Law.

                                                  (17.33)

where ao is the absorptivity in atm-1cm-1 of ozone at the wavelength of measurement, Ta is the

absolute temperature of the gas being measured, P is the pressure in mm Hg of the gas, and L is
the cell pathlength in cm.  At 253.7 nm, the absorptivity of ozone in the gas phase (at 760 torr,
293oK or 20oC) is 134 atm-1cm-1 (Inn & Tanaka, 1959; Hearn, 1961; DeMore & Patapoff, 1976). 
There is quite a bit of fine structure to ozone's broad UV absorbance band in the gas phase.  For
this reason, narrow band widths are preferable.  When expressed as mass per volume, the terms
for pressure and temperature drop out and one gets equation 17.34a.

                                                         
                                    (17.34a)

which for a wavelength of 253.7 nm and a pathlength of 0.2 cm reduces to equation 17.34b.

                                                                                    (17.34b)

Use of equations 17.32 and 17.33 is more convenient in the laboratory.  If necessary, one can
convert back to percent-based concentrations by equation 17.34c.

                                  (17.34c)

            2. Iodometric Methods

            Gas phase concentrations of ozone are also easily measured iodometrically.  Wet
chemical analysis of ozone in the gas phase is less convenient, however, it allows one to calibrate
or check spectrophotometric analyzers.  Also, the wet chemical tests measure mass flow rather
than concentration directly.  If there are uncertainties in the gas flow rate, the wet chemical
methods may provide greater accuracy in estimating mass ozone application rates.

            Many variations on the classic iodometric method have been proposed,  Basically, a
portion of the gas stream is directed to a gas bubbler filled with 2% KI solution for an exact
period of time.  Ozone reacts stoichiometrically to form an equivalent amount of iodine.  In the
presence of excess iodide, the triiodide ion (I3) is formed.

                 O3  +  2KI  +  H2O   --------->   I2  +   O2  +  2OH-  +  2K+                                  (17.35)

                               I2  +  I-   <--------->    I3-                                                                                           (17.36)

The iodine formed is then titrated with sodium thiosulfate using starch as an indicator to
accentuate the endpoint (APHA et al., 1985).  Phosphate buffered KI solutions are to be avoided,
as the phosphate anions appear to catalyze the formation of hydrogen peroxide (Flamm, 1977). 
This can lead to a shifting endpoint and unpredictable stoichiometry.  Flamm (1977) reports that
a borate buffered modification of the standard KI procedure has shown excellent agreement with
direct gas-phase UV absorption.  He measured the triiodide by spectrophotometry (352 nm).
Calibration of this method requires the spectrophotometric analysis of standard triiodide
solutions.  These solutions are prepared by oxidation of iodide using standard iodate solutions
(equation 17.37).   The stoichiometry of this reaction is preserved when conducted in 0.1 to 1.0
N acid.  The direct preparation of triiodide from iodine crystals and iodide is less accurate.

                                           (17.37)

From equations 17.35 and 17.37, one concludes that one mole of iodate liberates the same
amount of iodine/triiodide as three moles of ozone.  Since iodine or triiodide has two equivalents
of oxidizing potential, ozone also has two, and iodate has six.
            a. Procedure

1. Bubble ozone gas through a gas washing bottle containing a convenient volume
of BKI solution (usually 250-500 mL).  Record the bubbling time and gas
flow rate or settings.  A sample of the original (time zero) BKI solution
should be saved for determination of UV blank.  For low-level
measurements, it is recommended that the BKI solution be very slightly
preozonated.  This removes small amounts or reducing materials that are
invariably present in commercial potassium iodide.
2. At the end of the ozone trapping period, disconnect the gas washing bottle, and
pour a small sample (e.g., 10 mL) into a 50 mL beaker.
3. Measure absorbance of the ozonated BKI at 352 nm after 1 minute (Abss).  If the
absorbance is greater than 1.2, the sample must be diluted.  The final
absorbance value must then be corrected for this dilution.
4. Obtain a blank measurement by determining the absorbance at 352 nm of an
aliquot of the same BKI solution that was present in the gas washing bottle
at time zero (Absb).
5. Calculate concentration with equation #8.  Obtain the calibration factor, b1, by
the standardization procedure in "b".

                               Ozone Concentration (mg/L as O3)  =  (Abss-Absb)/b1                           (17.38)

            b. Standardization

1. Prepare a set of standard triiodide solutions.  This is done by first cleaning a

series of 100 mL volumetric flasks, and half filling them with the Acidic KI
solution.  Add aliquots ("V" mL) of the Standard Potassium Iodate Solution
to each and fill to the mark with additional acidic KI solution.  Each mL of
iodate will be roughly equivalent to 1.5 mg/L ozone.
2. Measure the absorbance of each of the standard triiodide solutions at 352 nm. 
Plot absorbance vs equivalent ozone concentration.

                                               Abs  =  b1 (Equiv. Conc.)                                     (17.39)

     where the equivalent concentration is given by:

    Equiv. Conc. (mg/L as O3)  =  MIO3 (100) (3Mole IO3) (48,Mole O3O3)         (17.40)

                               =  1,440  MIO3 V

     Note that MIO3 is the exact molar concentration of the Standard Potassium Iodate
Solution (see equation 11).  Try to work only in the linear range, or the
range from 0-1.2 absorbance units.

            c. Reagents
 1. BKI Reagent (0.1 M Boric Acid, 1% Potassium Iodide):  Add 6.2 g H3BO3 and
10.0 g KI to 1 liter of distilled water.  Stir to dissolve.
2. Standard Potassium Iodate Solution: Dry the primary standard at 120 oC for 2
hours.  Then, weigh out about 0.021 g of the dried material.  Record the
exact weight to 4 significant figures.  Dissolve in distilled water in a 100
mL volumetric flask and fill to the mark.  Calculate exact molar
concentration:

                          MIO3  =  wt in grams /  21.402                            (17.41)

3. Acidic KI Solution (1% potassium iodide in 0.1N acid): Add 5.6 mL of

concentrated H2SO4 to a 1 liter volumetric flask.  Slowly fill the flask about
half-way with distilled water and stir.  Then add 10 g KI, stir, and fill to the
mark with distilled water.

            Strictly speaking, all of these iodometric methods are non-selective.  That is, they measure a
wide range of oxidants, not just ozone.  For this reason, they should not be used to measure aqueous
ozone concentrations.  Because ozone is by far the major oxidant species produced by corona ozone
generators, and because ozone is far more easily stripped from water, the measurement of ozone in
the gas phase is not subject to significant problems with interferences.  Thus, iodometric methods
may be used in this case without reservation.

e. Ozone: Aqueous Phase

            1. Direct UV Method

            Aqueous ozone concentrations in pure (e.g., distilled) water may be conveniently
determined by direct spectrophotometric measurement at 260 nm.

               CO3 (mg/L as O3)  =   14.59*(Abs @260 nm)                   (17.42)

Equation 12 is based on a molar absorptivity of 3290 M-1cm-1 (Hart et al., 1983).  Unfortunately,
most solutes will interfere at this wavelength, so with actual environmental samples another method
must be used.  Since the iodometric method is too nonspecific for aqueous determinations, the
indigo method of Bader and Hoigne (1981) is recommended.

            2. Indigo Method

            This colorimetric procedure uses solutions of indigo trisulfonate (Bader & Hoigne, 1981). 
Ozone will stoichiometrically bleach this intense blue dye, and the loss in absorbance at 600 nm
may be translated directly into an ozone concentration.  The reaction product is relatively unreactive
to further ozonation.  The reaction is best carried out at low pH to minimize ozone decomposition,
and preserve the 1:1 stoichiometry.  Bader and Hoigne (1981) report a sensitivity factor or apparent
absorptivity for indigo trisulfonate of 20,000 M-1cm-1.  This is based on an aqueous ozone molar
absorptivity of 2900 M-1cm-1.  If one adopts the higher value reported by Hart et al. (i.e., 3290 M-
1
cm-1), the sensitivity factor for indigo trisulfonate becomes 22,700 M-1cm-1.
            This method is quite selective, however, it is subject to a few notable interferences.  The
presence of residual chlorine will cause a positive bias.  Addition of 500 mg/L malonic acid to the
Indigo Reagent solves this problem by out-competing the indigo for the chlorine.  Oxidized
manganese species will also result in bleaching of the indigo.  Here it is recommended that
duplicate samples be analyzed, one according to the standard procedure, and one following addition
of glycine.  The glycine selectively reduces residual ozone without affecting oxidized manganese
species.  The true ozone concentration may then be estimated from the difference of these two
measurements.

            a. Procedure

1. Prepare an indigo blank by adding 1 mL of the Standard Indigo Stock to a 25 mL

volumetric flask and filling to the mark with Phosphate Buffer.  Stopper and
mix.  Measure the absorbance of this solution at 600 nm (Abs i).  When the
Indigo Stock is new, it should be about 0.650.  With time this value will
drop.  When it falls below 80% of the original value, prepare a new Indigo
Stock and repeat procedure.  If low ozone concentrations are anticipated
(i.e., < 0.3 mg/L) prepare a Secondary Indigo Stock by adding 20 mL of the
Standard Indigo Stock to a 100 mL volumetric flask and diluting to the
mark with Super-Q water.
2. Soak a series of volumetric pipets in a dilute ozone solution for several minutes. 
These pipets are to be used to transfer the solution to be measured to the
indigo-containing flasks.  They must therefore be rendered ozone-demand-
free.  The capacities of the pipets will depend on the range of anticipated
ozone concentrations (see table below).
3. Assemble a series of clean, glass-stoppered 25 mL or 50 mL volumetric flasks
(see table below) and fill each with 1.00 mL of Standard Indigo Stock (25
mL flasks) or Secondary Indigo Stock (50 mL flasks) using a volumetric
pipet.  Wash this down from the inner surfaces of the neck with about 10
mL of Phosphate Buffer.
4. Quickly pipet the recommended sample volume to an indigo-containing
volumetric flask (see table below).  Be sure that the tip of the pipet is below
the meniscus.  Fill the flask to the mark with Phosphate Buffer, cap and
invert several times to mix.

(Vs) (Vt) (L)

Anticipated Recommended Recommended Recommended
Ozone
 Concentration  Sample Total Volume Pathlength
Volume
0 - 0.2 mg/L 25 mL 50 mL * 10 cm
0.1 - 0.3 mg/L 15 mL 50 mL * 10 cm
0.2 - 0.5 mg/L 10 mL 50 mL * 10 cm
 0.3 - 2.0 mg/L 15 mL 25 mL 1 cm
1.5 - 3.0 mg/L 10 mL 25 mL 1 cm
1.8 - 3.5 mg/L 8 or 9 mL 25 mL 1 cm
2.3 - 4.5 mg/L 6 or 7 mL 25 mL 1 cm
3 - 6 mg/L 5 mL 25 mL 1 cm
4 - 7 mg/L 4 mL 25 mL 1 cm
5 - 10 mg/L 3 mL 25 mL 1 cm
7 - 15 mg/L 2 mL 25 mL 1 cm
15 - 30 mg/L 1 mL 25 mL 1 cm

                                                          *
Secondary Indigo Stock must be used with 10 cm pathlength cells

5. Measure absorbance (Absf) of each sample at 600 nm using cells of the indicated
pathlength (L).  Concentration is calculated from a slope or calibration
factor determined by calibration against the direct UV method (see b.
Calibration).

                     Ozone Conc. (mg/L as O3)  =  Vt(Absi - Absf)/b1VsL                   (17.43)

            b. Calibration

1. Prepare a series of aqueous ozone standards in slightly acidified (0.1 mM HNO3)

Super-Q water.  This is generally done by first bubbling ozone gas through
about 500-1000 mL of the acidified water until near saturation (~1 hour). 
Under optimal conditions for high ozone output (i.e., high voltage, low gas
flow, low temperature) this should result in aqueous concentrations of about
10 mg/L.  Then, aliquots of this water are removed and diluted with varying
amounts of un-ozonated, acidified Super-Q water.  The degree of dilution
will depend on the range of ozone concentrations anticipated in the samples
of interest.
2. One-by-one measure each of these diluted solutions by the indigo method above
and by the direct UV method (equation 17.42).  Use the same sample
volume, Vs, for all standards.
3. Plot absorbance (from indigo method) versus aqueous ozone concentration (from
direct UV method).  The slope of this line multiplied by Vt/Vs gives b1L (see
equation 17.44).  Based on the presumed sensitivity factor for indigo
trisulfonate of 22,700 M-1cm-1, the calibration factor, b1, should be about
0.47 abs/cm per mg-O3/L.

                    Absf =  Absi - (b1LVs/Vt) (Ozone Conc.)                    (17.44)

            c. Reagents
 1. Standard Indigo Stock (1 mM in 20 mM phosphoric acid): Dissolve 1.36 mL
conc. H3PO4 in 1 liter of super-Q water and mix.  To this add 0.6 g indigo
trisulfonate, mix and store in a brown glass bottle.
2. Phosphate Buffer (pH 2): Dissolve 28 g NaH2PO4.H2O and 20.6 mL (35 g) conc.
H3PO4 in Super-Q water and dilute to 1 liter.

f. Aqueous Total Oxidant Concentration

            Ozone reacts in aqueous solution to give a variety of oxidant species.  These may include
simple inorganic oxygen-containing free radicals, hydrogen peroxide, organic free radicals, and
organic peroxides.  The iodometric method of Flamm (1977) coupled with the use of a catalyst
(Taube & Bray, 1940) is sufficiently non-specific to be useful for the determination of total oxidant
concentration.

            a. Procedure

1. Place between 1 and 9 mL of BKI reagent in a 10 mL volumetric flask.

                     2. Add 1 mL of the Molybdate Catalyst solution.

3. Quickly add sample to the mark, stopper and invert several times to mix.

                     4. Measure absorbance at 352 nm after 1 minute (Abss)

                     5. Repeat steps 1-4 substituting distilled water for the sample (Absb)

                     6. Calculate total oxidant concentration:

            Total Oxidant Concentration (mg/L as O3)  =  b1*(Abss-Absb)/Vs                                                                          

(17.45)

     Follow the procedure in part "b" using iodate standards to get the calibration
factor, b1.  Experience indicates that it should be about 21.6 mg/L as O3 per
mL of sample per cm pathlength.

            b. Standardization

1. Prepare a set of standard triiodide solutions.  Clean a series of 100 mL

volumetric flasks, and half fill them with the Acidic KI solution.  Add
aliquots ("V" mL) of the Standard Potassium Iodate Solution to each and fill
to the mark with additional acidic KI solution.  Each mL of iodate will be
roughly equivalent to 1.5 mg/L ozone.
2. Measure the absorbance of each of the standard triiodide solutions at 352 nm. 
Plot absorbance vs equivalent ozone concentration.

   Equiv. Conc. (mg/L as O3)  =  MIO3 (100) (3Mole IO3) (48,Mole O3O3)          (17.46)
                               =  1,440  MIO3 V

     Try to work only in the linear range, or the range from 0-1.2 absorbance units.

            c. Reagents

1. BKI Reagent (0.1 M Boric Acid, 1% Potassium Iodide):  Add 6.2 g H3BO3 and
10.0 g KI to 1 liter of distilled water.  Stir to dissolve.
2. Molybdate Catalyst Solution:  Add 0.9 g Ammonium Molybdate to 10 mL
distilled water and stir to dissolve.
3. Standard Potassium Iodate Solution: Dry the primary standard at 120 oC for 2
hours.  Then, weigh out about 0.021 g of the dried material.  Record the
exact weight to 4 significant figures.  Dissolve in distilled water in a 100
mL volumetric flask and fill to the mark.  Calculate exact molar
concentration:

                          MIO3  =  wt in grams /  21.402                            (17.47)

4. Acidic KI Solution (1% potassium iodide in 0.1N acid): Add 5.6 mL of

concentrated H2SO4 to a 1 liter volumetric flask.  Slowly fill the flask about
half-way with distilled water and stir.  Then add 10 g KI, stir, and fill to the
mark with distilled water.

g. Hydrogen Peroxide

            Hydrogen peroxide is one of the secondary oxidants produced by the action of ozone.  It
may be produced as a byproduct of aqueous ozone decomposition.   It is also an important
byproduct of the reaction of ozone with unsaturated organic compounds.  Hydrogen peroxide is also
important, because it may be added to waters undergoing ozonation for the purposes of accelerating
the formation of hydroxyl radicals.  These radical species are capable of oxidizing many types of
organic structures that are not affected by molecular ozone.  The two preferred methods for
hydrogen peroxide analysis involve the formation of heavy metal complexes (Parker, 1928;
Masschelein et al., 1977).  Both are susceptible to interference from aqueous ozone.  Therefore,
ozone must first be purged.  The titanium method (Parker, 1928) is recommended here.

            a. Procedure

                     1. Sparge sample to remove residual ozone as necessary.

                     2. Pipet 1 mL of the Titanium Reagent into a 10 mL volumetric flask.

                     3. Add sample to the mark, stopper and invert several times.

                     4. Measure absorbance at 415 nm.

                     5. Calculate concentration from a calibration curve prepared as below.

            b. Standardization

1. Prepare two serial dilutions of the commercial 30% solution.  Add 1 mL of the
commercial solution to a 500 mL volumetric flask and dilute to the mark
with distilled water.  This is the concentrated peroxide stock solution.  In a
second 500 mL volumetric flask dilute 5 mL of the concentrated stock to
500 mL with distilled water.  This is the dilute peroxide stock.
2. Standardize the dilute peroxide stock using the BKI spectrophotometric method
for total oxidants.  In this case the standard curve is most conveniently
based on equivalent hydrogen peroxide concentrations in mg/L.

Equiv. Conc. (mg/L as H2O2)  =  MIO3 (100)(3 Mole IO32)(34,Mole H2O2O2)        (17.48)

                              =  1,020  MIO3 V

                          or:

  Equiv. Conc. (mg/L as H2O2)  =  0.708 Equiv. Conc. (mg/L as O3)  (17.49)

3. Into a series of 10 mL volumetric flasks, place 1 mL of the Titanium Reagent. 

Add varying amounts of the dilute peroxide standard, and fill to the mark
with distilled water.  Measure the absorbance at 415 nm.  Plot concentration
versus absorbance.  The data should be linear with a slope of about 57 mg/L
per absorbance unit per cm.

            c. Reagents

1. Titanium (IV) Reagent:  Place 10 mL of 6N HCl in a beaker on a crushed ice

bath under a fume hood.  Place TiCl4 in an ice bath and allow both to cool. 
Carefully add 10 mL of the TiCl4 dropwise to the chilled HCl.  Allow the mixture to stand in the ice bath
until the solids that form dissolve.  Transfer this mixture to a 1 liter volumetric flask using 6N HCl.  Dilute to the mark with
additional HCl.

                     2. 30% Hydrogen Peroxide Solution (Commercially available).

3. BKI Reagent: (This is the same solution prepared for ozone gas phase analysis
and the total oxidant analysis; 0.1 M Boric Acid, 1% Potassium Iodide): 
Add 6.2 g H3BO3 and 10.0 g KI to 1 liter of distilled water.  Stir to dissolve.
4. Molybdate Catalyst Solution: (This is the same solution prepared for the total
oxidant analysis) Add 0.9 g Ammonium Molybdate to 10 mL distilled water
and stir to dissolve.
5. Standard Potassium Iodate Solution: (This is the same solution prepared for
ozone gas phase analysis and the total oxidant analysis) Dry the primary
standard at 120oC for 2 hours.  Then, weigh out about 0.021 g of the dried
material.  Record the exact weight to 4 significant figures.  Dissolve in
 distilled water in a 100 mL volumetric flask and fill to the mark.  Calculate
exact molar concentration:

                          MIO3  =  wt in grams /  21.402                            (17.50)

6. Acidic KI Solution (1% potassium iodide in 0.1N acid; This is the same solution
prepared for the ozone gas phase analysis and the total oxidant analysis):
Add 5.6 mL of concentrated H2SO4 to a 1 liter volumetric flask.  Slowly fill
the flask about half-way with distilled water and stir.  Then add 10 g KI,
stir, and fill to the mark with distilled water.

REFERENCES

Bader, H.; Hoigne, J. Wat. Res. 1981, 15:449-456.

Bolleter, W.T., C.J. Bushman and P.W. Tidwell (1961) "Spectrophotometric Determination of
Ammonia as Indophenol," Anal. Chem. 33(4)592-594.

Davenport, W.H. (1949) Anal. Chem. 21.

DeMore, W.B.; Patapoff. M. Environ. Sci. Technol. 1976, 10:897.

Flamm, D.L. Environ. Sci. Technol. 1977, 11:10:978-983.

Hart, E.J.; Sehested, K.; Holcman, K. Anal. Chem. 1983, 55:46.

Hearn, A.G. Proc. Phys. Soc. 1961, 78:932.

Inn, E.C.Y.; Tanaka, Y. In Ozone Chemistry and Technology 1959, ACS Advances in Chemistry
Series #21, ACS, Washington, pp.263-268.

Masschelein, W.J.; Denis, M.; Ledent, R. Wat. Sewage. Wrks. 1977, August, 69-72.

Parker, G.A. In Colorimetric Determination of Nonmetals, 1928, Boltz, D.R.; Howell, J.A., Eds.
Wiley, New York, pp.301-303.

Taube, H., Bray, W.C. J. Am. Chem. Soc. 1940, 62:3357.

new stuff

III. MEASUREMENT OF OZONE

A. Gas Phase Concentrations

            Commercial gas phase ozone monitors are based on the direct measurement of ultraviolet
absorbance.  With bench-scale studies, it is often convenient to use a laboratory UV-Vis
spectrophotometer equiped with a flow-through quartz cell (0.1-0.2 cm pathlength) as a
substitute for a dedicated ozone gas monitor.  Gas phase concentrations of ozone are also easily
measured iodometrically.  A portion of the gas stream is directed to a gas bubbler filled with 2%
KI solution for an exact period of time.  Ozone reacts stoichiometrically to form an equivalent
amount of iodine.  In the presence of excess iodide, the triiodide ion (I3) is formed.

                          O3  +  2KI  +  H2O   --------->   I2  +   O2  +  2OH-  +  2K+                     (20)

The iodine formed is then titrated with sodium thiosulfate using starch as an indicator to
accentuate the endpoint (APHA et al., 1985).  Flamm (1977) reports that a borate buffered
modification of the standard KI procedure has shown excellent agreement with direct gas-phase
UV absorption.  He measured the triiodode (iodine + iodide) formed by spectrophotometry (352
nm).

B. Aqueous Concentrations

            Aqueous ozone concentrations in pure (e.g., distilled) water may be conveniently
determined by direct spectrophotometric measurement at 260 nm.

                                     CO3 (mg/L)  =   14.59*(Abs @260 nm)                                (21)

Equation 47 is based on a molar absorptivity of 3290 M-1cm-1 (Hart et al., 1983).  Unfortunately,
most solutes will interfere at this wavelength, so with actual drinking waters another method
must be used.  Since the iodometric method is too nonspecific for aqueous determinations, the
indigo method of Bader and Hoigne (1981) is recommended.  This colorimetric procedure uses
solutions of indigo trisulfonate.  Ozone will stoichiometrically bleach this intense blue dye, and
the loss in absorbance at 600 nm may be translated directly into an ozone concentration.

old stuff:

********************************  Research Topic  *****************************

a. UV Absorbance

            Very little structural information may be obtained from the ultraviolet-visible absorption
spectrum of humic substances.  Figure 17.1 shows a typical scan of an aquatic fulvic acid.  Note
the total lack of features.  One sees only a monotonic decrease in absorbance with increasing
wavelength.  Nevertheless, the absolute value of the absorbance has been found to correlate with
certain water quality characteristics, such as DOC and THMFP, and this is often the reason for
measuring UV absorbance.

Figure 17.1

UV-Visible Absorbance Scan of an Aquatic Fulvic Acid

            It is now reasonably well established that the absorbance of light in the UV-Visible range
by aquatic humic substances conforms to Beer's law (Black & Christman, 1963a,b, Packham,
1964).  That is, the concentration of humic materials in a natural water which has been diluted
with distilled water is linearly proportional to its absorbance.  Therefore, one can calculate a
specific absorbance which corresponds to the absorbance per mg/l of DOC.  Some specific
absorbances for extracted humic materials are listed in Table 17.6.  As a result of the
conformance to Beer's law the DOC of a raw water can be estimated by measuring its absorbance
at a fixed wavelength provided that the relative concentration of all the constituent organics
remains fixed.  Although this is never strictly true, changes in the nature of the raw water may be
small enough to be unimportant.  Table 17.7 summarizes some attempts to correlate UV
absorbance (254 nm) to DOC for raw waters.  Note that the reciprocal of the slopes in Table 17.7
correspond to the specific absorbances in Table 17.6, and that the average slope (~25) gives a
specific absorbance of 0.004 which is similar to those reported for riverine fulvic acids.

            Note that groundwater humics and humics from eutrophic lakes are less colored than
riverine humics.  Humics from bogs and swamps are most colored.

            Since color increases with increasing pH, it is important to make all color measurements
at the same pH.  By convention, a pH 7 phosphate buffer is generally used.

Table 17.6

Specific Absorbance of Aquatic Humic Substances

Sample           Specific Color (400nm)        Specific UV abs (254nm)   Ref.

                 Average       Range           Average      Range 

Groundwater

  Fulvic Acid      0.001                                                   1

  Humic Acid       0.002                                                   1

Rivers & Streams

  Fulvic Acid      0.005                                                   1

                   0.0033    0.0023-0.0047       0.035    0.029-0.043      2

  Humic Acid       0.007                                                   1

                   0.0081    0.0071-0.0089       0.054    0.049-0.059      2

Wetlands

  Fulvic Acid      0.010                                                   1

  Humic Acid       0.014                                                   1

References

 1. Thurman, 1985

 2. Reckhow, 1988

Table 17.7

UV Absorbance (254 nm) to DOC Regression Equations+

    Water           Slope   Intercept      r           Reference

Tjeukemeer Lake      24.7*     2.7       0.925         De Haan et al., 1982

Grasse River         19.8      0.65      0.93          Edzwald et al., 1985

Glenmore Reservoir   28.1     -0.3       0.71          Edzwald et al., 1985

                     24.1      2.2       0.96          Smart et al., 1976

N.C. Surface Wat.    29.1      1.44      0.83          Singer et al., 1982

+
Model:  DOC (mg/l) = (slope)*(UV absorbance, cm  ) + intercept

* Based on absorbance at 250 nm, neutral pH.

            Estimation of THMFP  Trihalomethane formation potential (THMFP) is the

trihalomethane formation that occurs under a specified set of extreme conditions.  Reaction
conditions are chosen so that THMFP approaches the maximum possible yield.  A number
of researchers have proposed the use of simple linear models for THMFP, based on a single
easily-measured property of the water (equation 7).

                 THMFP  =  b0 + b1(independent variable)                     (17.12)

            Ultraviolet absorbance of a water has also been used as a surrogate for THMFP.  Its
utility is based on the presumption that conjugated and electron-rich structures that are
responsible for much of the absorption of 254 nm light, are also readily attacked by
chemical electrophiles such as chlorine.  A certain fraction of these reactions (e.g., ~3%)
will ultimately lead to the formation of THMs.  UV absorbance is generally determined
using light of 254nm wavelength, and it should be measured only on samples that have been
previously filtered.  Table 17.8 shows results of some studies where UV absorbance
measurements were made on filtered samples.

Table 17.8

Correlations Between UV Absorbance* and THMFP*

 Sample Type         b0       b1       r2       Origin           Ref 

Humic Substances            1570-150         Aq. Fulvic Acids     4  

                            1320-220         Aq. Humic Acids      4  

Raw Waters          102     2106      0.86   N. Carolina Waters   1  

                     38     1573      0.94   Grasse River         2  

                     49     1596      0.82   Glenmore Res.        2  

                    101-74  2120-180  0.80   Colored Waters       3  

                    -62     1975      0.741  Colored Waters       8  

Treated Waters       38     1629      0.97   Canton WTP           2  

                     42     1578      0.94   Oneida WTP           2  

*
THMFP is in units of ug/l, UV Absorbance is in abs. units/cm.

References

1. Singer et al. (1981); THMFP conditions - 7 days, pH 6.7, sufficient chlorine to

maintain a residual of 15-20 mg/l; Chlorine demand - 7 days
2. Edzwald et al. (1985); THMFP conditions - 7 days, pH 7.5, 3-5 mg chlorine per
mg TOC; Chlorine demand - 2 hours, 10-35 mg/l dose.
3. Singer (1987); THMFP conditions - 7 days, pH 7, 2-3 mg chlorine per mg
TOC.
4. Reckhow (1984); THMFP conditions - 3 days, pH 7, 20 mg/l chlorine, 20 oC;
values adjusted to 7-day reaction time.
 5. Urano et al. (1983); General THM Model adapted to the following conditions:
7 days, pH 7, 20 mg/l chlorine dose, 20oC
6. Morrow & Minear (1987); General THM Model adapted to the following
conditions: 7 days, pH 7, 20 mg/l chlorine dose, 20oC
7. Oliver & Thurman (1983); THMFP conditions - 7 days, pH 7, 10 mg/l chlorine,
20oC.
8. Amy et al., 1987); THMFP conditions - 3 days, pH 7, chlorine to carbon mass
ratio of 3.0, 20oC; General THM Model adapted to the following
conditions: 7 days, pH 7, 20 mg/l chlorine dose, 20oC

*************************************************************************
**

The old mercury vapor lamps, once common in spectrophotometers, had an especially high
intensity at 254 nm.  Therefore, measurements at this wavelength were favored, because they
were especially free of interferences posed by stray light, etc.