>A 660/2-73-039
ugust 1974
                             Environmental  Protection Technology  Series

                                         Office • of Research and

-------
            RESEAKCK REPORTING SERIES
Research reports of  the   Office  of  Research  and
Monitoring,  Environmental  Protection Agency, have
been grouped into  five series.   These  five  broad
categories  were established to facilitate further
development  and   application   of   environmental
technology.   Elimination  of traditional grouping
was  consciously   planned  to  foster   technology
transfer   and  a  maximum   interface  in  related
fields.  The five  series  are:

   1.  Environmental Health Effects Research
   2.  Environmental Protection Technology
   3.  Ecological  Research
   1.  Environmental Monitoring
   5.  Socioeconomic Environmental studies

This report has been assigned to the ENVIRONMENTAL
PROTECTION   TECHNOLOGY   ; series.    This   series
describes   research  performed  to  develop  and
demonstrate   instrumentation,     equipment    and
methodology  to  repair   or  prevent environmental
degradation from point and   non-point  sources  of
pollution.  This work provides the new or improved
technology  required for  the control and treatment
of pollution sources to meet environmental quality
standards.
                   EPA REVIEW NOTICE

The Office of Research and Development lias reviewed this
report and approved its publication.  Mention of trade names
or commercial products does not constitute endorsement or
recommendation for use.

-------
                             ABSTRACT
This report is concerned with  modification of  the amperometrie titration
method for measuring chlorine  residuals  in cooling  tower blowdown.  This
modified procedure can  be applied  to  other water systems with a concen-
tration of metal  ions similar  to those  found in cooling tower blowdown.
The addition of sodium  pyrophosphate  as  a complexing agent  removes the
interferences contributed by Fe  (III) and Cu (II) in the water matrix.

Procedure recommendations are  made to increase the  efficiency of both
sampling and the actual  titration  procedure in order to allow a residual
determination in the minimum amount of  time.   Equipment recommendations
along with a design of  a biamperometrie  end point system which allows
greater titration speed  along  with portability are  described.
                                  111

-------
                             CONTENTS
                                                            Page
Abstract                                                     iii
List of Figures                                              v
List of Tables                                               vi
Acknowledgment                                               vii
Sections
I      Conclusions                                           1
II     Recommendations                                       2
III    Introduction                                          3
IV     Method Familiarization                                7
V      Characterization of Interferences                     11
VI     Biamperometric End Point                              30
VII    References                                            37
VIII   Appendix                                              39
                                  iv

-------
                              FIGURES

No.                                                      Page

1.   Meter Drift vs Time                                 17

2.   Calibration Curve for Total  Chlorine Residual        27

3.   Chlorine Probe (DPE)                                 31

4.   Schematic Diagram of DPE Control  Box                32

5.   Background Current vs Applied Voltage Iron           34
     Interference Study

6.   Background Current vs Applied Voltage Copper        35
     Interference

-------
                              TABLES
No.                                                      Page
 1   Recovery of Free Chlorine Residual  (pH7)               8
 2   Recovery of Free and  Combined  Chlorine  in             10
     Chlorine Demand-Free  Water
 3   Loss of Chlorine in Distilled  Water as  a              12
     Function of Stirring
 4   Chlorine Stability in Distilled  Water  in              14
     an Unstirred Sample
 5   Chemical Analysis of  Slowdown  Water                  16
 6   Effect of Cu (II) on  Recovery  of                     19
     Chlorine Residual
 7   Effect of Citrate on  Copper Interference              19
 8   Effect of Citrate on  Iron Interference                22
     in Determination of Chlorine Residual
 9   Recovery of Cl  in the Presence of Kaolin              22
10   Recovery of Cl  Spikes on Cline's Pond  Sediment        24
     and Corvallis Clay
11   Recovery of Cl  Additions in tKe  Presence of  a         24
     Water Treatment Chemical as a  Function  of Time
12   Precision of Titration                               26
13   Chemical Analysis of  Natural Water Samples            29
14   Recovery of Chlorine  Spikes on Blowdown              29
                                 VI

-------
                         ACKNOWLEDGEMENT

The author would like to thank the personnel  of  the  Laboratory Services
Branch at the Pacific Northwest Environmental  Research Laboratory  for
their assistance to the project.   Personal  thanks  are given to David
Maier who helped extensively in the experimental portion of the project,
Daniel F. Krawczyk, Chief Laboratory Services  Branch, for  valuable
advice and orientation and John Jacobson  for  the construction of needed
equipment.
                                vii

-------
                            CONCLUSIONS

Experimental results have shown that the analytical  difficulties  in
the application of the amperometric technique to cooling  tower  blow-
down can be eliminated by the addition of a pyrophosphate reagent to
the currently used amperometric procedure.   Copper and iron  in  con-
centrations found in typical blowdown do not interfere in the presence
of pyrophosphate.

Suspensions of sediments exhibit stable end points with good recovery
of incremental additions of chlorine once the initial  chlorine  demand
of the system is satisfied.

Various synthetic blowdowns are evaluated as to stability of the  end
point.  All of these exhibit stable end points with  good  recovery of
the chlorine spikes once the demand constituents were  satisfied.   The
effect of agitation when using a commercial  instrument does  cause an
appreciable error in results if the titration is not performed  rapidly.
The source of the error may be due to actual flashing  off the residual
chlorine or enhanced photodecomposition resulting from an increase in
surface area provided by the agitation.

A modified biamperometric titrator is described and  evaluated.  When
using this titrator, results on the recovery of chlorine  along  with
precision of the measurement are comparable to that  of a  commercially
available instrument.  The dual electrode system offers numerous  ad-
vantages for field operation in that a liquid reference electrode is
no longer needed.  The system is battery operated eliminating the need
for an AC power source.   The big advantage  is the rapid determination
of the end point.  The meter response is faster for  the biamperometric
method over the commercial  instrument.   This permits more rapid addition
of titrant and easily meeting the goal  of performing a titration  in less
than 1.5 minutes.

-------
                          RECOMMENDATIONS

Further tests of the pyrophosphate  system  should be made on site on
actual blowdown systems to check  its  effectiveness in the elimination
of interferences due to metal  ions.   Feedback  from other laboratories
would be helpful in determining whether the  pyrophosphate system would
be effective in most blowdown  matrixes. Due to the lack of time com-
plete evaluation of the bioamperometric endpoint detector was not made.
Further tests on its application  for  the analysis of total chlorine
residuals should be made, perhaps by  other interested laboratories.
If the pyrophosphate system is not  adequate  in some systems, a mixture
of complexation reagents should be  looked  into.  A precipitation re-
agent could also be used.

-------
                           INTRODUCTION

The goal of the project is the development of a  simple,  accurate  method
for the analysis of residual  chlorine by the quantification  or elimi-
nation of certain interferences typically present in  cooling tower
blowdown.  Suspected interferences  are turbidity, copper,  iron and the
chlorine reaction with phosphonates and polyesters in the  water treat-
ment chemicals used as corrosion inhibitors.

There are basically two classes of  chlorine analysis, colorimetric and
titrametric.  The problem with colorimetric procedures is  the use of
comparison standards along with a spectrophotometer or color comparator.
The colorimetric class of analysis  requires the  ability  to differentiate
the ways light interacts with the colored species in  a sample.  The
physical characteristic of the blowdown water, having a  high degree of
turbidity and suspended solids, does not allow easy discrimination by
an instrument or the human eye.  The field stability  of  colored stan-
dards is also highly unreliable. For these reasons a titration method
is a more reasonable approach in the field than  a colorimetric method.

There are at present two titration  methods available  in  Standard
Methods (1) and ASTM (2).  The amperometric end  point technique is the
Referee Method in ASTM (2).  This involves the reduction of  an active
halogen species, either hypochlorite ion in the  free  residual  test or
liberated I~ in the combined  residual  determination,  by  phenylarsine
oxide (PAO).

-------
     CgH5AsO + HOC1  + H20 -»• CgH5AsO(OH)2  +  HC1

     CgH5AsO  + I2 + 2H20 -> CgH5AsO(OH)2  +  2HI

By lowering the pH to 4 the addition of iodide  produces  free  I,,  in
solution by oxidation of I" by both  the free hypochlorite  and combined
chlorine not available as chloride.   The  effectiveness of  the phenyl-
arsine oxide as a titrant is pH independent.

Also mentioned is a tentative procedure using ferrous  ammonium sulfate
as the titrant with DPD (N,N-diethyl-p-phenylene diamine sulfate) as
an end point indicator.  This procedure has certain  features  which
detract from its desirability in the field  application to  blowdown
water.  One flaw is the chemical instability of the  titrant which
necessitates frequent standardization.   The second flaw  is that  in
extremely colored water systems with a  high degree of  suspended  solids
along with high turbidity the end point is  not  sharp and easily  recog-
nized.  In clear water systems, however,  the DPD titration technique
allows differentiation among all active chlorine forms at  the same  pH.
This is a big advantage over the amperometric method which requires a
change in pH from 7 to 4 to detect free and combined chlorine residual.
An abrupt change in pH can change the chemical  eequilibria thus  dis-
torting the true free and combined chlorine fractions  (6).

Certain weaknesses of the amperometric titration are mentioned by
Nicolson (4) and Hasselbarth (5).  The primary  one is  the   flashing of
chlorine out of solution or enhanced photodecomposition  by the stirring
action of the amperometric titrator.  Another deals  with the  incomplete
reaction of the phenylarsine oxide titrant with the  chlorine  residual.
Hasselbarth suggests the addition of a catalytic quantity  of  iodide to
allow the titrant to react completely.   He notes that  monochloramine
does not increase the diffusion current between the  electrodes nor does
monochloramine require any titrant after the addition  of KI solution.

-------
His work, however, conflicts with the procedure in Standard Methods
for the determination of monochloramine.   Standard Methods  suggests
the addition of 0.2 ml of a 0.1N KI solution at the end point of the
free residual titration to obtain a monochloramine fraction.   The in-
terpretation of the results when KI is added is at this moment in ques-
tion.  Since this report deals with only  total  residual chlorine, it is
not necessary to resolve this question.

Guter and Cooper (6), in a paper evaluating various field test proce-
dures for the determination of free chlorine residual  in aqueous solu-
tions, suggest the following improvements in the amperometric procedure:

     1.   Exposing the samples to as little light as possible.
          (VERY IMPORTANT).

     2.   Adding a considerable quantity  (90 percent)  of the
          titration solution before turning on  the cell stirrer.

     3.   Completing the titration as quickly as possible.

Guter and Cooper (6) also note that the determination  of free residual
and combined residual on the same sample  gives  erratic results on the
free residual determination of subsequent samples.  They suggest rinsing
the cell four times with distilled, chlorine demand-free water between
each sample to remove all traces of iodine before another sample is  run.
This interference can be overcome by the  use of two separate titrators:
one for the free and one for the total residual chlorine fractions.
Thus, taking into account the physical characteristics of cooling tower
blowdown and the desirability of a fast,  simple technique for the deter-
mination of chlorine residuals, the amperometric titration  procedure
shows the greatest promise.  Reviews of alternate methods and their
limitations are given in papers by Marks  (7), Nicolson (4)  and Pal in (3).

-------
The experimental  work of the project is  divided into  three  phases.
Phase one is devoted to method familiarization.  Phase two  deals  with
an overview in which characterization of various waters as  to  major
cations, anions and nitrogen compounds is made.  Characterization
allows prediction of changes in chlorine residual  due to interfering
reactions.  Phase three involves the development of a fast, accurate
procedure free of the interferences rioted in cooling  tower  blowdown.

-------
                      MLTilOL) FAMILIARIZATION

Double distilled water as used in the Pacific tlorthwest Environmental
Research Laboratory had a negligible cnlorine demand as determined by
the amperometric titrator.  Tne water was subjected to chemical  analysis
weekly and the concentration of the major demand agent for free residual
chlorine, ammonia, was always less than 0.001 mg/1.

Tiie chlorine stock solutions were prepared from commercial sodium
hypochlorite solutions (Clorox-5.25 percent NaOCl).

Each ml of sodium hypochlorite in bleach contained  50 rng/ml  of active
chlorine.  A stock solution of 100 mg/1 was produced by dilution of 2
ml of bleach solution to one liter with cnlorine demand-free water.
Tne dilute chlorine stock solution was standardized against thiosulfate
as outlined in Standard Methods (8).

Aliquots of this stock were added to the amperometric titration vessel
containing (2x) distilled water.  Reagents were added as outlined in
Standard  Method 114B, and tne titration run.  The  amperometric titrator
used was a Wallace & Tiernan and the procedure employed is outlined in
Standard Methods (9).  The results are shown in Table 1.

-------
                              TALJLL 1
               Recovery of Free Chlorine  Residual  (pH7)

Cone, of Chlorine                Chlorine Residual           Net  Loss of
Added (mg/1)	Recovered (mg/1)	Chlorine  Residual  (mg/1)
     1.05                              0.96                     0.09
     1.05                              0.98                     0.07
     1.05                              0.95                     0.10
     1.05                              0.95                     0.10
     0.52                              0.48                     0.04
     0.52                              0.47                     0.05
     0.52                              0.48                     0.04
     2.10                              1.98                     0.12
     2.10                              1.97                     0.13

-------
As indicated in Table 1, tiie free residual as determined by the ampero-
metric procedure in the absence of iodide gives a lower value for the
free residual chlorine on the order of 6-10 percent.   Similar low
readings were presented in the Analytical Reference Service Report
Number 40 (10).

Ammonia in tiie form of ammonium chloride was added to known amounts of
a chlorine solution to check on the recovery of free and combined frac-
tions of the residual chlorine.  Results are shown in Table 2.

The importance of the data in Table 2 is that the added step to deter-
mine both free and combined residual  chlorine (1) caused consistently
lower recovery of total  residual chlorine.  Tin's lower reading is pro-
bably due to the flashing off of the combined chlorine residual during
the free residual determination.  This effect was also noted by Marks
(7).   Direct lowering the pH to four from the start and determining
only total residual allows more complete recovery of the chlorine
residual at a low pH.  The loss of Io has been reported to be neg-
ligible (7).

The effect of varying the concentration of ammonia and looking at the
free and combined residual fraction was not pursued due to tiie fact
that the project scope was concentrated only to total  chlorine resi-
dual.   However, the "flashing off" effect was investigated.

-------
                                TABLE 2

                 Recovery of Free and Combined Chlorine
                                  in
                       Chlorine Demand-Free Water
mg/1
Chlorine
Total
Added
2.04
2.04
2.04
2.04
2.04
2.04
2.04
2.04
2.04
2.04
2.04
2.04
2.04
2.04
mg/1
Ammonia
Nitrogen Added
0.00
1.00
1.00
1.00
1.00
1.00
2.00
2.00
1.00
1.00
1.00
1.00
1.00
1.00
mg/1
Free Chlorine
Residual Determined
1.98
__*
	 *
__*
__*
	 *
0.74
0.78
1.18
1.20
1.26
1.28
1.26
1.26
mg/1
Total Residual
Chlorine
Determined
	 *
1.98
1.90
1.88
1.97
1.90
1.80
1.88
1.74
1.83
1.60
1.84
1.81
1.83
*Determination not made
                                     10

-------
                 CHARACTERIZATION OF INTERFERENCES

TIME INTERFERENCE

Samples of chloramine T were used as a source of combined chlorine  resi-
dual.  A 10 ml aliquot of a 100 mg/1 solution of chloramine  T  was allowed
to stand in the titration vessel with stirring.    See Table  3.

An assumption was made that the rate of flashing is  proportional to the
concentration of the chlorine in the sample vessel.   This type of rate
                                                                 |/4-
law is exponential in nature.  The rate law has  the  form  f(t)  = be   ,
where t = elapsed time before the titration is performed  and b is the
initial concentration of total residual chlorine present. Note at  time
t=0 for a residual concentration of 1.20 mg/1

     f(t) = be"0 = b                                   (1)

     thus b = 1.20 mg/1

For t = 5 minutes, f(5) = 0.90 (from Table  3)  and

    0.90 = 1.20 e+K^5^                                 (2)

Solving for K yields the following expression  for the concentration
as function of time.

     fUJ-l^Oe-*0'058**                             (3)

To test the model another sample was allowed to  stir for  10  minutes.
Duplicate samples yielded 0.67 and 0.69 mg/1 with an average value  of
0.68 mg/1.
                                 11

-------
                         TABLE 3
              Loss  of  Chlorine in Distilled Water
                           as
                    Function of Stirring
                         Stirring                   mg/1
  mg/1                     Time                    Chlorine
Chlorine	(min)	Determined
 1.20                       0                       1.22
 1.20                       0                       1.17
 1.20                       5                       0.92
 1.20                       5                       0.89
 1.20                      10                       0.67
 1.20                      10                       0.69
                            12

-------
Using the model to predict what the value would  be  and  knowing starting
time yielded 0.67 mg/1.

     f(t) =1.20e-<°-058><10>=0.67                   (4)

Nicolson (4) states that after 90 seconds of  stirring he noted a 5
to 7 percent loss of chlorine.  Applying our  model  at t = 1.5 minutes,
a predicted loss of 8 percent is obtained, which compares quite favorably
to Nicolson values.  For a 30-second stirring time  before the addition
of titrant a 3 percent error is obtained.  This  experimental and theo-
retical treatment lends  support to the  improvement  stated by the U.S.
Army on the survey study of field test  kits  (6).  They  suggest the
addition of 90 percent of the titrant before  stirring action is ini-
tiated.  This drops the  chlorine or iodine concentration in solution
to 10 percent allowing a slower rate of flashing of the active species
in solution.  This, however, is very difficult to do in the field if
the exact end point is not known.  In this case  the titration should
be carried out as rapidly as possible.  Once  experience with a given
dynamic system is obtained, addition of 90 percent  of the titrant
should be practical.

Another time interference study was made.  A  chlorine residual stock
solution was prepared and eight samples were  poured at  once into 250
ml beakers.  The beakers were then stored in  the dark with one sample
pulled out at 5-minute intervals.  See  Table  4.   Total  chlorine residual
determination was run on the samples.

From the data in Table 4, it is apparent that without stirring in the
dark, the loss of chlorine residual is  negligible up to at least 35
minutes for chlorine demand-free systems.
                                  13

-------
                         TABLE 4
          Chlorine Stability in Distilled  Water
                  in an  Unstirred Sample
  mg/1                    Storage                   mg/1
Chlorine                  Time                    Chlorine
 Added	(min)	Determined
  0.30                      0                       0.30
  0.30                      5                       0.30
  0.30                     10                       0.30
  0.30                     15                       0.31
  0.30                     20                       0.30
  0.30                     25                       0.30
  0.30                     30                       0.29
  0.30                     35                       0.30
                            14

-------
METAL ION INTERFERENCE

It was thought that some interferences to the amperometric  procedure
might be due to heavy metal contamination in the blowdown water.   Chem-
ical analysis of both the makeup and blowdown water for  one cooling
tower gave the results in Table 5.  The ratio of the metal  ion  content
in the blowdown water to that in the makeup water was  approximately
10:1 — giving 10 cycles of concentration.   Metal  ions contained  in
the blowdown in appreciable concentration are Zn,  Mn,  Cu, and Fe.
Marks has stated that cadmium, trivalent chromium, divalent nickel
and zinc in concentrations up to 1000 mg/1  had no  immediate effect
on the amount of current flowing though the titration  cell  or poi-
soning the electrode surface 01)-  He has  also noted  that  copper  in
high concentrations (500 mg/1) poisons the  electrode.  Marks also
states that manganese in its higher oxidation states does not inter-
fere if the pH is above 3.5.

Based upon the literature data and heavy metal  ion content  known to
exist in blowdown water, two elements (Fe CHI), Cu (II)) were  present
in sufficiently high concentrations to cause a possible  interference.

These two metal  ions were added to distilled water to  investigate  their
possible interference.  Solutions of copper at concentrations as low as
0.05 mg/1 exhibit a poisoning effect after  two titrations.   The poi-
soning effect manifested itself by a reduction in  sensitivity and  in-
stability of the current flowing through the cell.   Copper  produced a
general  drift in the meter reading as a function of time as shown  in
Figure 1.  Copper also gave a false chlorine residual  when  none was
present.   See Table 6.

The meter deflections exhibited erratic behavior at the end point along
with the deposition of a crust on the platinum electrode.   Note that
                                    15

-------
                             TABLE  5
                Chemical  Analysis  Of  Slowdown Water
                       Cooling  Tower
                             I
Cooling Tower
     I-A
Cooling Tower
    II
Characteristic
QIC TOC1 (mg/1)
Kjel N2 (mg/1)
Ammonia N (mg/1)
Nitrite N (mg/1)
Nitrate N (mg/1)
Ortho P3 (mg/1)
Chromium (yg/ml)
Copper (yg/ml)
Iron (yg/ml)
Manganese (yg/ml)
Lead (yg/ml)
Zinc (yg/ml)
Makeup
22.4
0.38
0.007
0
0.52
0.011
< 20
20
440
10
—
—
Slowdown
87.1
2.4
0.067
0.025
6.0
1.4
< 20
200
6000
220
—
—
Makeup

< 0.01
< 0.001
0.003
0.031
—
10
5
39
4
< 10
68
Slowdown
5
0.4
0.0025
0.0525
0.765
—
26000
29
39
6
14
40
Bl owdown
„
—
0.0065
0.006
0.113
—
7600
11
240
1.2
240
1200
1  Total  organic carbon using Oceanography International  Corp.  instrument

2  Organic + anmonia nitrogen
                   3«
3  Phosphorus as PO*
                                   16

-------
  100
  90
  80
  70
Q 60
(T
LJ

UJ40
  30



  20



   10
     FIGURE  I
METER DRIFT VS. TIME
     Cu ~ 5mg/l
                                 I    I
        2530  60   90   120  150  180  210  240

              STIRRING TIME (sec.)
                     17

-------
for a residual  chlorine level  of 1  mg/1  one  series  gave low results
and another high results.   A cumulative  effect was  noted in which the
drift increased with each  determination.

Iron in concentrations of  5 mg/1  causes  a  similar loss in sensitivity;
however, no cumulative effect was seen as  in the case of copper.  Low
recoveries of chlorine were also obtained  (approximately 85 percent).

It was concluded that copper and iron did  interfere with the chlorine
residual test.   Other metal ions present may also interfere.  To eli-
minate these interferences a suitable complexing agent was sought which
would form a complexation  product that was inert under the chemical
conditions of the amperometric determination.  Complexing agents tried
were sodium citrate, NTA,  sodium oxalate and sodium pyrophosphate.

Sodium Citrate

A 20 percent solution of sodium citrate  was  prepared as a stock solution,
Aliquots of this stock solution were added to the titration vessel.
Addition of citrate to a 5 mg/1 copper solution stabilized the drifting
problem.  Table 7 shows a  data summary of  behavior  of copper on the
residual chlorine test.
                                  18

-------
                         TABLE 6

              Effect of Cu (II) on Recovery
                   of Chlorine Residual
Cl Cone.
mg/1
0
0
0
0.98
0.98
0.98
0.98
0.98
0.98
Cu Cone.
mg/1
0.05
0.10
0.15
0.05
0.05
0.10
0.10
0.15
0.15
Residual Chlorine
mg/1
0.01
0.05
0.04
0.96
1.00
0.88
1.07
0.99
1.05
                         TABLE 7

         Effect of Citrate on Copper Interference

Cl  Cone.        Copper Cone.        Cone,  of Citrate    Residual  Cl
 mg/1	mg/1	grams/liter	mg/1	

1  mg/1              0                     0                0.99
1  mg/1              5                     10               0.99
                            19

-------
Iron in a concentration of 5  mg/1  gave  low  recoveries of residual
chlorine.  See Table 8.

Iron (Fe) does not seem to form a  stable  complex sufficient to prevent
interference.  The citrate used with  the  iron and copper solution pro-
duced an incrustation on the  platinum electrodes and this was viewed
as undesirable.
NTA
The chlorine demand of NTA (nitrilo  triacetate  acid sodium salt) was
checked by the addition of 1  gram of NTA to  the titration vessel which
contained a chlorine residual  of 0.97 mg/1.   NTA exhibited a chlorine
demand of 0.06 mg/1.  Addition of 5  mg/1  copper and 25 mg/1 iron
yielded the same chlorine demand of  0.06 mg/1.

Sodium Qxalate

Sodium oxalate exhibited the  same type of behavior as NTA.  A 1 gram
sample added to a 1  mg/1 solution yielded a  chlorine demand of 0.06 mg/1.

Sodium Pyrophosphate

Sodium pyrophosphate was used by Mizuno (12)  to mask ferric iron in the
determination of ferrous iron.  Since a significant concentration of
iron is present in cooling water blowdown, sodium pyrophosphate should
be successful in eliminating  the effects of  iron on the  residual chlorine
determination.

A 1 gram sample of sodium pyrophosphate was  added to the titration vessel,
For a 1 mg/1 chlorine spike all of the chlorine was recovered.  Addition
of copper (5 mg/1) and iron (25 mg/1) gave an average chlorine residual
of 0.98 mg/1 or a 2 percent lower value.  This, however, is better than
any of the previous complexing agents.
                                  20

-------
The pyrophosphate chelating system gave good stable end points  with  no
drift.  Meter sensitivity was down slightly but tolerable due to  the
lack of drift.  To quantify the amount of pyrophosphate to effectively
complex interfering ions a 10 percent stock solution of Na. PJ3j  • lOHLO
was prepared as stated in the appendix; 1.4 ml  was  added to the titration
vessel.  This quantity was found to give stable end points for  copper
solutions up to 0.5 mg/1.  This quantity of pyrophosphate was sufficient
to complex 10 mg/1 of iron by allowing a stable end point.  A composite
solution of 0.5 mg/1 Cu (II) and 10 mg/1 Fe (III) with  1.4 ml of  pyro-
phosphate gave a slight drift in the end point. A  2 ml  portion of
pyrophosphate however, stabilized the end point. The pyrophosphate
system gave no positive chlorine residual analysis  in the presence of
copper.  The platinum electrodes remained shiny and clean.

EFFECT OF TURBIDITY

Varying amounts of turbidity were simulated by  Kaolin suspensions in
varying concentration.  Two mis of a 0.10 mg/ml chlorine stock  solution
was added to the titration vessel containing 200 ml of  the Kaolin sus-
pension.  The Kaolin suspension was prepared by adding  1  g of Kaolin
to one liter of water and performing appropriate dilutions of this
stock.  Two mi Hi liters of sodium pyrophosphate were added to the
titration vessel. See Table 9.

Good recovery of the residual was obtained in Kaolin concentrations  up
to 500 mg/1.  The effects of other solid and colloidal  suspensions were
examined.  Two samples tested were Corvallis clay from  the vicinity  of
the Oak Creek bank near 35th street in Corvallis and dried bottom
sediment from Cline's Pond.  Both sediments were sieved through 120
mesh screens to separate the fine particles for suspension.  Each sam-
ple exhibited an initial chlorine demand, i.e., total recovery  of the
chlorine spike was not achieved.  Additional spikes of  chlorine were
made on the system and good recovery was achieved.   See Table 10.
                                21

-------
                         TABLE  8
         Effect of Citrate on  Iron  Interference
          In Determination of  Chlorine  Residual
Cl Cone.
mg/1
1
1
1
1
Fe Cone.
mg/1
0
0
5
5
Cone, of Citrate
grams/liter
0
5
5
40
Residual Cl
mg/1
0.99
0.98
0.85
0.95
                         TABLE 9
        Recovery of Cl  in Presence of Kaolin
mg/1 Kaolin	JTU	Cl  Residual  (mg/1)
      0                                             0.96
     25                     (25)                    0.96
    100                     (86)                    0.96
    500                                             0.94
                            22

-------
WATER TREATMENT

A 100 mg/1  solution of a phosphonate-type water  treatment chemical was
prepared.   Solutions with no chlorine  added  gave no positive  indication
of a chlorine residual.   Solutions  of  the water  treatment chemical were
spiked with known amounts of chlorine.   Chlorine residuals were imme-
diately run.   The solutions  were  allowed to  stand with a chlorine resi-
dual run at varying intervals.  A data  summary appears in Table 11.

From the data it is evident  that  the water treatment chemical does ex-
hibit a chlorine demand.   Additional spikes  of chlorine to these sam-
ples allowed recovery of greater  than  90 percent.
                                 23

-------
                        TABLE 10

       Recovery of Cl  Spikes  on Cline's  Pond Sediment
                    And Corvallis  Clay

                              Recovery of Cl  Spikes
                         Spike #1       Spike. #2    Spike #3
Sample	Initial  mg/1	mg/1	mg/1

Cline's Pond Sediment      0.70          0.97        0.97

(100 mg/1)                 0.10*         0.47*       0.43*
Corvallis Clay             0.57          0.95        0.97

(500 mg/1)
Chlorine spikes = 0.98 mg/1
*Chlorine spikes = 0.49 mg/1
                        TABLE 11

     Recovery of Chlorine Additions in the Presence
   of a Uater Treatment Chemical  as a Function of Time
Contact time
0
5
69
(hours)



Initial Cl #1
Concentrations 1.01
Solution #1
(mg/1) Cl
0.66
0.59
0.38
#2 #3
2.52 5.04
Solution #2
(mg/1) Cl
2.18
0.89
0.00

Solution #3
(mg/1) Cl
4.09
3.33
1.46

                           24

-------
PRECISION

Twelve replicate runs using the procedure outlined  in  the appendix
using the Wallace & Tiernan Titrator,  were made  on  each  concentration
of residual chlorine.  An average value of the chlorine  residual  along
with the standard deviation and relative standard deviation  appear  in
Table 12.  A calibration curve was plotted for the  average values of
Table 12.  See Figure 2.  A least squares analysis  was run with  a cor-
relation coefficient of 0.9999.
                                 25

-------
                        TABLE  12
                 Precision  of  Titration
Cl spike
 mg/1
Average
 mg/1
Standard
deviation
  mg/1
  Relative
  Standard
deviation (%}
 0.053
 0.53
 1.06
 2.12
 0.047
 0.52
 1.02
 2.05
 0.0039
 0.009
 0.008
 0.0263
  8.3
  1.7
  0.7
  1.28
                            26

-------
                                              in
                                              CJ
                                              cvi
  10
"-^
CVI

LU
   II
  it
0<
o:
CDLJ
OO
  -J


  O
             O  CD  (0  *-
              *   »  	•   •
             cvi  —  —  —
                            CVJ  O  00  CD   Tf  CM
                             •   •   •   •    •   •


                       "N t-9QOO' OVd JO 'M
                                              8
                                              10
                                              N
                                                CP|
                      27

-------
RECOVERY RATE OF CHLORINE IN SYNTHETIC  AND  NATURAL BLOWDOWH

Synthetic blowdown water was prepared by slow  evaporation of natural
waters on a hot plate.   Solutions  were  allowed to evaporate to an
eighth of their original volume.   This  allows  an 8-cycle concentration
of cations and anions along with organic compounds.  Samples chosen for
synthetic blowdown were from the Buffalo River, Cuyahoga River and
Cline's Pond.  Chemical analysis of these sources is shown in Table 13.
All samples were spiked with a chlorine stock  solution  to a residual
chlorine level of 5 mg/1 and were  allowed to stand overnight to satisfy
any long-term chlorine  demand. A  residual  chlorine analysis was run on
each sample.  Approximately 0.5 mg/1 was added to another aliquot of
the natural samples and an additional  residual  chlorine analysis was
run.  The data are presented in Table  14.   The last two entries to
Table 14 are actual blowdown systems.

Cooling Tower I blowdown in the presence of pyrophosphate allowed a
stable end point to be  obtained with no positive indication of chlorine
when none was present.
                                   28

-------
                             TABLE  13



            Chemical  Analysis of Natural Water Samples



Characteristic           Cline's Pond      Cuyahoga  River
Buffalo River
QIC TOC (mg/1)
Ammonia N (rng/1)
Nitrite N (mg/1)
Nitrate N (mg/1)
Chromium (yg/1)
Copper (yg/i)
Iron (yg/1)
Lead (yg/1)
Manganese (yg/1)
Zinc (yg/1)

Recovery of
5 10
0.018
0.014
0.79
5 10
26 14
2100 900
20 25
33 210
16
TABLE 14
Chlorine Spikes on Slowdown
Total Chlorine
Residual Chlorine Residual
Source of Chlorine Addition After Spike
Sample mg/1 * mg/1 mg/1
Buffalo River 1.15
Cline's Pond 2.05
Cuyahoga River 1.12
Cooling Tower I 1 .60
Cooling Tower II 1.22
0.49 1.60
0.52 2.55
0.52 1.61
0.49 2.08
0.51 1.71




6
43
1300
31
58



Recovery
of
Addi ti on
mg/1
0.45
0.50
0.49
0.48
0.49
*After satisfaction of initial  demand and  overnight  standing.
                                29

-------
                     BIAMPEROMETRIC END POINT

An alternate amperometric titrator was  sought which  might  be  more  field
adaptable.  Marrow (13) described a dual,  polarizable  electrode  (D.P.E.)
method for the determination of the end point.   Many advantages  have
been cited for this method.   The first  is  simplicity of the electrode
configuration.  The electrode consists  of  two platinum sheets mounted
in plastic, see Figure 3.  This arrangement eliminates the need  for
maintenance of a reference electrode.   The power supply for a D.P.E.
method is much simpler in that an opposition voltage need  not be ap-
plied to the electrodes to get the end  point on  scale  as required  by
a single polarizable electrode.  For a  D.P.E., the end point  of  a
reversible electrochemical system titrated with  an irreversible  titrant
will cause the end point to occur close to zero  microamps  (14).

EQUIPMENT

A dual platinum electrode system similar to that, described by Marrow
(13) was constructed.  See Figure 3.

The dual polarizable electrode system was  designed in  the  shape  of a
probe for simplicity.  The stirring system used  was  a  La Pine "Porta
Stirrer."  This magnetic stirrer has a  rechargable battery pack  al-
lowing portable operation.

The schematic diagram of the control unit  appears in Figure 4.   The
value of the shunt resistor R is chosen such that the  10 yamp meter
scale now reads 100 yamps full-scale.

The shunt system is always in the circuit  unless the push  button switch
is physically depressed.  Removal of the shunt allows  full meter sen-
                                30

-------
    FIGURE 3
CHLORINE PROBE(DPE)
      Top View

      (0    0]

              Wire To Control Box


"

Platinum f
Foil IE
Y'

[LI

w
"V


r
j
Icn
• >





i.
N
=
S
/
t
-

\
3

Probe Made From
6" x 1" Lucite SI
Connection Wire
IP mm
|C_IIIIII
f ~~~?~~
1 10mm
       Bottom View
                31

-------
             FIGURE 4
SCHEMATIC DIAGRAM OF DPE CONTROL BOX


              1.35V Mercury (D Size)
                                 i
             1KI1 10Turn Pot.
             AAAAA/VW	
        lOjJdmp
Push Button Switched
For ]0=1 Shunt
                               R is Chosen For
                               a lOx Shunt
                  T  To Electrode
                 32

-------
sitivity which permits a more precise detection  of  the end  point.  Ex-
treme care must be exercised to insure that the  currents  are less than
10 yamps before the shunt is bypassed or meter damage will  occur.

Thus, knowing that the end point will .occur at zero current, rapid
addition of the titrant until the vicinity  of the end point is possible,
speeding up the titration procedure.   The stirring  mechanism for the
titrator causes less vigorous agitation than that of the  Wallace &
Tiernan titrator minimizing the flashing problem.   The stirrer can be
of any configuration and may be battery operated to be field adaptable
along with a battery power supply of  the D.P.E.

EFFECT OF COPPER AND IRON ON D.P.E. METHOD

The effect of copper and iron in concentration analogous  to those found
in cooling tower blowdown was studied.   The interference  studied was
the effect of varying the potential between the  dual indicator elec-
trodes and measuring the background currents.  Plots of the relationship
of background currents vs applied voltage appear in Figure  5 and Figure
6.  The iron concentration was 10 mg/1  and  copper concentration was 0.5
mg/1.  The pH was 4.

For the iron interference plot note that at applied voltages of less
than 200 mv the pyrophosphate system  caused a reduction in  the back-
ground current to the blank level.  At higher voltages the  pyrophos-
phate system greatly reduced the background currents.

For the copper plots the effect of copper upon the  background current
is nil up to about 300 mv.   Beyond this the pyrophosphate system re-
duces the background currents of copper although not as dramatic as
the iron curves.  Note that as long as  the  applied  voltage  is kept
less than 200 mv background contribution is nil.  This is an indica-
tion that these heavy metals are electrochemically  inactive at these
potentials.
                                33

-------
                FIGURE 5
    BACKGROUND CURRENT VS APPLIED VOLTAGE
         IRON INTERFERENCE STUDY
4.0
3.0
           200       400
             APPLIED EMFtm.v.)
600
800
                     34

-------
                    FIGURE  6
       BACKGROUND CURRENT VS APPLIED VOLTAGE
               COPPER INTERFERENCE
  2.0
o
o:
o
z
Lul
Cu
Od

O
   1.0
.Smg/ICu
  .5mg/l Cu-f
Pyrophosphate
                                                :er
                                              Blank
               200        400
                APPLIED E.M.F (m.v.)
600
800
                        35

-------
For the determination of the end point in  a  total  chlorine  residual
analysis an applied voltage of 150 mv is  recommended.

The presence of copper and iron did not exhibit the  drifting  problems
noted for the Wallace & Tiernan titrator.  The  meter readings  at  high
currents (7.5 yamps) showed some instability.   This  is  probably due  to
nonconstant stirring.  At the end point,  however,  the meter readings
are very stable.

The recovery rate for both forms of the amperometric end  point are the
same and the precision comparable for total  residual  chlorine.  The
biamperometric technique offers an advantage by allowing  the  end  point
to be reached much faster.
                                  36

-------
                            REFERENCES

1.   Standard Methods for the Examination  of Hater  and Haste Mater,
     13th ed., American Public Health Association,  New York, 1971
     p. 110-142.

2.   Annual Book of ASTM Standards,  Part 23.,  American Society for
     Testing and Materials,  Philadelphia,  PA,  1972, p. 222-227.

3.   Palin, A.  T.,  Chemistry and Control of Modern  Chlorination.
     LaMotte Chemical Products Company.  Chestertown, Maryland
     (1973).

4.   Ni col son, N. J., "An evaluation of the Methods for Determining
     Residual Chlorine in Hater." The Analyst, 90:187-198, April,
     1965.

5.   Hasselbarth, V., "Bestimmung von freiem und gebundenem warksamen
     Chlor."  Zeitschrift fur Analytische  Chemie. 234:22 1968.

6.   Guter, K.  J.,  Cooper, W. J., "Evaluation  of Existing  Field Test
     Kits for Determining Free Chlorine Residuals in Aqueous Solutions."
     Army Medical Environmental  Engineering Research Unit.  Publication
     Number AD752440.

7.   Marks, H.  C.,  "Residual  Chlorine Analysis in Water and Waste Water."
     In:  Hater and Water Pollution  Handbook.  Leonard L.  Ciacczo (ed.)
     Marcel Dekker, Inc., New York,  1972 (Vol. #3).

8.   Standard Methods for the Examination  of Wastewater, 13th Edition,
     American Public Health  Association, New York,  1971, Sec. 114A.
                                37

-------
9.   Ibid, 114B.

10.  Lishka, R.  J., and McFarren,  E.  F.  "Water Chlorine (Residual)",
     No. 2 Report Number 40.   Environmental Protection Agency,
     Cincinnati,  OH, 1971.

11.  Marks, H.  C. "Determination of  Residual Chlorine in Metal Finishing
     Wastes."  Analytical  Chemistry  24:1885-1887.  1952.

12.  Mizuno, T.,  Talanta. 19,  369-72  (1972).

13.  Marrow, James J.  "Residual-Chlorine Determination with Dual
     Polarizable Electrodes."   Jour.  ANNA 58:363-367.  1966.

14.  Stock, John T.  Amperometric  Titrations.  New York Interscience
     Publishers,  1965.   p.  45-89.
                                 38

-------
APPENDIX
   39

-------
                 AMPEROMETRIC  TITRATION  PROCEDURE

INSTRUMENTATION

The titrator employed in  this  report was  the Wallace & Tiernan Series
A-790 amperometric titrator.   The  buret  furnished with the unit was
substituted with a 5 ml autofilling microburet with calibration marks
down to .01 ml.  The flow of titrant was  transported to the titrator
through appropriate tygon tubing and tubing adaptors.

S tandard Meth o ds recommends that the volume of titrant used be less
than 2 ml.   For determining chlorine residuals in cooling tower blow-
down within 3 minutes, a  larger addition  must be tolerated, hence,
the use of a 5 ml  buret.   If chlorine  residuals of greater than 5
mg/1 are anticipated, a smaller sample size should be used, i.e., 100
ml instead of 200 ml.  If one  desires  ease of titration of extremely
low residuals, < .2 mg/1, dilution of  the titrant would allow greater
refinement of the equivalence  of titrant added to the system.  A
separate buret system would be needed  to not contaminate the stock
(PAO) solution buret (0.00564N).   Appropriate concentration factors
must be applied to the buret readings  to correlate the results to
mg/1 residual chlorine.

To facilitate rapid addition of these  reagents of the sample vessel,
syringes or autopipets are recommended.   For measuring out the 200 ml
samples a graduated cylinder sawed off at the 200 ml calibration mark
makes rapid sample measurement easy.   A  small lip should be fashioned
to eliminate possible spills when  transferring samples to the titration
vessels.  A 200 ml aliquot of  sample is  automatically measured by dip-
ping the cylinder into the water sample, thus filling the cylinder to
overflowing.
                                   40

-------
The contents of the cylinder are carefully poured  into  the  titration
vessel without spilling.   Between runs,  the titration vessel  should be
rinsed with chlorine-free water.  The sample vessel  need  not  be  dried
however, since the graduated cylinder contains  the active chlorine
residual.  All that is important is  that a consistent amount  of  sample
water is added to the vessel.   Dilution  by water adhering to  the walls
has no effect.

The titration vessel  supplied with the Wallace  & Tiernan  titrator should
be adequate—provided that a differentiation of free and  combined resi-
dual is not made on the same sample.   If the titrant volume is kept low,
sample overflow should not be a problem.

A suitable pH meter is required to check the buffering  capacity  of the
water sample and to check the proper  amount of  buffering  agent required
to bring the sample plus  reagents to  the appropriate pH.

REAGENTS

All reagents were prepared as  in Standard Methods  114.B,  13th (1)
edition.

The standard phenylarsine oxide was  purchased from Wallace  &  Tiernan.

Sodium Pyrophosphate

Dissolve 100 grams of sodium pyrophosphate Na,  PpO^lOhLO in  1 liter
of water heated to 50°C.   Place mixture  on a mechanical stirring device
to bring the salt into solution.  Allow  the solution to cool.  Add 85
percent phosphoric acid to the solution  until a pH of 6 is obtained.
For total residual chlorine analysis  a low pH of the pyrophosphate
chelate is desirable  as not to overwhelm the pH4 buffer system.
                                 41

-------
TITRATION

     1.    Preliminary Work.

The electrode surface should be sensitized  to  iodine in the following
procedure:

          a.    Fill  the titration  vessel with  distilled water.

          b.    Add 1  ml of 5% KI., 1  ml  pH4 Buffer.

          c.    Add sufficient hypochlorite  to  cause a yellow to
               yellow-brown coloring  of  the solution.  Place
               vessel on titrator  and allow to stir for 10
               minutes.

          d.    After the time limit direct  a stream of distilled
               water from a polyethylene wash  bottle around the
               electrode to remove as much  iodine  as possible.

          e.    Fill  titration vessel  with distilled water and
               place on titrator and  add with  stirring .5 ml of
               titrant.  This will remove any  traces of Ip present.

          f.    Remove titration vessel  and  rinse electrodes
               thoroughly by directing a stream of distilled
               water around the electrode.  The titration vessel
               containing the sample  can be used as a catch for
               the rinse water.

While the titrator is stirring in  the sensitizing  step a 200 ml sample
of water can be obtained with the  specially prepared graduated cylinder
as noted in the equipment section. The purpose of this sample is  to
                                   42

-------
determine the amount of buffering  needed  to  drop the pH to 4.  To the
water sample add 1  ml  KI, 2 ml  of  the  10% pyrophosphate solution.  Place
pH probe into sample and add increments of pH4 buffer until pH is in the
range of 4 to 4.2.   Add approximately  0.25 ml in excess to compensate
for variations in the buffering capacity  of  the natural water.  Two mis
of a 10 percent pyrophosphate solution should be sufficient for 10 mg/1
iron and .5 mg/1 copper.  If metal  ions are  significantly higher or
lower, adjustment of the amount of pyrophosphate reagent should be made.
A criterion for the proper amount  of pyrophosphate is the stability of
the end point (No drift).

If the meter deflections become erratic,  this is due to salt build-up
around the electrode sockets.  To  cure this, the electrode and the front
faceplate must be removed and the  salt scrapped away.  A stream of dis-
tilled water can be directed at the socket holes to remove the remaining
salt.  The entire socket area should then be wiped dry to prevent any
possible leakage currents to flow.

     2.   Titration Procedure.

During the actual titration run, all steps should be performed with
both speed and good laboratory  technique.  All reagent syringes should
be calibrated to deliver the proper amount of reagents.  Practice with
the sampling graduate should be done such  that sampling is rapid and
without spills.   Past experience with  the system in question would aid
in the prediction of the amount of titrant required to reach 90 percent
of the end point to eliminate loss  due to agitation.  Sunlight should
be kept to a minimum and the titration done  in subdued light.

Syringes, calibrated for proper delivery  of  appropriate reagents, should
be placed into the opened bottles  to act  as  lids.  The potassium iodide
solution should be stored in a  brown bottle.  Fresh KI should always be
                                43

-------
used.  If any trace of coloration  is  present, the solution should be
discarded.  (If there is  any  question about  the KI, a titration should
be performed on distilled water  to which  the proper amount of KI was
added.  If a positive indication is noted, fresh KI should be substituted.)

The actual titration procedure is  similar to that in Standard Methods.
(9, 10)  Modification to  the  procedure involves the preparation and
use of an additional reagent  and addition of .1 large increment of
titrant before the agitation  starts.   The procedure is as follows:

          a.   Obtain a 200 ml sample in  the precalibrated graduated
               cylinder and pour without  spilling into the titration
               vessel rinsed  with  distilled  water.

          b.   Using the  precalibrated syringes add proper amounts of
               KI, pH4 buffer and pyrophosphate solution.  One ml of
               KI solution is sufficient.

          c.   Place titration vessel on  the amperometric titrator and,
               if feasible, add  approximately 90 percent of the final
               titrant volume.   Start stirring action.

          d.   Continue the addition  of titrant until the needle no
               longer deflects left.   After  the first titration the
               scale sensitivity control  should be adjusted until
               the meter  reads approximately 20.  For subsequent
               titrations the end point should appear approximately
               at the same reading.   Early  in the titration the
               meter will be  off scale.  Rapid additions of the
               titrant can be made until  the meter reads on scale.

          e.   Read the buret for the concentration of the chlorine
               residual.
                                  44

-------
BIAMPEROMETRIC TITRATION

The equipment required is  a biamperometric  probe, control box and mag-
netic stirring device.  These pieces  of  equipment are described in the
main body of this report.

Experimental  Procedure Biamperometric End Point

     a.   The sample is obtained in the  same manner as in the
          previous  procedure.   The sample vessel, in this case
          is  a 400  ml  beaker.   (Electrodes  need not be sensitized
          to iodine.)

     b.   The necessary reagents are  added  as described earlier.

     c.   A magnetic stirring bar is  added  to the beaker and the
          beaker placed on the magnetic  stirring device.

     d.   The power is turned on the  control box.  This can remain
          on  until  the end of the sampling  period.

     e.   The electrode probe is lowered into the solution and the
          stirring  action  started.

     f.   Addition  of titrant is done rapidly until the scale reads
          less than 10 yamps.   At this point the X10 button is
          depressed and titrant added rapidly until the meter reads
          close to  0 microamps.   Titrant is added in small increments
          until the meter  no longer deflects left.  The last increment
          is  then subtracted.
                                45

-------