NO2 coulometric method đo nồng độ NO2

The air sample enters through the air inlet, is drawn into the sensor by way of the narrow annulus where it comes into intimate contact with the solution contained on the electrode suppo

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=uawm16

Journal of the Air Pollution Control Association

ISSN: 0002-2470 (Print) (Online) Journal homepage: http://www.tandfonline.com/loi/uawm16

Nitrogen Dioxide Detection Using a Coulometric Method

Royal E Rostenbach & Robert G Kling

To cite this article: Royal E Rostenbach & Robert G Kling (1962) Nitrogen Dioxide Detection Using a Coulometric Method, Journal of the Air Pollution Control Association, 12:10, 459-463, DOI:

10.1080/00022470.1962.10468113

To link to this article: http://dx.doi.org/10.1080/00022470.1962.10468113

Published online: 19 Mar 2012.

Submit your article to this journal

Article views: 389

View related articles

NITROGEN DIOXIDE DETECTION Using a COULOMETRIC METHOD

ROYAL E ROSTENBACH, Former Research Director, and ROBERT G KLING,

Experimental Chemist, Mast Development Company, 2212 East 12th Street, Davenport, Iowa

Instrumentation for the

quan-titative measurement of nitrogen

di-oxide is used today in areas of smog

control and industrial hygiene It is

well known that nitrogen oxides are

emitted from motor vehicles' exhaust,

chimneys, stacks, and other combustion

processes By photochemical and

ther-mal reactions nitrogen dioxide combines

with ozone and organic chemicals

resulting in unpleasant atmospheric

smog mixtures In the field of

propel-lants nitrogen dioxide is a fuel oxidizer

As a low cost selective oxidizer for

industry, nitrogen dioxide (or nitrogen

tetroxide) is currently gaining

recogni-tion with a resultant increase of

produc-tion in tank car quantities The

toxic level presence of this gas in any

working area is an industrial hygiene

problem, and its unwanted presence in

any area may present an air-pollution

problem

The adaptability of the Mast Model

724 Series Meter for the monitoring of

nitrogen dioxide has been

demon-strated Wartburg, Brewer, and

Lodge1' 2 have shown that the Meter

operates extremely stable in the range

of 0 to 3 ppm of nitrogen dioxide

when other oxidizing agents are

ab-sent Van Nattan, Drake, et al.,3

have demonstrated the adaptability of

the'Meter as a monitor up to 25 ppm of

nitrogen dioxide Within the past

year the authors have demonstrated

that the Meter will sense accurately

0-5000 ppm of nitrogen dioxide

The Mast Nitrogen Dioxide Meter

is based upon a coulometric system

This is the same one which has been so

successful in the sensing of ozone

Only a brief description of the sensing

system is included here as it has been

reported upon heretofore.4 Figure 1

shows a typical construction of a

micro-coulomb sensor including the basic

components essential.to operation A

chemical solution containing the proper

amounts of reagents is pumped into the

sensor The solution flows in a thin

film down the electrode support, upon which are wound many turns of a fine platinum wire cathode and a single turn of a platinum wire anode, and is deposited in the waste reservoir The air sample enters through the air inlet, is drawn into the sensor by way

of the narrow annulus where it comes into intimate contact with the solution contained on the electrode support, and exits by means of air pump "P."

A small potential is supplied across the cathode and anode by a battery

"B." The current flow is measured by

the microammeter, or a strip chart recorder is used to measure the voltage drop across a range resistor placed

in the circuit

In this coulometric system the current

is measured and this related to the concentration of nitrogen dioxide

When a sensing solution containing potassium iodide is used, iodine is released in solution by the chemical reaction of the nitrogen dioxide Sim-ilarly when a bromide solution (with

no iodide present) is used, bromine is released The current flows, and is measured, only when released iodine or bromine is present in the sensing solution

Some of the typical reactions taking place in the bromide sensing solution are:

NO 2

2NoO4

NO + H2O + Br2 (1)

2 HNO + 2 NO + Br2 (2)

* Presented at the 55ttt Annual Meeting

of APCA, Sheraton-Chicago Hotel, May

20-24, 1962, Chicae-o, Illinois

At normal atmospheric conditions of temperature and pressure the gas present is primarily nitrogen dioxide

Hence, the first reaction is more typical

of what is taking place Nitrogen dioxide and nitrogen tetroxide are usually considered together There is a rapid disassociation and association which may be written as a reversible reaction:

N2O4 ?± 2 NO2 The reaction is dependent upon tem-perature and pressure Nitrogen

tetrox-ide is a colorless gas while nitrogen dioxide is a brown gas

The sensing solution system feeds the solution to the sensor at the rate of approximately 2.5 ml per hour This

is done by a micro-feed pump The pump is driven by a synchronous motor operating a cam which rotates one revolution every hour The bellows

of the pump is actuated by the cam movement An elastic diaphragm sep-arates the liquid in the bellows from the solution being pumped The solution moves through the pump chamber and

is controlled by elastic check valves One complete pumping cycle takes one hour which is governed by the cam and motor noted above Reservoirs, 125

ml capacity, are provided for the new and used sensing solution because the solution is used only once

In the pumping cycle, fluid in the bel-lows is displaced, thus filling and ex-panding a flexible diaphragm The expansion of the diaphragm in turn displaces solution in the solution cham-ber, which flows through the outlet check valve During the refill portion

of the cycle the spring loaded bellows draws liquid from the diaphragm, causing it to collapse This in turn draws solution from the solution reser-voir through the inlet check valve

Fig 1 Microcoulomb sensor.

LLJ

LU

a

o

u

a:

U

o

CO

Z

LU

0 1000 2000 3000 4000 5000

NITROGEN DIOXIDE CONC ( P P M / VOL.)

Fig 2 High concentration nitrogen dioxide measurement with a 1 0 % potassium bromide solution.

The air sample is drawn continuously

through the microcoulomb sensor by a

small precision constant-volume piston

pump The pump is driven by a 300

rpm synchronous electric motor The

pump is adjusted to operate at 140 ml

per min capacity

The instrument requires only

115-volt, 60-cycle power supply and a small

1.34-volt mercury battery The ac

cur-rent is used by the motor for the

solu-tion pump and air pump The battery

supplies a dc potential to the sensor

The microammeter is in series with

the sensor and indicates the current

flowing through the sensor When the

microammeter is by-passed, the elec-trical signal of the sensor can be con-ducted through a phone jack inserted into the socket of a remote recording meter In this way a continuous record

of the sensor's operation may be made

A battery-operated portable unit was developed for nitrogen dioxide measurements at locations where an external power supply is not available

This unit was designed to operate from two rechargeable nickel-cadmium bat-teries or from a 115-volt, 60-cycle power supply The batteries used were 12 volts each (Burgess CD29) and were connected in parallel

.3 4 5 APPLIED SENSOR VOLTAGE

Fig 3 Effect of sensor voltage on sensor current.

.6

A 12 VDC Veo rpm Haydon motor was used to drive the solution pump while drawing an average current of

10 ma A 12 VDC Amglo reed-con-trolled 300 rpm motor was used to drive the air pump while drawing an average current of 50 ma With a total drain of 60 ma, the two batteries, when fully charged, were capable of powering the unit continuously for up

to 16 hours An eight-hour minimum battery-powered period between charge periods was desired This was ob-tained with the two batteries More cells in parallel would have increased this period but also would have in-creased instrument size and weight somewhat

A simple charging circuit was con-structed with a small transformer and

a 500-ma silicon rectifier plus resistors and filter capacitors This charging unit delivered about 95 ma to the two batteries The unit could be operated while batteries were charging In such use, the two motors drew 60 ma current leaving 35 ma available for recharging the two Burgess CD-29 cells at the same time

The sensing cell, solution pump, and air pump used on the battery-powered unit are the same as those used on the Model 724 series ac-powered units and were described previously

Wartburg, et al.,1 and Lodge2 have reported upon the use of the Mast Model 724-1 Ozone Meter as an in-strument for the quantitative detec-tion of nitrogen dioxide For example, the contents of a photolysis chamber made up of hydrocarbons, aldehydes, ozone, and nitrogen oxides were passed through a coarse glass frit heated to about 200 °C Under these conditions the instrument measured nitrogen di-oxide in the range of 0 to 3 ppm in direct relation to the meter reading These quantities of nitrogen dioxide were confirmed by chemical laboratory determinations The response to nitro-gen dioxide on a molar volume basis was 11.6% of that to ozone Van Nattan, Drake, et al.,3 demon-strated the adaptability of the same model meter as a monitor up to 25 ppm of nitrogen dioxide when ozone was practically absent The foregoing investigations used a sensing reagent buffered to pH 7 with phosphates and containing two percent potassium iodide, five percent potassium bromide, and three percent ethylene glycol

Effenberger5 has shown that nitrogen dioxide has greater oxidizing efficiency

at lower sensing solution pH Gluckauf,

et al.,6 have reported work along similar lines on the reaction:

NO2 + 2H+ + 2 I - - *

H2O + NO + 12

In addition, an iodide solution is more readily oxidized than is a bromide

solution in a coulometric system.

In other words, a given nitrogen dioxide

concentration will result in a

con-siderably smaller current when a

bro-mide solution is used in place of an iodide

solution With either solution, the

sensor output is proportional to the

nitrogen dioxide passing through the

sensing cell Therefore, in the effort

to sense and monitor nitrogen dioxide

in the range of 0-1000 and 0-2000

ppm, sensing solutions containing

bro-mide ions at a lower pH of 5.9 were

investigated These solutions were 20

and 10% aqueous potassium bromide

solutions

Operating characteristics of the

micro-coulomb sensor as used in the Mast

Meters are that:

(1) A fixed dc voltage is applied

across the sensor cathode and

anode;

(2) The chemical solution flows over

the electrodes at a fixed flow rate

(1.25 ml/hr typical for ozone

meters and 2.5 ml/hr typical for

nitrogen dioxide meters) ;

(3) The gas sample containing an

oxidant to be measured flows

through the sensing cell at a fixed

flow rate (140 ml/min typical for

ground level applications)

The above characteristics when

con-trolled at fixed conditions enable the use

of the Mast microcoulomb sensing cell

in accurate oxidant measurements

This is true when the

oxidation-reduc-tion reacoxidation-reduc-tion of the total oxidant

con-tent in the gas sample stream is 100%

complete during rapid passage of the

gas sample stream through the sensing

cell or when the oxidation-reduction

reaction is only partially complete,

even less than 1% complete, during

passage through the sensing cell For

example, the oxidation of an iodide

solution by ozone in the gas sample is

approximately 100% complete as the

reactions occur at a very rapid rate

Thus, with conditions of 0.2 VDC

sensor voltage, 1.25 ml/hr KI solution

flow rate, and 140 ml/min gas sample

flow rate, a concentration of five pphm

by volume of ozone results in a sensor

current of one ua Using the same KI

solution, flow rate, and sensor voltage at

the same gas sample flow rate, a

con-centration of about one ppm by volume

of nitrogen dioxide results in a one ua

sensor current The slower reaction

rate of NO2 required a concentration of

20 times that of ozone to obtain the

same one ua sensor current This

suggests that the NO2-KI reaction in

the sensing cell proceeded to only

five percent completion during the

passage of gas sample through the cell

Also, the oxidation of a bromide

solution by ozone in the gas sample is

only partially completed under similar

conditions With sensor voltage of

DC

o

to

Z

uu to

0 100 2 0 0 300 4 0 0 5 0 0

N I T R O G E N D I O X I D E C O N C ( P P M / V O L

Fig 4 Medium concentration nitrogen dioxide measurement with an acidified bromide solution.

0.56 VDC, a solution flow rate of 2.5 ml per hour of 10% KBr solution, and a gas sample flow rate of 140 ml/min, a concentration of about 125 pphm or 1.25 ppm by volume of ozone will result

in a sensor current of one ua With the same conditions of 0.56 VDC sensor voltage, 2.5 ml/hr flow rates of 10%

KBr solution, and 140 ml/min gas sample flow rate, a concentration of about 20 ppm by volume of nitrogen

dioxide in the gas sample will result in

a current of one ua Thus with a bromide solution as used in the above conditions the 03-KBr reaction ap-parently proceeds to about four percent completion and the NO2-KBr reaction apparently proceeds to about one quarter of one percent completion

To restate, with certain sensing cell factors controlled, the sensing cell output current is representative of oxidant

Fig.

0 2 4 6 8 10 12 NITROGEN DIOXIDE CONC ( P P M / V O L )

5 Low concentration nitrogen dioxide measurement with an acidified bromide solution.

461

* 2 " '

n

Fig 6 Nitrogen dioxide sample delivery

apparatus.

concentration in the gas sample whether

or not the oxidant reacts to completion

with the sensor solution By choosing

proper conditions of sensor voltage,

solution composition, solution flow rate,

and gas sample flow rate, the Mast

microcoulomb sensor can be used in the

construction of instruments to

continu-ously measure various concentrations of

ozone or nitrogen dioxide

The original goal of this project was

to develop an instrument to

con-tinuously monitor NO2 over a range of 0

to 250 ppm by volume for unattended

periods of 30 days minimum The KI

solutions were not considered suitable

for measuring as great a concentration

as 250 ppm, although it was realized

that use of a very low (approximately 50

ohms) range resistor with a suitable dc

amplifier circuit would allow for

measure-ments as high as 200 ppm of NO2

Early success was obtained in

measur-ing higher NO2 concentrations with a

solution of 20% potassium bromide and

6% ethylene glycol in water This

solution was used for several months

while efforts were directed to securing

test data that could lead to a dependable

system capable of reproducible NO2

measurements

The slower reaction rate of the

bro-mide solution allows measurement of

high concentrations of oxidant; for

example, 5000 ppm/vol or more of

NO2 Figure 2 is a graph of nitrogen dioxide measurements in the 0-5000 ppm range obtained with a 10% potas-sium bromide solution

When gas measurements cause the magnitude of sensor current to vary, there is a possibility of large sensor currents causing a drop in sensor voltage of 10% or more It was desired

to employ a sensor voltage that could

be varied considerably without causing

a variation in magnitude of the sensor output signal for a given gas sample composition Figure 3 is a graph of data obtained by measuring the sensor output signal while sampling a stable gas sample from a large polyethylene bag and varying the sensor voltage

As shown by the plateau on this graph, varying the sensor voltage from 0.43 to 0.65 volt resulted in the same sensor output signal On this basis, a sensor voltage of 0.55 to 0.57 volt was chosen for applications of the microcoulomb sensor requiring a bromide solution containing no iodide ions All sub-sequent NO2 measurement investiga-tions were made with an applied sensor voltage of 0.55 to 0.57 volt This voltage was obtained usually from a small 1.34 volt mercury cell A zener diode-controlled dc voltage from an

ac source was also proved satisfactory

in one system tested

Some salting out was observed in the 20% KBr solutions A change of sensor solution to 10% KBr in water with no ethylene glycol added was made and there was no apparent change in performance Other solution composi-tions could be used in the microcoulomb sensor for nitrogen dioxide

A potassium bromide solution con-taining citric acid was found to have a sensitivity between that of the iodide and the bromide solutions Other solutions could probably have been investigated with probability of suc-cess The use of mercury salts, how-ever, is not recommended in this type

of sensor solution

Figure 4 is a graph of tests in the 0 to

500 ppm range using a modified acidic bromide solution of pH 2 This solu-tion consisted of 10% potassium bro-mide, 1% citric acid, and 3% ethylene

glycol Figure 5 is a graph of tests in the 0 to 10 ppm range using the same modified bromide solution It is be-lieved that improved low concentra-tion measurements in ranges of 0 to

1 ppm, and 0 to 10 ppm of nitrogen dioxide, could be obtained with a modified iodide solution

The matter of nitrogen dioxide sampling for calibration is a unique one The method was used originally by Wright Air Development Center and

is shown in Fig 6 A 100-ml glass syringe is held in an upright position with clamps A small rotor is mounted

on the plunger The rotor and plunger are turned by means of a stream of compressed air or gas When an appropriate volume of nitrogen dioxide gas is injected into the syringe holding some air, and when more air is drawn into an exact volume, a capillary tube is attached by means of an adapter to the normal opening of the syringe When the rotor and plunger are turned, a small constant volume per unit of time

of nitrogen dioxide is released to the Nitrogen Dioxide Meter

The Nitrogen Dioxide Meter was calibrated using the above described apparatus Specific amounts of nitro-gen dioxide used over definite time peri-ods led to the formation of the calibra-tion and other curves shown herein Solution temperature is recognized as

a factor The results reported here were conducted within the range of 70°

to 95 °F No correction was made within this temperature range Be-cause the sensing solution is a salt solution it is not planned to operate below 35 °F without means of heating the instrument

Since nitrogen dioxide is a liquid below 21.3 °C and strongly attracted

to moisture, sample preparation for meter calibration has been found difficult at low ambient temperatures and at very high relative humidities Since many of the investigations were made at or near 100% relative humidity, care had to be exercised in maintaining the sample transfer syringes free of moisture droplets The presence of a small droplet of water in a syringe was found capable of altering the NO2

Fig 7 Mast Model 724-11 nitrogen dioxide

meter with a recorder, front view.

462

Fig 8 Mast Model 724-11 nitrogen dioxide meter, front view, cover open. Fig 9 Mast Model 724-11 nitrogen dioxidemeter, rear view, cover open.

concentration of a sample in either

direction The moisture could absorb

a considerable portion of an NO2 sample

and then later release a considerable

quantity of NO2 to a more dilute gas

sample

This test program led to the

develop-ment of the Mast Model 724-11

Nitro-gen Dioxide Meter This instrument

is similar in size and construction to

the Mast 724-2 Ozone Meter Figure 7

is a front view of the Meter and a strip

chart recorder, and shows the

micro-ammeter on the front cover Figure 8

is a front view of the Nitrogen Dioxide

Meter with front cover open and shows

the microcoulomb sensor, control knob,

and waste solution reservoir Figure 9

is a rear view photo with the rear cover

removed and shows the solution pump,

air pump, and fresh solution reservoir

The interrelationship of ozone and

nitrogen dioxide as detected by the

Mast Model 724 Series Meters has been

covered Some observations have been

made of the interference of sulfur dioxide

Using the modified bromide solution

containing citric acid, the Nitrogen

Dioxide Meter was found to give very

low positive readings to quantities of

sulfur dioxide in an air sample A

sample of 390 ppm gave a reading of

only 0.28 microampere Successive air

dilutions indicated that even this low

output signal was representative of the

sulfur dioxide concentration, as 195

ppm gave a 0.18 ua reading, 98 ppm gave

a 0.11 ua reading, and 49 ppm gave a

0.06 ua reading The same solution

would have required about 2 ppm of

nitrogen dioxide to give a 0.28 ua

reading, as obtained from 390 ppm

of sulfur dioxide Preliminary tests

were made in measuring nitrogen

dioxide and sulfur dioxide mixtures

When the ratio of sulfur dioxide to

nitrogen dioxide was one part to eight

parts by volume, the interference was

not significant and the nitrogen dioxide

reading appeared normal As the

pro-portion of sulfur dioxide was increased,

a negative interference was observed

resulting in lowered nitrogen dioxide

readings This interference appeared

greater at higher gas concentrations

For example, at a one to two ratio the

output signal for 330 ppm was reduced

by about one third but appeared norma1

for 40 ppm of nitrogen dioxide with the

same ratio or 20 ppm of sulfur dioxide

present A one to one ratio of sulfur

dioxide reduced a 330 ppm nitrogen

dioxide reading by 60%, and a three to

one ratio of sulfur dioxide reduced a

165 ppm nitrogen dioxide reading by

about 63% Normal readings were

again obtained as soon as the sulfur

dioxide content was reduced or omitted

from the gas sample

The sensitivity to sulfur dioxide was

very low with the solution composition

and other sensor characteristics chosen

for nitrogen dioxide measurement in these tests It appears feasible that these factors could be varied to increase sensitivity to sulfur dioxide if desired, and a similar instrument be developed for the continuous measurement of sulfur dioxide This approach is sched-uled for preliminary investigation in the near future

On the matter of performance, the response time of a sustained level of

2000 ppm of nitrogen dioxide is 50%

of full reading in 0.2 minutes, 87% in one minute, and full reading response occurs in less than five minutes The detector recovers in less than 12 seconds after removal of nitrogen dioxide

Recovery down to 20% of an actual concentration reading occurs in less than

60 seconds Full recovery occurs in less than five minutes

The above response time values are considered typical for a straight bro-mide solution Faster and slower values have been observed Usually 85% of full reading is obtained in less than one-half minute with an acidified bro-mide solution Various factors can affect this response time Conditions

or characteristics of various sensors and sensor gas sample inlet tubes are among these factors Clean, dry Teflon sur-faces would be considered an optimum carrier for NO2 sampling

REFERENCES

1 A F Wartburg, A W Brewer, and James P Lodge, Jr., "Evaluation of a Coulometric Ozone Sensor." Presented

at 138th Meeting, American Chemical Society (September 1960).

2 James P Lodge, Jr Private communi-cations (May 16, 1960, and April 24, 1961).

3 W R Van Nattan, Tamas A Drake,

et al., "NOs Calibration for Mast De-velopment Company's Ozone Meter MDC Model 724-1," a Martin Com-pany, Denver, Colorado, communica-tion (November 1960).

4 Gifford M Mast, "Research and De-velopment on the Instrumentation of Ozone Sensing," Instrument Society of America, 18-SF60 (May 1960).

5 E Effenberger, L Anal Chemic, 134:

106-109 (1951).

6 E Gluckauf, H G Neal, G R Marten,

and F A Paneth, J of Chemical

Society, 48: 2045(1946).

CLASSIFIED

SENIOR RESEARCH SCIENTIST (AIR POLLUTION) Conducts

origi-nal research in air pollution field Re-quire college graduation and three years of professional research experience

in the field Experience requirement varies for holders of master and doctoral degrees in field Salary range $10,520

to $12,575 All Civil Service benefits

Write: Mr Richard H Mattox, Direc-tor of Personnel, New York State Department of Health, 84 Holland Avenue, Albany, New York for applica-tion procedure

PHS TO SPONSOR TRAINING COURSES

The Public Health Service, through its Division of Air Pollution, will conduct the one-week training course,

"Measurement of Airborne Radio-activity," December 10 to 14, 1962, at Cincinnati, Ohio The course is de-signed for engineers It focuses atten-tion on sampling and analysis of radio-activity in air, with time devoted to interpretation of the resulting measure-ments, in terms of health significance Completion of the course, "Basic Radiological Health" (given in Cin-cinnati, October 22 to November 2 and

at Rockville, Maryland, November

26 to December 7) is recommended prior training

The Public Health Service, through its Division of Occupational Health, will conduct a series* of related training courses November 26 through December

21, 1962, at the Occupational Health Research and Training Facility, Cin-cinnati, Ohio "Industrial Hygiene Engineering," designed for industrial hygienists and engineers in the field of occupational health, and "Industrial Hygiene Chemistry," for chemists and chemical engineers in this field, are given concurrently, November 26 to December 7 Trainees in both courses meet together for the first week for instruction in industrial hygiene and medicine, toxicology, and principles pertaining to the evaluation of the environment Meeting separately the second week, work for the hygienists covers temperature and humidity meas-urements, illumination, noise measure-ment and control, and industrial ventila-tion; for the chemists, laboratory analyses for lead, free silica, and solvents; and spectroscopy, polarog-raphy, X-ray diffraction, electron microscopy, and gas chromatography Related courses, "Industrial Exhaust Ventilation" and "Emission Spectro-graphic Techniques" follow on December 10 to 21 and 10 to 20 re-spectively The first provides valuable training for design engineers and hygiene engineers charged with evaluation of the industrial environment The second acquaints the trainee with the principles and applications of emission spectro-graphic methods as applied to analysis

of environmental and biological samples Descriptions of these courses are

given in the Training Program Bulletin.

Applications or requests for information should be addressed to the Chief, Training Program, Robert A Taft Sanitary Engineering Center, 4676 Columbia Parkway, Cincinnati 26 Ohio, or to a PHS Regional Office