This article was downloaded by: [University of Saskatchewan Library]

On: 23 February 2013, At: 01:55

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Ozone: Science & Engineering: The Journal of the

International Ozone Association
Publication details, including instructions for authors and subscription information:
[Link]

Residual Ozone Measurement: Indigo Sensitivity

Coefficient Adjustment
a b
Gilbert Gordon & Bernard Bubnis
a
Miami University, Department of Chemistry, Oxford, Ohio, 45056
b
Novatek, A Division of EBB Inc, Oxford, Ohio
Version of record first published: 24 Oct 2007.

To cite this article: Gilbert Gordon & Bernard Bubnis (2002): Residual Ozone Measurement: Indigo Sensitivity Coefficient
Adjustment, Ozone: Science & Engineering: The Journal of the International Ozone Association, 24:1, 17-28

To link to this article: [Link]

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: [Link]

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 OZONE SCIENCE & ENGINEERING 0191-9512/02 $3.00+ .00
VoL 24, pp. 17-28 international Ozone Associaiion
Printed in the USA. Copyright 0 2002
Residual Ozone Measurement:
Indigo Sensitivity Coefficient Adjustment
Gilbert Gordon'
Miami University, Department of Chemistry, Oxford, Ohio 45056 (* Corresponding Author)

Bernard Bubnis
Novatek, A Division of EBB Inc., Oxford Ohio
Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

Received for review: 5 January 2001

Accepted for publication: 9 October 2001

Abstract

The determination of residual ozone using indigotrisulfonate (Standard Methods 4500-03B) assumes a molar
absorptivity of -20,000 ~ - ' c m - and
l a constant sensitivity coefficient of 0.42 L mg-'cm-I. Data are presented
showing large molar absorptivity differences and subsequent differences in the calculated sensitivity
coefticients for various sources of indigo. The variation in molar absorptivity (up to 15%) emphasizes the
need for measuring the sensitivity coefficient on a regular basis. A revised simple calculation that
compensates for decomposing indigo reagents is proposed. A modified equation that can be used with the
current Standard Method procedure is given. However, for the most accurate results, it is recommended that
the dry indigo reagent be periodically calibrated.

Key Words

Ozone; Indigo; Molar Absorptivity; Sensitivity Coefficient; Residual Ozone;

Introduction

The standard method for measuring residual ozone the blank is prepared with 90 mL of water plus 10 mL
(0,) is based on the reaction of 0 3 with of indigo solution. Be aware that when applying the
indigotrisulfonate as reported by Bader and HoignC gravimetric method, the volume of the blank could be
(1,2,3,4) and accepted by the International Ozone variable.
Association QAIQC Committee(5). To calculate the Spectrophotomebic, volumetric method
residual O, concentration, Bader and HoignC
determined an indigo sensitivity coefficient ( f ) of
0.42 L mg-'cm-' using a molar absorptivity of
-20,000 ~ - ' c m - 'for the change of absorbance (@
600 nm) per mole of added O3per liter.
where: ABlanl; absorbance of blank
Two indigo analytical methods (Standard Method Asml, absorbance of sample
4500-03B) are routinely used to measure residual 03. f indigo sensitivity coefficient, 0.42
The methods are based on an absorbance difference b pathlength of cell, cm
as measured between a sample and blank at 600 nm V volume of sample, mL (normally 90
as shown in Equations 1 and 2. It is important to note
that each of these methods as presented presume that
 18 G . Cordon a n d B. Bubnis

Spectrophotom etrk, gravim etric method have sensitivity coefficients that deviate
substantially from the generally accepted value o f
( AB,",,~ , c V T)
X 100 ) -( A ~ a m ~ X
0.42 L mg!'cm!'.
mg o J /L = [21
f xVsxb Table I is a typical data set used to determine the
..
indgo senstivity coeficient by the ozone
ere: absorbance of blank
calibration method A shrinking bottle (8) was used
absorbance of sample to deliver stock ozone s o l ~ i o n (measured
s by UV @
volume of sample, m L 260 nm, - = 3,000 ~ ! ' c m ! 't)o 100 m L volunetric
= [(final wt tare wt) g H I .O mL/gl
f l a k s containing indlgo. T h e decrease in indgo
-10mL absorbance was measured using 2 crn measuring celk.
total volume of sample plus indgo,
mL
, T h e data in Table I has the following features. T h e
Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

= (final wt tare wt) g H 1.0 mL/g

decrease in the absorbance of the indgo soluion in
b pathlength of cell cm the presmce of ozone is linear. There is good
f indgo sensitivity coefficient, 0.42 agreement betwen the y-intercept of the
calibration line and the blank absorbance of the
T l ~ c a s s m p t i o nin each of the procedures is that all indgo soluion. T h e x-intercept indicates the
s o ~ m c s of indigo have a constant sensitivity maximum 0, measlrable r a n g . T h e slope of the
cocficient. Recently, G o r h n and co-wrkers have calibration line (f)equals 0.42 L mg!'cm!'.
rcported (6,7) that different sources of indgo
can

Table I. Typical Ista S b for ktermirdng the Sensitivity Coefficient Using the O m e Calibration Techmique

0 3 0 3 Calalated
(mgiL) (moleslL) Absorbance . -Absorbance (M"cml')-

AVG 20.274 V 294

y-intercept: 0.497 slope ( f ) : 10.420 r2: 0.9990

x-intercept: 0.592 (max Oj measurable r a n g )
under the same carefully controlled condtions (v 1-
Tnblcs Ila and Ilb show data sets for 2 sources of 2%) demonstrated in Table I. T h e decrease in the
indigo where the sensitivity coefficient is less than absorbance of indgo was measured using 2 cm
0.42 L mg!' cm!'. Each calibration was performed measuring cells.
 Residual Ozone Measurement: Indigo Sensitivity Coefficient Adjustment 19

Table 11. Typical Data Sets for Determining the Sensitivity Coefficient Using the Ozone Calibration Technique

A. Aldrich I f = 0.390)
0 3 0 3 Calculated e
(mg/L) (moles&) Absorbance absorbance (M-' cm-')
blank --- 0.452 --- ---
0.0588 1.22 x 0.401 0.05 1 20.81 6

0.1177 2.45 x 0.355 0.097 19,779

0.1765 3.68 x 0.316 0.136 18,493

0.2354 4.90 x 1 0 . ~ 0.267 0.185 18,862

Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

0.2942 6.13 x 0.219 0.233 19,007

AVG 19,391 * 827

y-intercept: 0.450 slope (f ): -0.390 2: 0.9989

x-intercept: 0.577 (max 0, measurable range)

0 3 0 2 Calculated e
(mg/L) (molesn) Absorbance absorbance (M-' cm-')

AVG 17.395 i 228

y-intercept: 0.394 slope (f): -0.364 2: 0.9995

x-intercept: 0.541 (max O3measurable range)

The data in Tables IIa and IIb also show a linear absorbance of the indigo reagents. The extension of
decrease in the absorbance of the indigo solution in the calibration line to the x-axis shows the maximum
the presence of residual ozone. These solutions have calibration range for each reagent. These data
good agreement between the y-intercept of the -
correlate well with the range limit (0 0.6 mg/L 0,)
calibration line and the blank absorbance. However, in Standard Method 4500-0,B. It should be noted
the slope of the calibration line does not equal 0.42 that the range of the method can be increased by using
for either indigo source. In the case of the Aldrich a smaller sample size or with appropriate sample
reagent, the difference is 7%. In the case of the dilution.
Riedel reagent, the difference is 13%.
Figure 1 also has 2 interesting characteristics. The
A graph of the 3 data sets is shown in Figure 1. The data show that as the sensitivity coefficient decreases,
negative slope demonstrates the decrease in the the useable range of the reagent decreases (the x-axis
 20 G. Gordon and B. Bubnis
intercept). Thus, in principle the method range for an purity indigo has a molar absorptivity of 23,800
indigo source can be calculated. More importantly, M-'cm-'. (Commercially available indigo typically
the graph shows that the initial absorbance of the 3 has a molar absorptivity <20,500 ~ ~ ' c m - Based
' ) . on
indigo sources differ. Each of the solutions had a the Beer-Lambert Law, the blank absorbance of the
starting concentration of 1.25 x M (a 11100 analybcal solution prepared from high purity indigo in
dilution of a solution containing 770 mg/L a 1-cm measuring cell would be
commercial indigo reagent). This feature implies that
the molar absorptivity ( e ) is not constant for all indigo Absorbance = molar absorptivity x pathlength
sources. x indigo concentration
As a note in context, the sensitivity coefficient for the
Riedel indigo was first determined in 1983184 and
found to have a value of 0.42 L mg-'cm-'. The recent
evaluation of the reagent determined the average
Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

sensitivity coefficient to be 0.363 -0.364 L mg-'cm-' -

- a decrease of about 13-14%. The observed In an early paper (I), HoignC notes that the purity of
deviation from 0.42 L mg"cm-' suggests that solid indigo from microanalysis and titration is 80%. An
indigo decomposes and emphasizes the need for a apparent stoichiometric factor of 1:0.8 (compared to
"calculation" of the sensitivity coefficient as proposed 1: 1) for the decrease in indigo absorbance by residual
in the early indigo papers (1,2,3,4). ozone is also noted. These statements suggest that
indigo purity could be important.

The sensitivity coefficient defined in the Standard

Method 4500-0,B is based on the decrease in indigo
absorbance in the presence of 1 m g L residual ozone
(Table I). Thus, as HoignC notes an apparent molar
absorptivity of 20,000 L mol-'cm-' (from the ozone
calibration method) corresponds to a sensitivity
coefficient equal to 0.42 L mg-'cm-' as shown below.
The molar concentration of 1 m f l O3is 2.08 x lo-'.

F = molar absorptivify x pathlength x indigo

concentration [4]
Ozone (mg/L)
mol mg
2.08 x 10.'- x 20,000 L mol-'cm-'=0.416 cm" per y
Figure 1. Decrease in indigo absorbance - ozone
calibration method
As shown in Tables IIa and IIb, the sensitivity
coefficient is not constant for all indigo sources. The
Standard Method 4500-03B recommended procedure for measuring the sensitivity
coefficient is to apply the Beer-Lambert equation (A
Standard Method 4500-03B is based on a 1:1 reaction = E ~ c ) .In other words, a plot is made of indigo
stoichiometry of ozone with indigo. The preparation absorbance vs. residual ozone concentration. The
of the indigo stock solution in the standard method slope of the line ( ~ b equals
) the sensitivity coefficient
requires 770 mg/L of commercial indigo reagent (1.25
as shown in Figure 1. It is important to note that as
x lo-' M). A 1/100 dilution of the stock solution
the initial absorbance of each indigo source presented
(1.25 x lo-' M) is used for the residual ozone
in Figure 1 decreases, the corresponding calculated
measurement. sensitivity coefficient (i.e. slope) also decreases.
There is some confusion in the literature regarding the Because the sensitivity coefficient (f ) is directly
origin of the indigo molar absorptivity chosen for the proportional to the molar absorptivity ( e ) , we propose
reaction with ozone. Preisler reports (9) that high
that the initial absorbance of an indigo solution can be
 Residual Ozone Measurement: Indigo Sensitivity Coefficient Adjustment 21

used to calculate a statistically significant "apparent" All stock and working indigo solutions should be
molar absorptivity which is proportional to the stored in the dark.
sensitivity coefficient.
Reagents
The objectives were:
To examine the relationship between the a. lndigo stock solution: An indigo stock
molar absorptivity ( e ) and the sensitivity solution is prepared (10) by dissolving 770 mg of
coefficient (f) of indigo solutions commercial potassium indigotrisulfonate reagent into
To evaluate the E for variety of indigo a 1L flask containing 500 mL of distilled water and 1
sources mL of concentrated phosphoric acid and diluting to
To compare the E determined by multiple volume with distilled water.
initial absorbance measurements with the E
b. lndigo reagent 11: A solution is prepared in a 1L
Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

calculated by ozone calibration curves

To estimate the potential error resulting from flask by adding 100 mL of the indigo stock solution
the use of non-standardized indigo with 10 g sodium dihydrogen phosphate, NaH,PO,
To assess the stability of indigo solutions and 7 mL concentrated phosphoric acid. The solution
To discuss the implications of using should be stirred and diluted to volume. This solution
decomposed solid indigo reagent and stock should be stored in the dark.
indigo solutions
lndigo Stock Solution Molar Absorptivity
Calculation
NOTE: For convenience, the term molar absorptivity
is used without dktinction. The authors 10 mL of each indigo working stock solution was
recognize that the molar absorptivity pipetted into a 100 rnL volumetric flask and diluted to
calculated from initial absorbance volume. The absorbance of the solution was
measurements b an "apparent" molar measured a minimum of 5 times at 600 nm in a 2 cm
absorptivity based on the initial indigo molar measuring cell and the molar absorptivity calculated.
concentration and its purity. Equation 3 can be used to calculate the molar
absorptivity of an indigo solution as shown below for
the Fluka indigo (shown in Table I and Figure 1).
Experimental

All solutions were prepared using double distilled Absorbance

Molar Absorptivity =
deionized water @DW) unless otherwise noted. All pathlength x indigo concentration
chemicals were used without fiuther purification.

Preparation of lndigo Stock Solutions

An indigo stock reagent for each source was prepared

following the Standard Method (10). 770 mg of Sensitivity Coefficient Calculation: Ozone
commercial potassium indigotrisulfonate reagent is Calibration Curve
dissolved in 500 mL of distilled water and 1 mL of
concentrated phosphoric acid in a 1L flask. The The calculation of the sensitivity coefficient from an
solution is mixed and diluted to volume with distilled O3calibration curve is tedious and time consuming.
water. Ozone readily decomposes and is volatile. Thus for
the experiments reported here, prior to each
Summary of stock solution concentrations: calibration fresh 0, was generated directly in a
shrinking bottle containing a pH 4 acetic acid buffer
Stock solution 1.25 x M (1 1). Because there is zero headspace in the shrinking
Working stock solution (1110 dilution)l.25 x M bottle, problems caused by O, volatility are
Analytical concentration 1.25 x M eliminated. The absorbance of the O3stock solution
 22 G. Gordon and B. Bubnis
was measured at 260 nm and the concentration Table III. Indigo Sensitivity Coefficients (I)
--- 0,
calculated using a molar absorptivity of 3,000 Calibration Method
~ - I c m -(5).
'
Indigo Source Calculated e Avgf
Ozone calibration curves were prepared by pipetting Aldrich 1999 20,529 0.27 i 0.006
I0 mL of the working indigo stock solution into a 100
Fluka 1998 20,096 0.418 i 0.008
mL flask and adding appropriate aliquots of the 0,
stock solution. The absorbance of each O3 "standard" Acros 1999 19,471 0.405 t 0.002
was measured at 600 nm. By plotting the Absorbance Aldrich 1998 18,798 0.391 i 0.005
vs. mg/L O,, a calibration curve can be drawn and the
Acros 1998 18,461 0.384 i 0.006
sensitivity coefficient calculated:
Riedel1983 17,452 0.363 i 0.007
calibration curve slope Siama 1998 17,308 0.360 i 0.001
Sensitivity Coefficient ( f ) = PI
Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

cell pathlength

Results and Discussion

As a matter of convenience, indigo solutions

will be discussed in the context of their molar
absorptivi& (c ) to more accurately refect
the d~fferencesbetween the variour indigo
sources in the context of this paper.

Molar Absorptivity of Multiple Indigo Sources

- o3Calibratlon Method
The molar absorptivities for 7 indigo sources are
presented in Table 111. Table I11 also tabulates the Sensitivity Coefficient
sensitivity coefficient calculated by the ozone
calibration method for each indigo source.
Figure 2. Molar absorptivity vs. sensitivity coefficient - 0,
The molar absorptivities presented in Table 111 are calibration method
averaged values from a minimum of 3 data sets. The
uncertainty of the measurement is < 2%. These data
indicate that the slope of a plot of the decrease in Because the calculation of the sensitivity coefficient
indigo absorbance vs. added ozone is not constant. In using ozone calibration curves is a tedious process,
other words, the indigo molar absorptivity changes laboratories have been reluctant to standardize their
and the sensitivity coefficient can deviate from the indigo solutions and have accepted de facto that the
generally accepted value of 0.42 L rng-Icm-'. sensitivity coefficient is 0.42 L mg-'cm-'. The
potential benefit to measuring the molar absorptivity
Figure 2 is a plot of the data in Table 111. These data of indigo solutions is that laboratories can easily
suggest that the indigo sensitivity coefficient is "standardize" the indigo and thus improve the
proportional to the molar absorptivity of the indigo accuracy of the residual ozone measurement.
solution. This means that there are 2 ways to
calculate indigo sensitivity coefficients: ozone
calibration and molar absorptivity. The indigo Molar Absorptivity Determination Using the
sources whose molar absorptivities are below 18,000 Indigo Blank Absorbance The molar
~ - ' c m - are
' "boxed" to focus attention on the fact absorptivities for 7 indigo sources by the initial
that indigo molar absorptivities in this region are absorbance method are presented in Table IV and
suspect. Figure 3.
 Residual Ozone Measurement: Indigo Sensitivity Coefficient Adjustment 23

Table IV. Indigo Molar Absorptivities (E) --- Initial Table V. Comparison of Indgo Molar Absorptivity
Absorbance Method
-

Indigo Source Avg E Calculated f Avg E Avg f %

Indigo 0, Calibration Absorbance Difference
Aldrich 1999 20,355 * 37 0.423 Source
Fluka 1998 19,987 i 379 0.416 Aldrich 1999 20,529 20,355 1
Acros 1999 19,043 i 121 0.396 Fluka 1998 20,096 19,987 1
Aldnch 1998 18,067 i 82 0.376 Acros 1999 19,471 19,043 3
Acros 1998 17,707 + 154 0.368 Aldrich 1998 18,798 18,067 4
15,800 * 182
Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

Riedel 1983 0.329 Acros 1998 18,461 17,707 5

Sigma 1998 17,326 * 94 0.360 Sigma 1998 17,452 17,326 1
Riedel 1983 17.308 15.800 10

Sensitivity Coefficient
(Thousands)
From Initial Absorbance Method
Figure 3. Molar absorptivity vs. sensitivity coefficient -
Initial Absorbance Method
Figure 4. Indigo Molar Absorptivity - 0,Calibration vs.
Comparison of Os Calibration and Initial Initial Absorbance
Absorbance Methods
The data in Table V and Figure 4 have a number of
Molar absorptivities calculated by the ozone interesting features.
calibration and initial absorbance methods are
presented in Table V. The dashed line in Table V 0 with the exception of the Sigma reagent, the
indicates the 18,000 ~ - ' c m - '"cutoff' for indigo molar absorptivity of the remaining 6 indigo
acceptability. The calculated molar absorptivities for sources show a linear relationship between
each method are compared in Figure 4. A line of the 2 methods, y = 0 . 6 7 0 4 ~+ 6736
equality is presented to indicate 100% agreement 0 the Sigma reagent "falls on" the line of
between the 2 methods. equality between the methods
0 at E values >18,000, the agreement between
the methods is within 2%
at E values 4 8 , 0 0 0 , the difference between
the methods can be 2 10%
 24 G. Gordon and B. Bubnis
Because the impact of misreporting residual ozone with Figures 1 and 2 clearly show that the
concentrations is considerable (in terms of costs and sensitivity coefficient is not constant for all
confusion), the routine standardization the indigo sources of indigo.
reagent takes on added importance. However, it is
generally acknowledged that many utilities applying This fact is not a crisis. It supports the
ozone would be resistant to using the ozone original conclusion noted by Bader and
calibration method for routine indigo standardization. Hoignt (1-4) that the accuracy of the indigo
Thus, Rakness proposes (12) the concept of an method requires that the indigo be calibrated
adjusted sensitivity coefficient (f-adjusted) based on or standardized.
the molar absorptivity calculated from the absorbance
of the indigo blank solution (Ab& B. Indigo Concentration lmplications
Because the reaction stoichiometry is 1 :I,
Error Evaluation a sample containing 1 m@ O3 (2.08 x
Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

lo-' M) requires 2.08 x lo-' M indigo.

Table VI compares the indigo sensitivity coefficients In other words, the amount of indigo in
calculated using a constant sensitivity coefficient the reaction flask is critical.
(Standard Method) with the sensitivity coefficient
calculated from the indigo solution molar absorptivity. For example, the analflcal concentration
of indigo in Standard Method 4500 is
1.25 x lo-' M or 60% of the amount of
Table VI. Sensitivity Coefficient Comparison (L indigo required to react with 1 mg/L Op.
mg-'cm-') The implication is that the upper limit of
the method range is 0.6 m@ 0,. By the
E Constant Indigo same reasoning (assuming a linear
(M-'cm-l) 0.42 Absorbance relationship), if the molar absorptivity of
an indigo solution is 10,000 M-'cm-I, the
upper limit of the method range is only
50% of the specified range (0.3 mg/L
0,).

Table W. Standard Method 4500-0,B --- Measurable

Range as a Function of c and "Useable" Indigo (f= 0.42 L
mg-'cm-I)

Maximum
c "Useable"1ndigo Measurable Range
(M-'em-') (M) ( m f l 0,)

The data in Table VI should be reviewed carefully in

the context that the stoichiometry of the indigo-ozone
reaction is 1:1.

A. Sensitivity Coefficient Implications The

accuracy of Standard Method 4500-03B is
based on the assumption that all sources of
indigo have a molar absorptivity of - 20,000
~ " c m - 'determined by the ozone calibration
curve method. The data in Tables 1-4 along
 Residual Ozone Measurement: Indigo Sensitivity Coefficient Adjustment 25

Table VII shows the corresponding useable Table VIII is an expansion of the example
concentration of indigo and the upper limit of residual case showing the calculated residual ozone
ozone that can be measured using Standard Method concentrations at varying indigo molar
4500-03B resulting from a change in indigo molar absorptivities.
absorptivity.
The example data in Table VIII provides a
C. Sensitivity Coefficient Calculation number of interesting observations.
Below is an example showing the difference
in the calculated O3 concentration using a As the indigo molar absorptivity
constant f (= 0.42) compared to using a decreases, the range of the method
calculated f by indigo absorbance. Assume decreases as shown in Figure 1
that the molar absorptivity of the indigo Calculations based on a constant 0.42
source is 18,000 M-'cm-' and the estimated L rng7'cm-' show no change in the
Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

residual ozone is -0.5 m a . Under these calculated 0 3 concentration. In other

conditions, each calculation method should words, the calculated O3
have an adequate indigo concentration to concentration can exceed the upper
meet the stoichiometry requirements. In this concentration limit based on
example, the concentration of residual ozone available indigo
will be determined using the following 0 3 concentrations based on the
sensitivity coefficients: absorbance of the indigo blank
increase as the molar absorptivity
Standard Methodr: 0.42 Indigo Absorbance: 0.374 decreases (i.e. f in the denominator
of Equation 2).
Assume the following parameters: At E = 18,000 M-'cm-', the
agreement in the calculation is most
Absorbance blank: 0.225 Tare weight: 90 g probably within the error of the
Total volume: 1 OOmL method.
Absorbance sample: 0.205 Sample volume: 10 mL At E < 18,000 M-lcm-', the error in
the calculation most probably
exceeds the error of the method.

D. Utility implications
Insertion of each sensitivity coefficient gives
the following: The use of unstandardized indigo at a utility
applying ozone has significant financial and
by-product formation impact. For example, if
f Q4 the indigo sensitivity coefficient is 0.36 L
Standard Methods: 0.42 0.48
Indigo Absorbance: 0.374 0.54 mg-'cm-' instead of 0.42 L mg-'cm-', the
residual ozone concentration will be
underreported by - 15%. This means that in
Based on this evaluation, the uncertainty in some cases utilities are adding too much
the calculated residual ozone concentration ozone and the ozone dose can be reduced
can be 0.06 m g L without losing CT credit. A reduction in the
ozone dose also has the added benefit of
It is important to recognize that the effect lowering the potential to form brornate ion
of a decreasing sensitivity coemcient is an (BG).
increase in the calculated residual ozone
concentration because the f value is in the
denominator.
 26 C. Gordon and B. Bubnis

Table W I . Residual Ozone Calculation Comparisons

e f = 0.42 Variable f YO Method Limit

(M-lcm-I) ( m f l 01) ow- 03) Difference ( m f l 0,)
20,000 0.48 0.48 --- 0.60
Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

Absorbance blank: 0.225 Tare weight: 100 g Total volume: 100 mL

Absorbance sample: 0.205 Sample volume: 90 mL

At a recent IOA/PAG conference, Rakness Stability of Indigo Solutions (6)

(12)showed examples from utility data that
the molar absorptivity of the indigo solutions The Rakness utility data c o n f m s that the stability of
was decreasing (reaching a low of -6,000 indigo solutions is variable. In our studies, 3 sources
~ - ' c m - ' ) . The following observations were of indigo were evaluated to determine their stability.
made: The data are presented in Table IX.

The molar absorptivity of the indigo

solution is changing over time.
The residual ozone concentration Table IX. Indigo Solution Stability Based on Molar

mg-'cm"
-
calculated usine a constant 0.42 L
does not deviate
Absorptivity

substantially regardless of the Initial Day 1 Day 2 Day 4 Day 10

indigo molar absorptivity. The
Dye 1 20,355 20,133 19,991 19,653 18,320
difficulty with this measurement is
% Decrease 1.09 1.79 3.45 10.60
that the operators were changing
the 0, dose. The operators had no Dye 2 19,043 18,829 18,664 18,281 17,462
indication that the demand of their %Decrease 1.12 1.99 4.00 8.30
high quality water source (Lake Dye 3 17,326 17,204 17,002 16,650 15,697
Michigan) changed. % Decrease 0.70 1.87 3.90 9.40
The residual ozone concentration
calculated using an adjusted f
resulted in an increase in measured
residual ozone that appeared to
correspond to changes in the ozone
dose.
 Residual Ozone Measurement: Indigo Sensitivity Coefticient Adjustment 27

The data show that the average decrease in solution Derivation

molar absorptivity is -1% per day. This means that
"indigo solutions" have a limited shelf life. If a molar Equation 2 is used to calculate the residual O,
absorptivity lower limit of 18,000 ~ - l c r n -is' applied, concentration by the indigo gavirnetric procedure.
solutions prepared from Dye 1 should be discarded By substituting Equation 4 for the f factor, the
within -2 weeks; solutions prepared from Dye 2 equation now becomes
should be discarded within 6 days. We recommend
that all indigo solutions have a minimum molar
absorptivity of 18,000 M-'cm-I. On this basis, Dye 3
should be replaced with a higher purity solid.

Conclusions To remove the molar absorptivity term, Equation 3 is

substituted
Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

Recent data show that the indigo sensitivity

coeffkient is not constant. These same data show that
the sensitivity coefficient can be calculated based on
the initial absorbance of the indigo reagent provided
the solution has a minimum molar absorptivity of
18,000 M-'cm-'. Because the measurement of the By incorporating the concentration of indigo (1.25 X
sensitivity coefficient using ozone calibration curves 10.' M) and canceling out the pathlength term, the
is a tedious process, an equation to calculate 0, equation simplifies to
directly is proposed based on the initial absorbance of
the diluted indigo reagent in Standard Methods 4500-
03B.

Example calculations are provided that demonstrate

the implementation of initial absorbance data for
determining the residual ozone concentration Further simplification provides the following equation
according to Standard Method 4500-03B. for the initial absorbance calibration method.

Proposed Calculation (e > 18,000 M % r - ' )

The following equations are valid if the amount of

indigo in the reagent and blank solutions is precisely
known. We recommend that dry commercial indigo Example Calculations: f = 0.42 mg'lcm-'
be periodically calibrated by weighing the commercial Compared to Initial Absorbance
indigo reagent (in triplicate) and precisely
determining the initial absorbance. A sample calculation is presented below for the
spectrophotometric gravimetric procedure. When the
indigo method is performed exactly as outlined in
Standard Methoak, an initial absorbance of 0.498
yields a sensitivity coefficient of 0.42 L mg-'cm-'
using a 2 cm measuring cell. As proposed, the
where: calculation of the 0, concentration using Equation 2
AB,As absorbance of blank and sample respectively (Standard Method 4500-0,B) and Equation 9 (initial
vs volume of sample, mL = [(final weight - tare absorbance method) should be identical. The small
weight) g x l .O m u g ] - 10 mL difference in the calculated results are due to the
VT total volume of sample plus indigo, mL = rounding of 0.416 to 0.42 L mg-'cm-I.
(final weight - tare weight) g x 1.0 mL/g
 28 G . Gordon and B. Bubnis
Bader, H.; Hoignb, J. "Determination of Ozone in
Water by the Indigo Method; A Submitted
Standard Method", Ozone Sci. & Eng., 4:169
(1982).
Equation 2 (Standard Method 4500-0,B):
Hoignb, J.; Bader, H. "Bestirnmung von Ozon
und Chlordioxid in Wasser mit der Indigo-
Methode", Sonderdruck aus Vom Wasser, 55,
261-280 (1980)

lntemational Ozone Association Revised

Guideline Document - 1998. "111. Colorimetric
Method for Manual Determination of Ozone
Equation 9 (Initial Absorbance Method):
Concentration in Water". Ozone Sci. & Eng., Vol.
Downloaded by [University of Saskatchewan Library] at 01:55 23 February 2013

20.6, p.443.

6. Gordon, G.; Gauw, R.D.; Miyahara, Y.;

Bubnis, B. "Calculation of -the Ozone
Sensitivity Coefficient Based on Indigo
Calculations for Variable Grams of Indigo Absorbance", Journal A WWA, Vol. 92(12), p.96
(2000).
Indigo stock solutions can have a small deviation in
Gordon, G.; Walters, B.; Bubnis, B. "The Effect
the amount of indigo used to prepare the stock of Indigo Purity on Measuring the Concentration
reagent. For example, the method calls for 770 mg of of Aqueous Ozone", Proceedings of the
commercial indigo reagent to be dissolved into 1 liter International Ozone Association -Pan American
of deionized water. If the measured weight is 0.775 g, Group Conference ZOO0 (Orlando).
how can this "extra" indigo be accounted for?
Silverman, R.A.; Gordon, G. "Use of the Syringe
Equations 10 and 11 can be used to adjust for small as a Shrinking Bottle", Anal. Chem., 46:1:178
differences in the measured weight of indigo during (1974).
the preparation of the indigo stock solution.
Preisler, P.W.; Hill, E.S.; Loeffel, R.G.; Shaffer,
Gravirnetric Procedure: Ph.A. "Oxidation Reduction Potentials, Ionization
Constants and Semiquinone Formation of Indigo
( AB x 1 0 0 ) - ( As x 100 ) grams indigo Sulfonates and Their Reduction Products", J. Am.
rng OJ/ L = x Chem. SOC.,81, 1991-1995 (1959)
A ~ 1X . 6 6 4 ~ ~ ~ 0.77
[lo1 10. Standard Methods for the Examination of Water
and Wastewater, 18" Edition. 1992, Method
Volumetric Procedure: 4500-0, B - 1ndko ~olorirnet& Method, p. 4-
106.
100 ( - A ~ ) grams indigo
rngOJ/L= u11
As x 1.664 x V s 0.77 Hoigne, J.; Bader, H. The Role of Hydroxyl
Radical Reactions in Ozonation Processes in
Aqueous Solutions. Water Res, 10:377 (1976).
References
Rakness, K., Gordon, G., Bubnis, B., Rexing, D.,
1. Bader, H.; Hoigne, J. "Determination of Ozone in Wert, E., Tremel, M. "The Impact of
Water by the Indigo Method". Water Res., 15:449 Underestimating Dissolved Ozone Residuals
(1981). Using Standard Method 4500-Ozone and
Outdated Indigo", IOAIPAG Conference,
2. Bader, H.; Hoignk, J. Ozonation Manual for Newport Beach, CA, (2001).
Water and Wastewater Treatment, Chapter 38.
John Wiley & Sons, New York (1982).