WATER POLLUTION CONTROL RESEARCH SERIES
17O1OEEDO7/7O
                 THE ELECTRO-OXIDATION
                           OF
            AMMONIA IN SEWAGE TO NITROGEN
ENVIRONMENTAL PROTECTION AGENCY  WATER QUALITY OFFICE

-------
             WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe the results
and progress in the control and abatement of pollution in our
Nation's waters.  They provide a central source of information on
the research, development, and demonstration activities in the
Water Quality Office, Environmental Protection Agency, through
inhouse research and grants and contracts with Federal, State,
and local agencies, research institutions, and industrial
organizations.

Inquiries pertaining to Water Pollution Control Research Reports
should be directed to the Head, Project Reports System, Office of
Research and Development, Water Quality Office, Environmental
Protection Agency, Room 1108, Washington, B.C.  20242.

-------
     THE ELECTRO-OXIDATION OF AMMONIA IN  SEWAGE TO  NITROGEN
                                by
                       Ionics,  Incorporated
                 Watertown, Massachusetts  02172
                             for the

                       WATER QUALITY OFFICE

                 ENVIRONMENTAL PROTECTION AGENCY
                        Project  #17010 BED
                       Contract  #14-12-800
                            July 1970
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price 55 cents

-------
                EPA Review Notice
This report has been reviewed by the Water
Quality Office, EPA, and approved for publication.
Approval does not signify that the contents
necessarily reflect the views and policies of
the Environmental Protection Agency, nor does
mention of trade names or commercial products
constitute endorsement or recommendation for
use.

-------
                           ABSTRACT
A study was undertaken to determine the feasibility of electro-
oxidation of ammonia as^a possible new method for ammonia removal
during sewage treatment.  Initial tests were conducted with plati-
nized platinum and showed that ammonia decomposition occurred in a
narrow potential range with high coulombic efficiency, assuming
a three-electron exchange per ammonia molecule.  Conversion of
nitrogen to the elemental form was shown by analysis of the gas
from the anode.  Analysis of the electrolyzed water for nitrite,
nitrate, hydrazine, and hydroxylamine indicated these materials were
not formed in measurable quantities.  Their absence suggests a high
degree of conversion of ammonia nitrogen to the elemental form.

Following the initial determination of technical feasibility, a
survey was made of electrodes having lower cost than platinized
platinum.  Only electrodes containing a significant amount of
platinum were found capable of converting ammonia nitrogen to
elemental nitrogen.  The cost of these electrodes is too high to
make this ammonia removal method competitive with other existing
methods.  A cheaper electrode that can operate at significantly
higher current densities is needed to bring the cost in line with
the cost of these other methods.

This report was submitted in fulfillment of Project Number 17010 EED,
Contract 14-12-800, under the sponsorship of the Water Quality
Office, Environmental Protection Agency.
                              111

-------
                             CONTENTS
                                                                Page
ABSTRACT                                                        iii
INTRODUCTION                                                      1
THEORY                                                            2
EXPERIMENTAL                                                      4
     Procedures                                                   4
     Electrochemical Cells                                        4
     Electrode Preparation                                       10
     Analytical Techniques                                       12
RESULTS AND DISCUSSION                                           13
     Platinized Platinum Micro-electrode                         13
     Platinized Platinum Macro-electrode                         27
     Evaluation of Other Electrocatalysts                        29
           Titanium, Tantalum, Zirconium, Niobium                29
           Proprietary Catalysts                                 29
           Rare Earth Oxides                                     30
           Non-Precious Electro-deposited Metals and Oxides      31
           Noble Metals                                          31
                 Bright Noble Metals                             31
                 Electro-deposited Noble Metals                  31
                 Commercial Electrodes                           32
                 Noble Metal Blacks                              33
                 Platinum Black Catalyst                         34
     Continuous Flow Apparatus                                   41
           Cell Design                                           41
           Cell Operation                                        41
     Estimated Unit Cost                                         42
CONCLUSIONS                                                      46
 REFERENCES                                                       47
                            IV

-------
                         INTRODUCTION
Ammonia in wastewater has two detrimental effects upon receiving
waters.  Because it is a plant nutrient, it encourages excessive
algal growth.  Because of its conversion to nitrate in the
receiving waters, it  exerts a very significant oxygen demand.
As a result of these undesirable characteristics there is an
increasing need for ammonia removal from wastewaters.  Air
stripping at high pH, bio-oxidation to nitrate, selective ion
exchange, and breakpoint chlorination are existing methods for
accomplishing this removal, but none of these  methods is
entirely satisfactory.  There is need for an effective and
inexpensive ammonia removal technique.

A very desirable property of an ammonia removal method is the
ability to convert the nitrogen in the ammonia to elemental
nitrogen which causes no further pollution of the environment.
Studies concerned with ammonia in fuel cells indicated that
electrooxidation of the ammonia had a good potential for
producing elemental nitrogen.  This fact, and the knowledge
that power requirements for applying this treatment method to
domestic wastewater should be very low, led to the present
feasibility study.

-------
                              THEORY
Recent years have seen the development of substantial interest in
ammonia as a possible fuel-cell reactant because of its high content
of hydrogen.  That interest has raised the question of the electro-
chemical behavior of ammonia, i.e., of the mechanism and kinetics
of its reactions at anode surfaces.  Several papers dealing with
this problem have appeared (1-4) with contradictory results in
regard to kinetics (2,4) but with general agreement that the overall
electrochemical reaction at the spontaneously established negative
potential is a three-electron exchange reaction yielding nitrogen
and water as products with almost 100% faradaic efficiency:

           NH3 + 30H~ - 1/2 N2 + 3 H^ + 3e~               [l]


A platinized platinum anode, a highly active electrocatalyst, was
used in experimental work by all of these authors.  T. Katan et _al  (1)
found that current efficiency of reaction [l] using platinum black
as the anode was 99.7% at current densities  (c.d.)^ 100 mA/cm2 and
anode potentials of -0.31 to +0.15V.

All of the experiments were conducted in strongly alkaline solutions
(in the range of 0.1 - 6.9N KOH) and ammonia concentrations ranging
from 0.03M to 6M.  H.G.Oswin e_t a_l  (2) suggested that oxides of
nitrogen would be expected only at potentials positive to a standard
hydrogen electrode.  All of his experiments were conducted in the
potential region negative to SHE where visible gas evolution was
observed.  The gas was identified as pure nitrogen.  He too stated
that overall reaction [l] proceeds with 100% coulombic efficiency.

Anodes of smooth platinum and other metals were tried  (3,5,6) but
nitrogen was never produced in quantities which corresponded to
complete stoichiometric reaction of ammonia as indicated by Eq.[l]  .
Instead substantial quantities of nitrate and lesser quantities of
nitrite ions were formed:


           NH  + 7OH~ - NO ~ + 5H O + 6e~                  [2]
           NH  + 9OH~ -> NO ~ + 6H O + 8e~                  [3]


Various workers (2,3,4), studying the electrochemical kinetics of
the ammonia oxidation reaction in alkaline solutions on platinum
black surfaces, reported the proposed kinetic scheme of consecutive

-------
reactions, theoretical current-voltage relationships for various
rate -determining mechanisms, and experimental cur rent- voltage re-
lationships .  There are four sequential steps in the proposed
anodic oxidation of ammonia,  in these equations the electrode
surface is represented by M.

           NH  + M 4- OH~ -* M-NH  + HO + e~                [4]
             O                     <,


           M-NH  + OH~ - M=NH + HO + e~                   [5]


           MNH + OH~ - MSN + HO + e~                     [6}
                       2M + N
                             b
The electrode surface would be then covered with -NH2/ -NH,' and -N
species.  Most of the authors agreed that step [5] is the rate-
determining step in the process of the anodic oxidation of  ammonia.

-------
                           EXPERIMENTAL
Procedures

Potentiodynamic procedures have been employed in most of this work
using a Heath Polarography System Model EUW-401.*  This system
consists of Polarography Module EUA-19-2, operational Amplifier
System Model EUW-19A, operational Amplifier Chopper Stabilizer Model
EUA-19-4 and Servo Recorder Model EU-20B.  The polarography module
includes the potentiostat, initial potential source, linear potential
sweep generator and current measuring amplifier.  In some experi-
ments, the steady-state potentiostatic procedure was also used,
current-potential relationships being determined point by point
 (20 mV steps) at controlled intervals of time (1 minute).  All
potentials are referred to the standard hydrogen electrode  (SHE)
and are designated E  .
.Electrochemical Cells

All-glass electrochemical cells of conventional design  (Figure 1)
were used for most of this work.  The central compartment contained
the working electrode and had a sufficient volume to minimize con-
centration changes during the experiments.  The platinized platinum
gauze counter electrode was in a compartment separated from the main
compartment by a fritted glass.  The reference electrode was mercury/
mercuric oxide in 1M KOH.  In some experiments a saturated calomel
electrode was used.

In the first part of the experimental work, the working electrode was a
platinized platinum electrode  (bulb-type rotating electrode after
Laitinen-Kolthoff) rotated by a Sargent Constant Speed Stirring
Motor  (600 rpm).  A standard electrode cleaning treatment (7) was
used prior to platinization.  The platinizing procedure from the
same source  (7) yielded black deposits.  The same procedure was used
before each set of measurements.  Geometric area of the electrode
was 0.13 cm^ .

Experiments were conducted at room temperature in various supporting
electrolytes (Nad, K2SO4, KOH, phosphate buffer and K2HPO4) with pH
values varying from 5.55 to 12.8 and ammonia concentrations from
0.00012M to 0.01M  (~ 2 to 170 mg/1).
 *Mention of proprietary equipment or products is for information
 purposes only and does not constitute endorsement by  the Water
 Quality Office,  Environmental Protection Agency.

-------
                 I
    A - gas inlet
    B - ref. electrode
    C - counter electrode
    D- working electrode
Figure 1. Experimental Cell for Investigating Anodic Oxidation of Ammonia

-------
To determine the course of the reaction, experiments were continued
with a larger platinized platinum electrode and reaction products
accumulated for analysis.  Constant current was applied for a com-
paratively long period of time through the cell shown schematically
in Figure 2 .  The apparatus consisted of an air-tight Pyrex glass
cell with rotating platinized platinum gauze anode C (geometrical
surface area 20 cm^), mercury (triply distilled) cathode F, Hg/HgO
reference electrode A, sweep-gas inlet B, solution overflow outlet E,
and sampling septum for both gas and solution D.  Solution content
was 1.2 liters.  The supporting electrolyte was a phosphate buffer
(pH 8.2) and initial ammonia concentration varied from 40 to 55 mg/1.
The initial experimental step was pre-electrolysis of the supporting
electrolyte at 1 to 2 mA for 2 hours.  Helium gas was bubbled through
the cell for at least two hours before the experiment was started to
exclude the air from the solution and to replace the air at the top
of the cell.  Flushing was continued until very little air was de-
tected in the cell by chromatographic analysis.  After flushing, the
ammonia was added and the experiment started by applying a current
of 10 to 15 mA.  The anodic potential was kept at all times below the
critical point where anode passivation occurs.  Samples of solution
and gas were taken for analysis every 1 to 2 hours.  Tests for the
most probable reaction products  (nitrates, nitrites, hydrazine and
hydroxylamine in solution, and nitrogen in gas phase) were performed
and determination of ammonia concentration changes during the oxidation
reaction was made.

In a series of experiments where the search for an electrocatalyst
other than platinum alone was performed, the platinized platinum
micro-electrode was replaced with other metals including bright or
black noble metals and with powder type catalysts of different com-
position layered over an inert base metal.  These experiments were
conducted at room temperature in phosphate buffer solution of
pH = 8.00 0.15  or pH = 6.60 0.05   (for metal oxides unstable in
basic solutions) with helium gas bubbling through the cell.  The
ammonia concentration was 50 to 100 mg/1.

A continuous flow electrolytic cell was used to simulate one of many
parallel channels in a full size prototype.  The cell, shown .schema-
tically in Figure 3, contained a platinized titanium anode and a
stainless steel cathode with the cell volume and active electrode
surface defined by a polyethylene spacer between the electrodes.  A
top view and a cross-section of the spacer is shown in Figure 4. The
active electrode area in the cell was 60 cm2 with a flow path length
of 90 cm and a path width of 0.7 cm.  No membrane or diaphragm was
used to separate the electrodes.  A test solution was prepared from
ammonia- and a pH buffer at values typical of wastewater effluents
(about 30 mg/1 ammonia, pH 8).

The experimental set up included a feed solution reservoir containing
13 liters of test solution.  The solution was fed to the cell by a

-------
                         rf
                J

u
        ^
L
   A- ref. electrode
  8-gas inlet
  C-anode
  Desampling outlet
  E-overflow
  F-cathode
Figure 2. Experimental Cell for Determination of Ammonia Oxidation Reaction Products

-------
Outlet
Lucite block
Rubber gasket
Cathode
Spacer
Anode
     Figure 3. Continuous Flow Bench Scale Electrolytic Cell

-------
             o
             o
             o
                                                o
o
o
Figure 4. Top View and Cross-section of the Tortuous Path Spacer

-------
centrifugal pump.  A rotameter, calibrated by effluent collection,
measured the flow rate.  Current was provided by a 25 amp DC power
supply.

The experimental parameters were selected on the basis of the de-
sign equation discussed later.  With assumption of 50% average
current efficiency, 90% ammonia concentration reduction and a tortuous
path spacer of 90 cm length and 0.1 cm thickness, an input velocity
(Vo) of 0.274 cm/sec was required with a corresponding retention
time (t ) of 327 sec.
       R

The experimental data were recorded with various solution feed rates,
spacer thicknesses and current densities.
Electrode Preparation

The small test electrodes were prepared from high purity 0.010 or
0.020 inch diameter wires which were sealed into glass tubing.
Electrical contact was established through a column of mercury in-
side the glass tubing and all prepared electrodes were tested at a
rotating velocity of 600 rpm.

Smooth metals tested anodically were platinum, rhodium, iridium,
palladium, nickel, titanium, tantalum, zirconium, and niobium.

Metals tested in electrodeposited form were palladium black (on  
palladium), platinum black  (on graphite), rhodium bright (on rhodium),
iridium black (on iridium), and nickel black  (on nickel).

Palladium catalyst was prepared by electroplating palladium from a
solution containing 1% palladous chloride in IN hydrochloric acid
at a current density of 30 mA/cm  for 30 minutes.  The procedure
used for preparation of platinized graphite was the same one as for
platinum.

Rhodium was electrodeposited from a rhodium sulfate bath containing
20 ml/1 concentrated sulfuric acid and 2 g/1 rhodium.  Plating was
carried out at a temperature of 42 C and at a current density of
100 mA/cm  for 30 minutes using a platinum anode.  A bright rhodium
deposit was obtained.

Iridium was"electrodeposited from an iridium plating solution using
the technique of Ovenden  (8).  A dark grey, not very coherent,
deposit of iridium was obtained by this procedure.

A nickel catalyst was prepared in the form of a very finely divided
black deposit by electrodepositing nickel from a bath containing
33 g/1 of nickel ammonium sulfate and 14 g/1 of Rochelle salt at a
current density of 100 mA/cm2, at 20C, and at a pH of 6.6 .


                                 10

-------
Manganese dioxide was deposited by anodic electrodeposition at 80C
and at a current density of 2 mA/cm  from a bath containing 50 g/1
anhydrous MnSO^ and 65 g/1 H2SC>4.  Deposition time was 25 minutes.
                                                            ;
The electrodes for testing electrocatalysts in a powdered form were
prepared by .coating 0.02 inch diameter titanium wire, titanium
coupons, or expanded titanium (9 mil overall thickness with mesh
dimensions of 0.055 in.  x 0.0285 in. and approximately 700 openings
per square inch) with a mixture of powdered catalyst and Teflon T-30
dispersion as binder.

Teflon T-30 is an aqueous dispersion of Tetrafluoroethylene (TFE)
resin particles containing about 60% solids, the rest being water and
"Triton X-100" stabilizer (a non-ionic wetting agent).  Special heat
treatment is required in order to remove water and stabilizer, which
decomposes cleanly at temperatures between 250 and 300C, leaving
a contaminant-free polymer.
Two coating procedures were tested.
follows:
Coating procedure #1 was as
           9   Mix conducting material throughout T-30 dispersion
               by very slow agitation  (rapid agitation causes
               irreversible coagulation).

         .     Dip titanium support into dispersion for even
               layering.  If too much pick-up occurs per dip,
               dilute mixture with water.

           e   Place coated electrodes in furnace, which is
               gradually heated from room temperature to 250C.
               Water is removed by this procedure from the
               coated catalyst mixture.

           e   Heat for 3 hours at temperatures between 250 and
               300C to remove stabilizer.

              Sinter for 4 to 5 minutes at 360C (melting point
               of TFE resin).

              For further applications of the catalytic mixture,
               cool electrodes rapidly, apply next coating and
               repeat heating procedure.  After last application
               cool the electrodes slowly to provide even annealing
               which is needed for good adhesion.

Technique #2 includes two steps:

           0   Dry the applied mixture for 5 to 10 minutes
               under an infrared lamp.

              Bake and sinter at 320C and 1500 to 2000 psi.
                                  11

-------
The desired pressure and temperature were provided by a hydraulic
press with heated platens.  Pretreatment of the titanium support
consisted of cleaning in hot (70C to boiling) hydrochloric acid
(8 to 12N) for 5 to 15 minutes.  Graphite and tungsten bronzes were
often used in mixture with the catalyst to improve the electrical
conductivity and to dilute the expensive catalyst.  It has been shown
that, by incorporation of tungsten bronzes into the platinum black
anodes, the electro-oxidation of carbon monoxide and reformer gas
can be appreciably enhanced(9).  The attempt was made to use the
platinum catalyst more effectively in the ammonia oxidation reaction
by addition of tungsten bronzes to the catalyst in different ratios.

Sodium tungsten bronzes with the general formula NaxW03 and high
sodium content were prepared according to a modified Straumanis (10)
procedure.

Platinum electrodeposition on titanium and tantalum was carried out
using the same procedure as for platinizing platinum.  Tantalum
electrodes were treated prior to platinization cathodically apply-
ing 100 mA/cm  for 15 minutes in 2% sulfuric acid solution. Titanium
was pretreated in hot hydrochloric acid the same way as for powder-
catalyst electrode preparation.
Analytical Techniques

A Hach DC-DR photoelectric colorimeter was used for liquid phase
analysis.  The ammonia nitrogen was determined by Nessler's method
with a minimum detectable nitrogen concentration of 0.05 mg/1.

The cadmium reduction method which is a modified diazotization  (1-
naphthylamine-sulfanilic acid) method was employed for the nitrate
nitrogen determination.  Minimum detectable nitrate nitrogen con-
centration was 0.1 mg/1 .

For determination of nitrite nitrogen a diazotization method  (1-
naphthylamine-sulfanilic acid) was used with minimum detectable
nitrite nitrogen concentrations of 0.01 mg/1.  Hydrazine tests were
performed using the 4-dimethylaminobenzaldehyde method.  The method
is very sensitive and is used for determination of small amounts of
hydrazine.  Minimum detectable concentration was 0.01 mg/1.

The hydroxylamine test was based on the reaction with benzoyl chloride
and ferric chloride and the method was modified for colorimetric
analysis in low concentrations.

Chromatographic analysis of the gaseous phase was conducted using an
Aerograph Gas Chromatograph Model A-90-P3.
                                 12

-------
                      RESULTS AND DISCUSSION
Platinized Platinum Micro-Electrode

The first set of experiments was conducted in alkaline solutions
(pH 12.8 to 13.0) with KOH as the supporting electrolyte and.
ammonia concentrations varying from approximately 0.001M to 0.003M
(17-51 mg/1).  Potentiodynamic and potentiostatic techniques were
employed.  Single potential sweep current-potential curves recorded
with stationary and rotating platinized platinum micro-electrodes
in the same solution are shown in Figure 5.  An anodic current peak
which appears at -0.1 V with the stationary electrode was slightly
shifted towards more positive potentials  (0.05 to 0.25 V) and in-
creased in height about five times when the rotating electrode was
used.  Current maxima for stationary and rotating platinized
electrodes were measured at various sweep rates and" ammonia con-
centrations with the experimental results given in Figure 6 and
Table 1.  The lowest of the set of curves in each graph of Figure 6
was recorded in the supporting electrolyte (0.1M KOH) without
addition of ammonia.  The increase of the current peak value was
observed with increasing ammonia concentration and potential sweep
rate.  After the importance of increased mass transport to the
electrode was established, all the experiments were conducted with
a rotating platinized platinum electrode.

Current-potential curves obtained potentiostatically at steady state
conditions are shown in Figure 7.  Current maxima are observed at
potentials of about 0.025 V versus SHE .    The maximum current in-
creases with concentration in the same manner as in the potentio-
dynamic experiments.

Maximum current density as a function of ammonia concentration is
shown in Figure 8.  Maximum current density was found to increase
linearly with the square root of the ammonia concentration.  This
relationship was established on the basis of purely empirical
determinations.  Experimental data were fitted to several types of
relation including linear relation but square root relation gave
the highest degree of correlation.  The experimental points taken
at higher potential sweep rates were more consistent than those
taken at lower rates or at steady state conditions.

Maximum current density is plotted versus the square root of the
potential sweep rate  (V) in Figure 9.  A linear relationship is
observed at four different ammonia concentrations.  At the condition
where V2 ~> 0, the ordinate of the figure gives the value of the
steady state current for given concentration.
                                13

-------
                               36 mg/l NH3
                               0.2 V/min
                               pH 12.8
Figure 5. Current Maxima for Rotating and Stationary Electrodes

-------
                        TABLE 1
'Current Maxima for Stationary and Rotating Platinized Platinum
  Electrode at Various Sweep Rates and Ammonia Concentrations
Sweep Rate
[v/min]

0.50

0.2

0.1

0.05

steady-state

Ammonia Cone.
[mole/lxlO3!
3.0
1.5
0.75
3.0
1.5
0.75
3.0
1.5
0.75
3.0 I
1.5
0.75
3.0
1.5
0.75
2
Maximum Current Density[mA/cm ]
Rotating - i_ Stationary - i^
R S
10.8
6.2
3.5
8.1
4.4
2.3
6.9
3.9
1.9
5.4
2.7
1.6
3.8
1.7
0.96
2.08
1.23
'0.85
1.15
0.72
0.5
0.96
0.58
0.38
0.88
0.46
0.29
0.67
0.40
0.19
V^S
5.19
5.04
4.12
7.04
6.11
4.60
7.19
6.81
5.43
6.14
5.87
5.52
5.75
4.25
5.05
Average  iR/ig
5.61
                           15

-------
rOB
                47mg/l NH9
                               SOirWmto
             0.2     0.4     0.6     0.8
                      i	i	I
tOOirM/min
                                                                0.2     0.4     O.6     0.8
                                                                        EH.V
           Figure 6. Current Potential Traces for Various Sweep Rates and Ammonia Concentrations

-------
-1
   -0.2
0.6
0.8
                                                      EH. V
          Figure 7. Current Potential Traces for Various Ammonia Concentrations from Steady-state Measurements.

-------
oo
        12
         10
S  6
0)
Q
4-1
c
CD
      o
      E 4
      I
      'x
                                                                                                           0.50 V/min
                                                                                                              0.20
                                                                                                        D ^-0.10
                                                                                                               0.05
                                                                                           A  Steady state
                                                NH3 cone., (mole/l)1/2  x 102
                Figure 8. Maximum Current Densities as Functions of the Ammonia Concentration for Various Sweep Rates

-------
      12
u
C
c

-------
All the preceding experiments were conducted in the supporting
electrolyte with pH = 12.8 .  The influence of pH on the maximum
current density was determined by conducting the experiments under
otherwise similar  conditions but with supporting electrolyte of
different pH-value.  The pH range between 5.5 and 12.8 was in-
vestigated with corresponding maximum current densities presented
in Figure 10.  At very low ammonia concentrations maximum current
density increases slightly and almost linearly with pH while at
higher concentrations a minimum is observed at pH about 7.5 .
Increasing the pH from 9 to 12, increases current density 2.5 to
3 times for 0.002M and 0.003M ammonia solutions.

The supporting electrolyte had no influence on the rate of oxidation.
Sodium chloride was of special interest as a supporting electrolyte
because of its presence in substantial amounts in wastewater.  There
is a possibility of interaction of the hypochlorite formed during
the electrolysis of a chloride-containing solution with the ammonia
yielding chloramines.  However at the low anodic potentials applied
in the electro-oxidation of ammonia  (0.2 to 0.5 volt vs. SHE)  no
chlorine evolution would be expected and consequently the subsequent
reactions would not occur.  With sodium chloride concentrations as
high as 500 mg/1, chlorine evolution would not be expected on platin-
ized platinum anode below 1.48 volts vs. SHE.

One series of experiments with the platinized platinum rotating
microelectrodes consisted of recording the current-time relationship
at chosen constant potentials  (100 mVf 200 mV, 300 mVf 400 mV and
500 mV).  To ensure that the surface of the electrode would be in
the same state for each test the electrode was subjected to a fairly
lengthy pre-electrolysis sweep.  The operating scheme is shown in
Figure 11.  This scheme consisted of a pre-test sweep from -0.2 volt
to +1.5 volts and return and a post-test sweep to 1.5 volts and
return to -0.2 volt.  The total duration of the sweep is not sufficient
to make any significant change in the test solution.

The first set of measurements was recorded with 0.05M K HPO   as
supporting electrolyte  (pH=9.05) containing 7mgp/lof ammonia  (Figure
12).  It can be seen that current is fairly steady during the in-
vestigated period of time at potentials up to 400 rnV, which is very
close to the potential of maximum current in the single potential
sweeps.  At greater potentials a large drop in current density  (i.e.
decrease in ammonia oxidation rate) with time was observed.

The same behavior was observed in experiments with 17, 34 and 51 mg/1
of ammonia as shown in Figures 13 to 15, the only difference being
that the potential of maximum current is shifted to lower values with
increasing, ammonia concentration.

Generally it can be concluded that the optimal potential for con-
tinuous electrolysis is very near the peak current potential de-
                                 20

-------
  10.0
   9.0
   8.0
 - 7.0
I  6.0



E


M
'(/)
C


  5.0

c



o

E
D

.  4.0
X
ro
   3.0
   2.0
   1.0
       0.003 M NH3.



     0.002 M NH3



  0.001 M NH3 ,



0.0005 M NH3 ,
                      J	1
                               8       9

                                  PH
      10       11       12
13
 Figure 10. Effect of pH on the Maximum Current Density at Sweep Rate of 0.2 V/min.
                                  21

-------
ro
N5
         I
         HI
             2.5
             2.0
             1,5
             1.0
             0.5
             0.0
            -0.5
            -1.0
                 -   Linear
                     potential
                     sweep
                     0.2V/min
                                                      15
30
                                                                   Time, min
                                    Figure 11. Potential Sweep Sequence for Current-Time Investigation

-------
0.1  _
                                         c = 7 mg/l NH3
                                       pH = 9.05
                                2             3
                                     Time, min
                Figure 12.  Current-Time Relationship at Constant Potential

                                 23

-------
1.0
0.9  -
0.1  _
                                    c= 17 mg/l NH3
                                  pH = 9.05
                                 2              3

                                    Time, min
               Figure 13. Current Time Relationship at Constant Potential

                                  24

-------
   1.0
            500 mV
   0.9
      I  -
   0.8
   0.7
    0.6, r
    0.5
 c
 0)
 Q

 4-
 C


'
                                                                  c = 34mg/l NH3

                                                                 pH = 9.05
    0.4
    0.3
    0.2
             100mV
    0.1
                                                                 -a-
                                      Time, min
                   Figure 14. Current Time Relationship at Constant Potential



                                       25

-------
   1.3
   1.2
   1.1
   1.0   -
   0.9   .
 . 0.8
>
c
01
Q

c

-------
termined from single potential sweeps at the given pH and ammonia
concentration.


Platinized Platinum Macro-electrode

The degree of .completion of the oxidation reaction, current efficiency
and reaction products were determined using the larger electrolytic
cell  (Figure 2) with rotating platinized platinum anode of 20-cm^
geometric surface area.  Test of nitrites, nitrates, hydrazine and
hydroxylamine using colorimetric methods were negative. Quantitative
analysis of evolved nitrogen as the only gaseous product of the
electro-oxidation of ammonia was successfully performed using gas
chromatography.  Known amounts of the head gas from a sealed cell
were used to obtain peaks on the chromatograms (Figure 16bf c, d)
which were compared with reference peaks obtained from dilutions of
air in helium (Figure 16a) .  Even with the greatest precautions,
traces of air could be detected on the chromatograms, probably due
to leakage into the gas-tight syringe.  To avoid any mistake about
the origin of the detected nitrogen, chromatographic peaks from the
test gas were always compared with reference peaks obtained with the
air-helium system.  Results of the analysis from one experiment are
shown in Table 2, and Figure 16.  The "observed" value of nitrogen
evolved refers to the value obtained by gas chromatographic analysis
and the "calculated" value is obtained from the ammonia loss from
the solution during the reaction assuming stoichiometric conversion
of ammonia to nitrogen.


                              TABLE 2
Charge
[Coulombs ]
123
204
2603
2811
Volume of N
Observed
3.5
11.1
30.2
33.7
evolved [ml]
Calculated
4.9
12.0
24.0
28.5
Deviation
from mean
16 . 7%
3.9%
11.4%
 8.3%
Good agreement between the observed and calculated values was noted
except at the very beginning of the reaction where amounts of the
evolved nitrogen are very small and chromatographic analysis could
fail to give accurate results.  This agreement, coupled with the
absence of other reaction products, indicates a high degree of con-
version of ammonia nitrogen to the elemental form.  Current efficiency
for the oxidation reaction  (with assumption of three electron ex-
change) varied between 50 and 80% depending upon the experimental
conditions.

                                 27

-------
00
        80
        70
        60
        50
        40
        30
        20
        10
                    (a)
Reference
(air sample)
N2/O2 = 3.85
80
                 70
                 60
                 50
                 40
                 30
                 20
                                  01
                             (b)
t  =  2 hr
N2/O2 = 4.7
                                                          90  -
80 t
                         70  L
                        60  L
                        50  \-
                        40  h
                        30  U
                        20
                                                          10
IUU
(c) 90
1

-
-
-
-
-
-
" , J
t = 20 hr 80
N2/02 = 13.3
70
60
50
40
N2 30
20
10
UT
M V- -. . - 0
-
t = 25 hr
N2/02 = 17.9
-
-
-
-
-
-
-J
u,
1




N2

02
1C
                           Figure 16. Chromatographic Analysis of Gas Samples Taken at Different Reaction Times

-------
Evaluation of Other Electrocatalysts

Part of the work on electro-oxidation of ammonia consisted of an
investigation of a number of different materials as possible electro-
catalysts for the ammonia oxidation reaction.  The main emphasis was
on investigation of noble metals but many non-precious metals, metal
oxides, rare earth oxides and commercial proprietary catalysts were
evaluated experimentally.  Numerous metals showed very high ionic
currents at low anodic potentials in phosphate buffer solution and
at pH of 8  (e.g. vanadium, rhenium, molybdenum, tungsten, lead, tin).
Dissolution of the metallic electrode occurred, followed in a few
cases by passivation due to oxide layer formation at high anodic
potentials.  In some cases a second dissolution region appeared
(corresponding to a higher oxidation state of the metal) until finally
the oxygen evolution region was achieved.  Consequently, the behavior
of the electrode surface in the supporting electrolyte and in the
anodic potential region which is of interest made investigation of
these electrode materials as ammonia oxidation electrocatalysts im-
possible.  All other tested materials are reported in following sub-
sections .


     Titanium, Tantalum, Zirconium and Niobium

     The metals classified in this group showed complete passivity
in the entire investigated anodic potential region due to instant
formation of an oxide film.  The only current observed was the "leak-
age current" in the oxygen evolution region.  Metal surfaces in this
state did not show any catalytic activity for the investigated
oxidation reaction.

     Proprietary Catalysts

     Proprietary metal oxide catalysts developed by Esso Research
and Engineering Company for electro-oxidation of hydrazine and for
oxidation of propylene and isobutylene were tested for use in the
ammonia oxidation process with the results given in Table 3.

     Samples designated with catalytic ability of A gave observable
oxidation currents;  samples designated  B gave a barely detectable
oxidation rate;  and samples designated  C showed no oxidation
current under the experimental conditions indicated.

     On the basis of data given in Table 3 it was concluded that the
proprietary oxide catalysts tested have little practical value for
ammonia electro-oxidation, since the highest oxidation rate obtained
(20 jUA/cm2 with catalyst 377-87-1) is too low to be feasible in
practical unit operation.
                                 29

-------
                               TABLE 3
          Catalytic Properties of Metal Oxide Catalysts
Catalyst
377-87-1
377-87-2
377-87-3
377_87-4
377-87-5
377-87-6
377-87-7A
377-87-7B
377-87-8
377-87-9
377-87-10
377-87-11
377-87-12
Note 1:
Note 2 :
Note 3:
Note 4:
Catalytic Maximum current General
'fy
ability pH density[jLtA/cm ] Observations
A 8.15
B 6.55
C 6.55
C 6.55
A 6.63
B 6.63
B 6.63
C 6.60
C 6.60
C 8.15
B 6.60
A 8.15
B 8.15
High background current
Unstable in base
Unstable at given pH
Poor reproducibility
20
6
0
0
16
3
7
0
0
0
8
10
6




Notes 1 and 2
Note 2
Note 3
Notes 1 and 2



Notes 1 and 4
Notes 1 and 4
Note 2







      Proprietary DuPont fuel cell catalysts with semi-conducting
properties and with non-precious metal base were recommended by the
supplier for testing in the ammonia electro-oxidation process.
Electrode samples designated IA-1 to IA-4 by the supplier were tested
as well as some of electrodes prepared in our laboratory from powder
catalyst designated Mo-0-5 AC-5, but no indication of catalytic
properties was found for the ammonia oxidation process.
      Rare Earth Oxides

      Gadolinium, cerium, neodymium, terbium' and thulium oxides were
tested anodically using rotating electrodes prepared according to the
procedure described in the experimental part of the report.  Rare
earth oxides were admixed with graphite in the ratio 15:1 and coated
on titanium substrates.  Potentiodynamic tests in solutions containing
110 mg/1 of ammonia at pH of 7.95 to 8.0 gave no indication of electro-
catalytic activity for any of the tested materials.
                                  30

-------
      Non-Precious Electro-Deposited Metals and Oxides

      A nickel catalyst when prepared in the form of a very finely
divided black deposit has been found to be a very good catalyst for
some other oxidation reactions.  Tests were performed using nickel
black as the anode in electro-oxidation of ammonia but no catalytic
activity was observed.  The same result was found with a smooth nickel
anode.

      Manganese dioxide electrodeposited on a porous substrate has
been successfully used as a fuel cell electrode.  The possible
application as an ammonia oxidation catalyst was tested by electro-
deposition of manganese dioxide on titanium wire and an electrode
so prepared was submitted as an anode to the usual testing procedure.
A dark grey deposit of Mn02 was formed on the titanium wire subjected
to the electrodeposition procedure described earlier, but no oxidation
current was observed when the electrode was tested anodically in the
ammonia oxidation reaction.

      Noble Metals

           Bright Noble Metals

           Iridium, gold, platinum, palladium and rhodium were tested
anodically.  A very low residual current was observed in experiments
with iridium.  Current was higher with rhodium and gold and sub-
stantially higher with platinum and palladium with a wide region of
electrostability.  No catalytic activity was observed.  Experiments
with smooth platinum were repeated with the solution pH varied from
8 to 12, but no meaningful results were obtained probably because the
exchange current of the ammonia reaction on this electrode was low
relative to electrolyte impurity currents.


           Electrodeposited Noble Metals

           Rhodium-plated rhodium, iridium-plated iridium, palladium-
plated palladium, platinum-plated tantalum and titanium and platinized
graphite were tested as the anodes for electro-oxidation of ammonia.

           Electrodeposition of rhodium gave a fairly bright metallic
deposit.  In use this electrode gave very low background currents and
no appreciable ammonia oxidation current.  This is typical of behavior
of polished metal surfaces previously tested as oxidation electro-
catalysts.

           Electroplating of iridium is known to be a very difficult
procedure with only a few electroplating references available for
obtaining iridium black deposits.  The procedure by Ovenden gave a
dark grey deposit of iridium, which was completely inert in the pro-
cess of ammonia oxidation.
                              31

-------
           The freshly electroplated palladium black catalyst showed
some activity at an ammonia concentration of 100 rng/1, but no
evidence of any activity at lower concentrations.

           Reproducibility of the experimental results obtained with
platinized graphite electrodes was not good, but resulting oxidation
currents were slightly lower than those obtained with platinized
platinum catalyst.  The general conclusions are in accordance with
those observed with platinized platinum.  The region of optimal
oxidation rate shifts to higher potentials as the concentration of
ammonia and pH of the solution decrease.  Oxidation rates increase
with ammonia concentration under otherwise similar experimental
conditions.

           Platinization of titanium and tantalum resulted in black
deposits.  These were coherent on titanium, but were not very coherent
on tantalum.  On the basis of the oxidation currents obtained and
physical appearance and properties of two electrodes i-t can be con-
cluded that platinized titanium is preferred to platinized tantalum.
Greater currents were observed under otherwise similar experimental
conditions in runs with platinized titanium  (0.3 mA/cm  compared with
0.18 mA/cm  for tantalum) and the platinum deposit on titanium showed
better coherence probably due to more favorable electrode surface
conditions obtained prior to platinization.

           Oxidation currents obtained with platinized titanium were
about one-eighth those obtained with platinized platinum and about
one-sixth those which resulted from using platinized graphite as the
electrocatalyst at the same experimental conditions.
           Commercial Electrodes

           Work on the electrocatalyst screening included testing of
DSA  (dimensionally stable anode) electrodes with heavy coating from
Electrode Corporation, Painesville, Ohio.  The coating consists
primarily of ruthenium oxide and titanium dioxide.  Very high back-
ground currents  (ranging from 200-250 mA/cm2), but negligible ammonia
oxidation rates, were observed in tests when  these electrodes were
used as anodes.

           Two different types of fuel cell electrodes  (American
Cyanamid- Company) with platinum load of 9 and 10 mg/cm  on titanium
screen backing were tested.  Maximum current  densities for both cases
 (130 and 150 JUA/cm2 in' a solution containing  50 mg/1 NH.J were much
too  low for the effective use of the catalyst.  Both low effective-
ness and high cost of the electrode are strong reasons against the use
for  this work.
                               32

-------
           Noble Metal Blacks

           Commercial platinum black, rhodium black, iridium black
and palladium black catalysts were tested alone or in admixture
with graphite coated on titanium substrate with Teflon T-30 binder.
Composition of the tested platinum black electrode was 74% Pt,
7.4% graphite and 18.6% binder with the total amount of noble metal
for surface unit averaging 70 mg/cm?.  Maximum currents observed
were 0.85 mA/cm  for this fairly high platinum loading.  Rhodium
black and palladium black electrocatalysts gave no observable
oxidation current.  Commercial iridium black catalyst was tested
alone and mixed with platinum black catalyst.  Similar catalytic
mixtures were prepared by diluting platinum black with rhodium black
and palladium black catalysts.  Results from potentiodynamic tests,
where single noble metal blacks and their mixtures were used as the
anodes for the ammonia electro-oxidation, are shown in Table 4.
                             TABLE 4
       Ammonia Oxidation Rates Obtained with Noble Metal Blacks
Catalyst
Maximum Current
   Density*
   [jUA/cm ]
                                  Portion of Noble
                                  Metal in Mixture
                                        [%_]
Weight Ratio
   of the
Components
                                                                 ***
Rh
Pd
Ir
Ir -
Pt -
Pt -
Pt -
Pt -
Pt -


h G**
I- Ir
h Ir + G
1- Rh + G
h Pd + G
f- G
0
0
17
0
750
450
175
175
750
90
90
90
79
85
71
69
67
74
.0
.0
.0
.2
.6
.5
.2
.8
.0
-
-
12
2
10
10
4:
10


:1
:1
:5:2
:6:2
23:1
:1
      *  Recorded at potential sweep rate of 0.2 V/min

     **  G = Graphite

    ***  Binder content varied between 10 and 20%
           Iridium black showed very minute oxidation currents at high
noble metal surface concentration but negligible currents at surface
concentrations at which considerable oxidation rates were observed
with platinum catalyst  (750 /^/cm2 at 74% Pt).
                                 33

-------
           Dilution of platinum with iridium resulted in a slight
decrease of the oxidation current compared to a value obtained for
pure Pt with the same total amount of noble metal per unit surface.
The addition of rhodium black and palladium black to platinum black
diminished oxidation rates considerably with the same surface con-
centration of noble metal.

           Overall, the results of mixing platinum black with the
other noble metals were not favorable.  In the best case  (dilution
with iridium) almost the same catalytic behavior as for pure platinum
catalyst was observed.  Dilution with iridium does not lower electrode
cost.  However the promoter effect of  Iridium (which is not active
itself) was a significant discovery and suggested that a search be
made for lower-cost promoter-diluents.


           Platinum Black Catalyst

           A review of the catalysts which oxidize ammonia with any
reasonable rate is given in Table 5.  On the top of the list is
platinized Pt followed by platinized graphite with a slightly smaller
oxidation rate, then commercial platinum black catalyst with 50% less
activity and finally platinized titanium and platinized tantalum.  As
seen from  Table 5 all previous work revealed platinum as the only
material tested thus far, which would sustain a significant current
density.

                             TABLE  5

             Comparison of Ammonia Oxidation Catalysts

                             Maximum     ^               Ammonia
    Catalyst              Current Density        pH   Concentration
                                                         [mg/1]
Platinized Pt
Platinized graphite
Platinum black (74% Pt)
Platinized Ti
Platinized Ta
2.3
1.65
0.75
0.3
0.18
7.80
7.85
8.10
8.15
8.15
100
51
100
100
100
        * Recorded at potential sweep rate of 0.2 V/min.


           For the process to be economically feasible the catalyst
cost should not exceed $5 to $6 per square foot of electrode.  This
demand requires even distribution and reasonably complete utilization
of the electrocatalyst.  Since cost limitations necessitate a very
small quantity of active material, feasibility would be aided enorm-
                                 34

-------
ously by the existence of a diluent which would not adversely affect
the properties of the material being diluted.

           Graphite, generally used for this purpose, and tungsten
bronzes, to which was*  attributed the property of promoter-diluent
for platinum catalysts in some electro-organic reactions, were used
as diluents in further experimental work.  The main emphasis of the
continuing work was placed on the preparation of high performance
electrodes with low platinum loadings.

           In the electrode preparation technique, attempts were made
to deposit the catalyst for optimizing the electrochemical character-
istics of the electrode.  This requires depositing all or most of the
catalyst in optimum location, i.e. on the exposed surface.

           A series of electrodes were prepared and tested anodically
in an attempt to optimize the catalytic and adhesive properties of
the coated layer and to minimize the amount of platinum per unit area
of electrode surface.  Coating techniques #1 and #2  (see 'Page 1-1)
were compared.  On the basis of photomicrographs  (Figure 17) there
appeared to be difficulties with the use of method #1.  Cracks on the
surface of a heat treated electrode can be observed  (Figure 17c) with
platinum distributed over the surface in the form of scattered
agglomerates (Figure 17d).  Somewhat better platinum distribution and
apparently closer adhesion to the tungsten bronze sublayer can be
noticed on the hot pressed electrode (Figures 17a, b, and 18),.  As a
result, coating technique #2 was used for the electrode preparation
in further work.

           To determine the effect of Teflon content, oxidation rates
were determined with electrodes containing equal amounts of platinum
but with Teflon content of 20 to 50%.  The maximum currents from
these experiments are plotted as a function of Teflon content and
ammonia concentration in Figure 19.  Independent of the ammonia con-
centration the best performance is observed with a Teflon content
between 25 and 30%.  With increased Teflon content oxidation rates
decreased.  This is explained by the hydrophobic properties of Teflon,
which cause poor contact with electrolyte.  With increasing Teflon
content the adhesion of the catalyst layer was increased as expected,
but this made a poorer electrode on account of the lower oxidation
rates.  The region of Teflon content below 20% was not investigated.
because of poor adhesion to the electrodes.  Giner, et  al  (11) have
succeeded in the preparation of electrodes with Teflon content in the
region of 10 to 20% Teflon.  They obtained relatively symmetrical
curves of current vs. Teflon content with low currents at low Teflon
content.

           Specific activity of the catalyst for the electro-oxidation
of ammonia was investigated as a function of platinum loading using
tungsten bronze and graphite-containing electrodes with a platinum
                                  35

-------
3
                                                                         3

                           Figure 17.   Comparison of the surfaces of hot pressed (a and b) and heat treated
                                       electrodes (c and d) each containing tungsten bronze and 1.0 mg Pt/cm2.
100X

-------
LO
-J
                   Platinum black catalyst
                   Tungsten bronze + T-30
                  Titanium metal
                               Figure 18.   Photomicrograph of a cross-section of a hot pressed electrode
                                            containing 1.0 mg Pt/cm2. M  50 X

-------
    150
    100 _
<
c
0)
D
O
-a
o
c
E
D



X
to
                  Platinum Load 20 mg/cm
                       10              20             30             40


                              Teflon Portion in Catalytic Mixture, %
               Figure 19.  Effectjpf Teflon Content on Performance of Electrode
                                       38

-------
                             2
load between 0.5 and 20 mg/cm .  With tungsten bronzes electrode
performance was erratic and non-reproducible  (Table 6).  As a result
the correlation between platinum loading and specific activity was
poor.  It was not possible to determine whether there was any ad-
vantage to lower catalyst loadings.  With graphite the results were
quite reproducible.  Data for one ammonia concentration are shown
in Table 7.  The improvement in specific activity with decreased
platinum loading was very clear.  A six-fold increase in specific
activity was observed between  20 mg/cm  and0.5 mg/cm .  At the
0.5 mg/cm  loading the specific activity with graphite was three
times as great as with tungsten bronzes.  This comparison is shown
in Table 8.


                             TAJBIE  6
      Specific Activity of Tungsten Bronze-Containing Electrodes
                     with various Platinum Loadings
  Sweep Rate =0.2 V/min                   pH = 7.8 0.1
    Platinum Loading       Ammonia Cone.     Specific Activity
         sQ2.!	Isa/il	
30
0.5 58
100
28
1.0 56
110
30
4.0 60
115
26
10.0 52
30
15.0 61
121
30
20.0 61
121
11.0
21.0
40.0
12.0
16.0
22.0
5.0
5.0
6.3
10.0
17.5
19.0
21.7
28.3
10.7
12.7
17.5
                                 39

-------
                            TABLE  7
       Specific Activity of the Electrocatalyst as a Function
                      of Platinum Loading
SR = 0.2 V/min
Platinum Loading
[nig/cm^]
0.5
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
15.0
20.0
pH = 7.8 0.1
Peak. Current
[/LlA/cm2]
56
139
114
120
124
150
174
175
240
270
200
328
360
C = 121 mg/1 NH^
Specific Activity
[/UA/mg Pt]
112
139
58
40
31
30
29
25
30
30
20
22
18
                             TABLE 8
         Comparison Between the Performance of Tungsten Bronze
         and'-Graphite Containing Electrodes with 0.5 mg Pt/cm

Sweep Rate = 0.2 V/min	gHj= 7.8 0.1	
   Ammonia Cone.                   Specific Activity CjLtA/mg Pt]
     [mg/1]                        Tungsten BronzeGraphite
       30                               11                  50
       60                               21                  78
      120                               40                 112
                                  40

-------
Continuous Flow Apparatus


      Cell Design

      The design equation was derived from a material balance around
a differential length > flew path.  In this derivation, F-, is flow
rate in ml/sec, C is local concentration in moles/liter, i is current
density in mA/cm  and T] is current efficiency.  L is total flow path
length and dL is differential flow path length both in cmf w is path
width in cm.  ^ is 96,500 A-sec/eg and  n  is 3 eq/mole.
F
 d
                                dL
                                       +dC
                              77iwdL
                               n5"
                                                        [8]
                                -?
If, as a first approximation, the current density is assumed to be
constant along the flow path, equation  [9] can be directly integrated
to give:

                                ?7i   wL
           F^(C.  - C   1  =      *Vg                  [10]
            d  in    out         rSf-
This approximation was used in setting up the experimental cell.  For
greater accuracy  i  should be expressed as a function of C.  This
refinement was not considered necessary for the present study.
      Cell Operation

      The results from four typical experiments employing electro-
plated platinum catalysts are presented in Table 9.  Feed concentrations,
volumetric feed rates and spacer thickness were varied from experiment
to experiment.  Ammonia conversion increased with increased current
density up to a limiting value of about 0.4 mA/cm2, when it started
decreasing as the corresponding potential suddenly rose to the value


                                 41

-------
where oxygen evolution starts.  An optimum anodic potential is found
to be in agreement with previous determinations  (micro-tests), i.e.
between 350 and 550 mV vs. SHE.  With the exception of a relatively
high current efficiency at very low current density (r? = 0.402 at
0.158 mA/cm2), the current efficiency varied between 12.8 and 28.2%
and averaged 20.9%.

      Commercial powdered platinum catalyst was tested in the con-
tinuous flow cell at two different platinum loadings, about 50 mg
Pt/cm2 and 1 mg"Pt/cm2.  The experiment with the high platinum loading was
carried out at two different current densities, 0.238 and 0.397 mA/cm^,
yielding 23 to 40% ammonia conversion and proceeding with 23 to 30%
current efficiency (Table 10).  No significant change in catalyst
activity was observed during the experiment (20 hours).  No meaning-
ful results were obtained when the low platinum loading anode was
used in the continuous flow cell experiments.
Estimated Unit Cost

An estimate was made of process economics for the ammonia oxidation
unit using a platinum catalyst.  This estimate is based on a plate
and frame design similar to Ionics electrodialysis equipment. Since
the total currents are low, a back-to-back arrangement with each anode
and each cathode serving two cells is contemplated.  This should
reduce electrode fabrication costs and allow the cathodes to be made
of a comparatively inexpensive material such as steel, carbon or
graphite.

The energy required for treatment of 1000 gal. of water, P, is given
by

                  P = I-t-V
                           3600-MW
                    = 0.018-ACV                      [11]
where P is in KWH/1000 gal., l-t is coulomb load in amp.hrs/gal., V is
voltage par cell, A C is the decrease in concentration in mg/1, n is
3 eq/mole, ^ = 96500 coulomb/eq and MW is 17 gin/mole.  The voltage
is estimated at 0.6 volt per cell.  This assumes an anode potential
of +0.4 volt, a cathode potential of -0.1 volt  (both with respect to
a standard hydrogen electrode) and 0.1 volt for overcoming the path
resistance.  This would result from a c.d. of 1.5 mA/cm2, a solution
conductivity equivalent to that of 0.01N NaCl and a path length of
0.03 inch.  Total voltage is virtually independent of current density.


                                 42

-------
                         TABLE  9

   Experimental Data fr.om. the Continuous Flow Electrolytic
      Cell with Platinum Electroplated Titanium Anode	

       2                                    3    Cl  ~ C
i[mA/cm ]        E [Volts]     C [mole/1 ]xlO     -i	     T)
                  H             O                   C.
                                                     1


C. = 1.69xlO~  mole/1, F  = 100 ml/hr,  V  = 0.402 cm/sec, t  = 227 sec
 i                      d                o                  R

0.158          0.352-0.392       1.19        0.295           ' 0.402
0.317          0.432-0.457       1.01        0.402            0.273
0.397          0.502-0.527       0.806       0.524            0.282
0.476          0.702             0.923       0.453            0.368


C. = 1.6xlO~3 mole/I^ = 75 ml/hr, V  = 0.29 cm/sec, t  = 300 sec
 1                    a              o                 R

0.238          0.422-0.437       1.01        0.37             0.237
0.397          0.442-0.457       0.69        0.57             0.220

C. = 1.78xlO~3 mole/1, F^ = 75 ml/hr, V  = 0.29 cm/sec, t  = 300 sec
 i                      a              o                 R

0.238          0.457-0.502       1.33        0.253            "0.125
0.317          0.477-0.522       1.22        0.315            0.165
0.476          0.627-0.682       0.935       0.475            0.170

C. = 1.78xlO~ mole/1, F, = 70 ml/hr, V  = 0.54 cm/sec, t  = 163 sec
 i                     a              o                 R

0.238          0.337-0.362       1.25        0.298            0.200
0.317          0.347-0.352       1.15        0.354            0.177
0.476          0.447-0.557       1.06        0.404            0.181
                             TABU: 10
             Experimental Data Obtained with Commercial
             _ Platinum Black Catalyst _

30 mg Pt/cm2   C^= 1.69xlo"3mole/l, F =110 ml/hr, V =0.442 cm/sec ^=206 sec
i[mA/cm2]     C [mole/1 ]x!03       ci " Co               T]
   0.238         1.29              0.233              0.23

   0.397         1.00              0.408              0.30
                                  43

-------
Using V of 0.6 volt and a power cost of K/KWH, the cost of power
for removing 20 mg/1 of ammonia would be 0.22C/1000 gal. at 100%
current efficiency.  The experimentally determined current efficiency
for the process was 25% which results in a cost of 0.9C/1000 gal.
The electrode surface area required to treat one gallon per hour,


                        0.018-A C
A  is given by:
 s
                  A  =
                   S        1

                          2       1
where As is the area in ft /gal'hr   and  i  is the current density in
A/ft  .  An average current density value of 0.05 A/ft2 was obtained
with platinum black-graphite containing electrocatalyst with platinum
load of about 0.5 mg/cm2.  Using this value, A = 7.2 ft2/galhr~1 .

The electrode cost is based on use of 10 mil titanium anodes with a
coating of 0.5 mg/cm2 of platinum.  This thickness was considered
optimal in the trade-off between material cost and satisfactory
physical properties .  Titanium sheet in this thickness currently sells
for about $2.00/ft  in quantity.  Since both sides are used the cost
per square foot of usable area is slightly over $1.00 .  A platinum
coating of 0.5 mg/cm  would require $4 of platinum/ft2 assuming that
gasketed areas were not platinized.  Other components and installations
are estimated at $2.00 per square foot of electrode based on installed
electrodialysis units.  Total installed cost of the unit per square
foot of electrode area amounts to $7.00.

The capitalized cost per 1000 gal., G, using a rate of 7% per annum
and 90% load factor, is given by:
where CA is the installed cost of the unit in dollars per square foot
of electrode.  Using 7.2 ft2/gal*hr~1 for Asand $7.00 per ft2 for CA ,
C = 44.8C/1000 gal.

Total of power cost and capital cost for removal of 20 mg/1 of ammonia
amounts to 0.9 + 44.8 = 45.7C/1000 gal.

Substantial improvement in catalyst activity would be required before
this ammonia removal method could be considered practical.  A twenty-
fold increase in current density from 0.05 to 1.0 A/ ft2 would be
needed to reduce the cost of equipment to a reasonable value of
2.2C/1000 gal.

Although reductions in electrode -cost may be possible, they could not
be great enough to make the process practical at the assumed current
                                44

-------
                    2
density of 0.05 A/ft  .  Installation of the unit and cost of com-
ponents other than anodes is estimated at $2.00 per ft  of anode.
The costs would result in an amortization cost alone of 13C/1000 gal.
This is too high a cost to make electro-oxidation competitive with
other ammonia removal methods.
                              45

-------
                            CONCLUSIONS
Technically, electro-oxidation has a number of desirable attributes
as a means of removal of ammonia from wastewater.  Under conditions
of ammonia concentration and pH typical of treatment effluents, the
reaction proceeds stoichiometrically to nitrogen with no other
oxidation products.  Oxidation occurs in concentrations as low as
2 mg/1.  For the electrochemical reaction the cost of power would
be very low, 1C/1000 9al or less.  The only electrocatalyst on which
the reaction was found to proceed with any reasonable rate, however,
was platinum.  Because of the high price of platinum, the capital
investment would be too high for the process to be economically
feasible.

To reduce the capital cost, a considerable search was made for other
possible electrocatalysts.  These included a wide variety of materials
including noble metals, metal oxides, commercial electrocatalysts
and proprietary catalysts.  No material was found which was an im-
provement over platinum.  Some economic advantage resulted from use
of extenders like graphite and tungsten bronzes.  However, neither
of these was sufficiently effective to lower the catalyst cost per
unit area to a suitable value.

While the process is not economically feasible at the present time,
future work directed toward other processes may result in the develop-
ment of a cheaper catalyst which would operate at a higher current
density. - No approaches to   more economical electrodes are known
at present which would justify additional work,,
                                 46

-------
                          REFERENCES
 1.   T. Katan and R. J. Galiotto,  J.  Electrochem.  Soc.  110,  1022
      (1963).
 2.   H. G. Oswin, and M. Salomon, Can. J. Chem. 41,  1968 (1963).


 3.   A. R. Despic, D. M. Drazic and P. M. Rakin, Electrochim.  Acta
      11, 997 (1966).


 4.   D. Sparbier and G. Wolf, Z. Naturforsch, 19a, 614 (1964).


 5.   E. Muller and P. Spitzer, Z. Elecktrochem, ll',  917 (1905).


 6.   Y.O. Reitlinger, Z. Elektrochem, 20, 268 (1914).
 7.   D. J. G. Ives and G. J. Janz, Reference Electrodes,  Academic
      Press, New York (1961).
 8   P. J. Ovenden, Nature, 179, 39 (1957).
 9.   L. W. Niedrach and I. B. Weinstock, Electrochem. Technol.  3,
      270  (1965).                                               ~
10.   M. E. Straumanis, J. Am. Chem. Soc. 71, 679 (1949)
11.   J. Giner, J. M. Parry, S. Smith and M. Turchan,  J. Electrochem.
      Soc. 116, 1692 (1969).
                                47

-------
    Accession Number
                           Subject Field & Group
                             05D
                                           SELECTED WATER  RESOURCES ABSTRACTS
                                                  INPUT TRANSACTION FORM
    Organization

       IONICS,  INCORPORATED, WATERTOWN, MASSACHUSETTS
    Title

    "THE ELECTRO-OXIDATION OF AMMONIA IN SEWAGE TO NITROGEN"
10
Authors)

Marincic, Ljiljana
Leitz, Frank B.
16
Project Designation

       17010 EED 07/70
                                    21
                                       Note
22
Citation

   Water
23
Descriptors (Starred First)

*Water Pollution  Control,  *Wastewater Treatment, *Cost Evaluation,
 Electrochemical  Treatment,  Method Development
25
    Identifiers (Starred First)
      *Ammonia electrooxidation
27  Abs
-------