Capnography : review of analysers for CO2 detection in gases

Capnography : review of analysers for CO2 detection in gases

Citation for published version (APA):

Nauta, A. P. (1987). Capnography : review of analysers for CO2 detection in gases. Eindhoven University of Technology.

Document status and date: Published: 01/01/1987 Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne

Take down policy

If you believe that this document breaches copyright please contact us at: openaccess@tue.nl

providing details and we will investigate your claim.

DEPARTMENT OF ELECTRICAL ENGINEERING EINDHOVEN UNIVERSITY OF TECHNOLOGY

Division of Medical Electrical Engineering

CAPNOGRAPHY

REVIEW OF ANALYSERS FOR COz DETECTION IN GASES.

Au~e P. Nauta

The investigation described in this report forms part of my training for electrical engineering. performed from 0'-'2-'86 to 20-02-'87.

supervised by prof. dr. i r . J.E.W. Beneken and drs. M. Stapper.

THE DEPARTMENT OF ELECTRICAL ENGINEERING OF THE EINDHOVEN UNIVERSITY OF TECHNOLOGY DOES NOT TAKE RESPONSIBILITY FOR THE CONTENTS OF REPORTS

SUMMARY.

In this report a literature review, of some methods for detection

of CO2 is given.

A major part of articles have been found in two indexes: Exerpta

Medica and Index Medicus. Via reference l i s t s and the Science

Citation Index more papers have been collected.

The basic principle of each method found is described, and every

principle is supplemented with one realisation, of which the

method-specific advantages and disadvantages are given. In some

cases alternative techniques are mentioned.

There are large differences in specifity, sensitivity, accuracy,

stability, speed and r e l i a b i l i t y , between the specific

applica-tions. Not every paragraph contains the same comparable

characte-r i s t i c numbers, dependent as I was upon literature. Prices are

never mentioned, sizes and power consumption in some cases.

Review articles

measurements in

not restricted to just detection of

gases, are referenced [O'J to [04J.

TABLE OF CONTENTS Chapter.

Introduction.

1 Some applications o~ CO2 detection.

page. 2 2. 1 2.2 2.3 2.4 2.5 2.6 3 3. 1 3.2 De~initions.

Selectivity or Cross talk. Sensitivity.

Accuracy. Stability. Response time. Calibration.

Quantities and units. Pressure. Concentration. 2 2 2 2 2 2 2 3 3 3 4 4. 1 4.2 4.3 4.3. 1 4.3.2 4.4 4.5 4.6 4.7 4.8 4.9 4. 10 4. 1 1 4. 12 5 6 Methods o~ analysis.

The infra-red analyser. The mass spectrometer. Gas chromatography.

Thermal conductivity detector. Ionisation detectors.

The light spectrometer. Raman spectrometry.

The flueric gas analyser. Photometric analyser. Conductometry.

Paramagnetic CO2 analyser. The PC02 electrode.

The Einstein CO 2 detector. Refractive index measurement. Discussion. Re~erences. 4 4 7 9 1 2 12 13 16 17 19 21 23 25 26 27 28 29 Appendix A Enclosure 1

The input of the computer search in the different databases.

The results of the computer search in Chemical Abstacts and Inspec.

INTRODUCTION.

In the 4t~ year of the study for electrical engineering at the

Eindhoven University of Technology students have to perform an

almost independent project. The work described in this report was

carried out at the Division of Medical Electrical Engineering,

and was supervised by prof. dr. ir. J.E.W. Beneken and

drs. M. Stopper.

The assignment was to make an inventory, from literature, of all

the methods that exist for detection of CO2 in respiratory gases,

also called Capnography.

The importance of CO2 analysis in medical applications is

explained, a short review is given about criteria, quantities and

unities, followed by a description of the methods for analysis;

this description covers only the most recent or most modern

realisation of each of the basic principles.

The literature was gathered by hand from Index Medicus and

Exerpta Medica covering 1976 to 1987. Via reference lists and the

Science Citation Index more papers were found. I did put all my

attention to Index Medicus and Exerpta Medica because, after a

short glance in some different indexes, those two delivered the

largest quantity of useful articles. Since the goal was detection

of CO2 in respiratory gases, the medical orientation of those two

indexes seemed an advantage.

After I gained some experience with the subject, a computer

search was performed in four indexes: Inspec, Chemical abstracts,

Medline (Index Medicus) and EM Base (Exerpta Medica). The inputs

to the different databases, and the amount of articles found is

shown in appendix A. A list with titles found from Chemical

abstracts, 1982+ and from Inspec, 1982+ is enclosed in this

report. Only from those two indexes because new methocs of

detection might be found in disciplines other than medical, and

because Medline and EM Base will overlap a great deal with the

literature already found manually. The titles found with the

-1-SOME APPLICATIONS OF CO z DETECTION. Clinical applications, [ 1 ' ] , [P~], [1:3]:

-Indication of the quality of ventilation and gas exchange of patients subjected to a r t i f i c i a l ventilation.

-Measurement of physiological dead space of the lungs and bronchi.

-Brain swelling during anaesthesia makes surgery difficult and is dangerous for the patients health. The brain volume is related to the brain-bloodflow. Which in turn depends on the CO z tension (PaCO z ) in the arterial blood. By hyperventilation the PaCO z can be reduced, respiratory CO2 analysis gives information about this process.

-Indication of esophageal intubation. Industrial applications:

-Analysis of flue gases. -Control of synthetic gases. -Measurement of air pollution.

-2-2 DEFINITIONS.

2.1 SELECTIVITY OR CROSS TALK. [2']

The ability of analysing one component in a sample without inter-ference by other gases. Influence of the result by a different constituent is coIled cross talk.

2.2 SENSITIVITY. [2']

The sensitivity is the smallest change in concentration, or the smallest concentration of the component under investigation that can be detected.

2.3 ACCURACY. [22 ]

The accuracy displays the measurement and the actual 2.4 STABILITY. [2']

maximum error between a result fro~ a value.

The reproducibility of stability. The change per time units, and is oscillating signal the

a measurement is characterised by the in both zero and amplitude is given in %

called drift. In case of measuring an standard deviation is given.

2.5 RESPONSE TIME. [23 )

In case of a stepwise change of the concentration under investi-gation at time 0 (zero) the following definitions can be given: delay time: from 0 to T~o~ (time when response is 50% of final value); rise time: from T~~ to T9~~ and fall time: from T9 0 % to Te~. Because many autors have their own definitions of response time, one might not find the above mentioned times in the descriptions of all the detection methods.

2.6 CALIBRATION. [24 ]

The accurocy of a gas analyser depends on the exactness of its calibration. Usually gas analysers are calibrated with 0 gas with know composition. Different types are:

- Separation gas, this gas contains all the components that are to be expected during a measurement. Chromatographic decomposi-tions and immunity to cross talk can be verified with such a gas. - Calibration gas, a mixture with all the components that are to be analysed with exactly known concentrations. The concentrations of the different components are about the value that can be expected during a measurement.

- Test gas, has a known composition. To be used for regular chec~ ups of the analyser.

- Blank gas, does not contain any of the components that are to be analysed. It can be used to chec~ zero.

-3-3 QUANTITIES AND UNITS. 3.1 PRESSURE. [3']

The 51-unit for pressure is the Pascal (1 Pa =

(1 bar

=

10e N/mZ ) and hydrostatic units li~e pressure. Some other expressions are: 1 psi

=

torr

=

1 mmHg

=

133.4 N/m z .

3.2 CONCENTRATION. [3']

1 N/m z ), the bar mmHg and mmHzO 6S97 N/mz _{and 1}

For concentration often smaller concentrations refers to the American "miljard". Using ppm weightfraction (w/w), Another representation etc.

ppm, parts per million, is used. For even ppb, parts per billion, is useable. This billion, 109 , _{in the Netherlands called} or ppb has to be specified with either volumefraction (v/v) or molarfraction. for concentrations is ng/g, ng/ml, ml/m3

-4-4 METHOD5 OF ANALY5I5.

The f i r s t ~pp~ratus for measuring carbon dioxide in gases was described by Haldane, [4.0' to 4.04

] , later followed by the

method of 5cholander. In these methods, ~ liquid ~bsorbs the component that has to be ~nalysed; from the ch~nge in volume or the change in pressure in the system the gas concentration could

be c~lcul~ted. Although the Haldane method is quite accurate ( ±

0.01'" ), i t is not described in this report, neither is the 5cholander method, because both are obsolete by now (time consuming, complic~ted ~nd operator dependent) .

4.1 THE INFRA-RED ANALY5ER.

Many different gases (CO, CO z , NzO, HzO, CH4 , Cz H4 ) have a

characteristic absorptionband in the infra-red spectrum between 2 ~nd 15 ~m. The absorption bands are quite sharp. This property can be used by exposing a gas to infra-red light and detection of the transmitted intensities with a photo detector (in-line system) or indire"tly (sample system) as will be mentioned under "Alternatives". Two types of analysers are possible: dispersive and non-dispersive. With non-dispersive systems the gas is exposed to the complete spectrum between 2 and 15 ~m. In order to detect one component a reference gas has to be used. With dispersive systems the light source is filtered and the gas is exposed to just one or just a few different wavelenghts.

A dispersive analyser is described here [4.1') I a non-dispersive

analyser is mentioned under "Alternatives". As can be seen in fig. 4.1 D _{, a broad spectrum beam of infra-red radiation passes} through the respiratory gas via sapphire windows in a tube connecting the patient to a ventilator. The radiating beam is mechanically chopped at a low frequency,

=

'80 Hz. An inter-ference f i l t e r transmits only light with a wavelength of 4.2 ~m at which carbon dioxide has a main absorption peak. The bandwidth is 0.07 ~m. The chopped and filtered radiation is received by a photoelectric Ge-As sensor, the signal of which is amplified. The carrier frequency generated by the chopper is ~mplitude modulated by the transmitted radi~tion. Thus the ~mplitude of the oscilla-tions is a function of carbon dioxide absorption.

"""'---11=3::\'

-fig. 4.10 _{The carbon dioxide} transducer ~nd the initial port of the circuitry of an infra-red CO2 analyser.

-5-A signal processor demodulates the signal by means of an

ampli-fier that is phase-locked to the chopper signal. After

demodula-tion a signal

"u"

corl'esponds to the amount of light received. At

the end of inspiration there is no carbon dioxide in the sensor

head. The signal is sampled and the inspiratory level U1 is held,

and to be used as a reference for the following expiration. An

analog divider gives the quotient U/U,. The division by U,

corrects for changes in the amount of emitted light, absorption

in the windOws, the sensitivity of the infra-red detector and the

primary amplification of the signal. If for example, the windows

become partly blocked by deposits, this will only influence the

current expiration. For the following expiration a new U, will

restore correct amplification.

Selectivity: Because of the use of a very narrow band-width

f i l t e r , there is no interference with other gases

(included water vapor). A contributory factor is th~t

the CO z analyser measures differences between expired

and inspired gases.

Accuracy: End tidal carbon dioxide concentration (%) was compared

with a mass spectrometer. In oxygen-air mixtures:

CO z % (IR) == 0.218 + 0.927 -::- _{CO z % (M5),}

(r = 0.987,50 = 0.22, SEM of intercept = 0.119, Sr:W of'

slope = 0.028). In oxygen-nitrous oxide mixtures:

COz % (I R) == O. 1 5 6 + O. 97 8 3

'C

COz % (M 5) ,

(r = O. 9 8 7, SO = O. 1 5, 5 E M 0 f i n t ere e p t = O. 068, S E

r:

c f

slope

=

0.017). A t-test showed that the differences

in intercept and slope between oxygen-air and

oxygen-nitrous oxide mixtures were not significant (P>D.05).

Carbon dioxide production (Vcoz) (ml/min) was also

compared with a mass spectrometer. For oxygen-air

mixtures: Vcoz (IR) = -0.649 + 1.047 ~c Vcoz (MS),

(r = 0.991, SO = 14.0, 5EM of intercept 2.84, SEM of

slope

=

0.0147). For oxygen-nitrous oxide mixtures:

Vcoz (IR) = 10.56 + 1.006 ·:c Vcoz (MS),

(r

=

0.993, 50 10.1, SEM of intercept

=

2.728, 5EM of

slope

=

0.0174). A t - t e s t showed that the difference

in slope between oxygen-air and oxygen-nitrous oxide

was not significant (P>0.05). The difference in the

intercept was significant (0.01>P>0.005).

Stability: The chopper generates a carrier frequency that is

modulating by the transmitted radiation. In this way

problems inherent in detection of small signals

superimposed upon large drifting ones are solved.

Response time: T,oo~ is < 6 msec.

-6-loop end the epplication of a divider, no zero adjust-ment or calibration is needed.

Feil conditions: When the inspired gas conteins carbon dioxide, the enelyser meesures the increese of cerbon dioxide over thet concentretion, not the ebsolute concentra-tion.

An increese of pressure in the sensor heed will cause en increesing cerbon dioxide signel for two reesons. First, the number of molecules increeses in proportion to pressure. Second, intermoleculer forces that increese with pressure will en hence ebsorption by cerbon dioxide.

Strong points: Porteble. Instant CO2 determination. A gas withdrewal system is unnecessary beceuse the epparetus is operating in-line. No need for calibration. Easy to handle. Two ways of monitoring ere relatively easy to realize; CO2 concentration egainst expired volume or expired CO2 volume against expired volume.

Weak points: Single gas detection. Has to be used in conjunction with a ventilator.

Alternative:A schematic diagram of the principal of e nondisper-sive analysis is shown in fig. 4.1b _{• Infrared light} interrupted with a chopper to get an e.c. signel, enters two sealed chambers containing some CO2 , Weve-lengths of 2.6 and 4.3 ~m are ebsorbed by this CO2 causing i t to expand; when the two chambers receive equal emounts of the incident energy et these wave-lengths, the pressure rise is the seme in both cham-bers. The chambers are separated by a very thin metal membrane pressure manometer, which will be displaced when some of the infrared light is ebsorbed by a semple placed in one of the light paths in the sample cell. The metal membrene forms one plate of e capacitor, the cepecity chenge provides en electricel signel.

~I

CHOPPER ~ ' , . _ .

lAS SAMPLE CELL

CAPACITY ..' PICIC·UP

METAL MEMSRANE

fig. 4.1b _{Schematic of a}

-?-Some figures from [4.12 ] are; no cross-sensitivity of nitrous oxide, end collision broadening where nitrous oxide elters the reeding in the presence of carbon dioxide wes less then 3% of the reeding. The eccuracy is ebout 0.1%, but cen be increesed up to 0.01% when preceutions ere teken [4.13 ] . The response time is dependent upon the semple flow, from 0 to

T.

o _ in 85

msec is possible.

Litereture: In the recent erticle [4.14 ] e deteiled eveluetion of experiments with 13 different CO2 infre-red monitors from 13 different menufecturers is given.

4.2 THE MASS SPECTROMETER.

Mess spectrometers can be distinguished in two types, i.e.

dipole spectrometers end quedrupole spectrometers. In dipolespec-trometers mass seperetion tekes pIece by leeding the ges through e homogeneous magnetic field. The ions will move elong e circle with radius proportional to i(m/e). (m/e) meens the mess/charge retio of the elementery particles. After moving halfway the circle the distinctive masses can be seperetely detected. See fig. 4.2a _.

fig. 4.2a _{The principle of} the dipole mass spectrometer. , ionisation chamber, 2 + 3 +

4 slots, 5 magnetic field, 6 collector,? resistor, 8 am-plifier, 9 display.

In quadrupole spectrometers the ions are led through e cylindri-cal chamber, shaped like the stator of a four poled electrical engine. The poles ere activated with e combination of direct and elternating volteges up to 2 MHz. This errengement works like a mess f i l t e r . Dependent upon the volteges on the poles only one (m/e) velue will reech the end of the cylinder, where e detector is situeted. Other (m/e) velues ere bent eside. By verietion of volteges e wide renge of (m/e) velues cen be detected.

A quedrupole mess spectrometer which consists of en ionizer, a mess f i l t e r end e detector, is shown in fig. 4.2b _{• The ionizer}

electricelly forms ions by sepereting etoms end molecules, or by edding or subtrecting electrons from etoms by e strong electric field. The mess f i l t e r (quedrupole) is excited by e combinetion of d.c. end RF volteges in such e wey es to pess only those perticles which fell within e nerrow renge of mess to cherge retios. The detector measures the ion current passing through the mass f i l t e r . [4.2']

-8-,..--.,

1 - - - + - 1 GAS I

...""-""-...;...

Lf!.l!.CT.J

b . . . . .~,.,:L"

-

....

SPECTROMETER ~--

--.,

:"~J~MA~~J

i

.oNiii.iiONj I fFFlClENCY , AND CRACkING I COllECtION I L..: :.J COMPUTER

fig. 4.2b _{Schematic of the}

quadrupole mass spectrometer. Sensitivity: 100 ppm. (w/w).

Accuracy: For CO2 there is a good quantitative 8eckman L82 CO2 analyser. agreement with a Stability: Over a 27 components are drift is < 0.1 min. period, less than 0.03 volume percent

output variations of all volume percent. Baseline per day.

Response time: 0 - T90~ is = 30 msec. Slower responses for gases which are readily adsorbed onto the flow tube wall and for high molecular weight gases.

Calibration:With (gravimetrically) prepared precision mixtures. Water vapor calibration is performed by humidifying a dry gas of known composition. The dilution of oxygen by water vapor can be measured with a Nernst cell (0 fuel cell) oxygen analyser. The flow measurement subsystem can be calibrated with a turbine pump flow source using timed spirometer collection as a standard of com-parison.

Weekly calibration is sufficient for routine applica-tions, daily calibration for critical applications. Likely causes for failure:ouring ionisation in mass spectrometers

using thermionic filaments as a source of ionising electrons, carbon compound production on such filaments can give rise to spurious carbon dioxide measurements. After sampling gas mixtures in which the volatile anaesthetic agent Halothane was present this CO2 production can rise to the order of 1%. Expired air

-9-contains a CO z concentration

an error of 16%. The solution

modification of the ion source

of 6% so this represents

of this problem is the

cage. [4.22 _{] .}

Strong

When an unknown, unmeasured gas is introduced into the

fixed collector mass spectrometer that computes the

percentage or partial pressure of each gas by

referen-cing it to the total sum of all gases measured, all the

readings will be in erJ'or. [4.23 _{] , [4.2"']}

points: Breath by breath analysis. No sampling pump

needed, because the vacuum in the spectrometer together

with an inlet valve provide a means for gas inta~e.

Multi patient operating possible. Multi gas detection.

The application of a computer provides the experimenter

with a menu of monitoring possibilities.

Weak points: Expensive. Not portable.

4.3 GAS CHROMATOGRAPHY.

In all chromatographic techniques separation of the sample in its

components taKes place by means of transporting thE sam~le

through a column. This column contains the so called stationary

phase, and the sample is led through the column with a carrier.

The stationary phase can be solid or liquid, the carrier a gas or

liquid. Thus four types of chromatography can be distinguished.

Separation takes place because of adsorbtion to the stationary

phase. Chemicals with a strong adsorbtion to the stationary phase

will flow slowly through the column, while chemicals with l i t t l e

:._-

tw~I.- ~:-_=!.-::-:_.

-1.-~.~. ~lll:~C>Z~~

...

-~~{~~~~jg

-...:=.. . ,.- :. SAMPl.E: •D.J ...1.J· II ..., ... .: _ . . I~'. CARil.lE:ll CAS· 40 1I " •. • ,H. :

-= '..

"!.-: _ ATTE:NUATIO!'l • Jl :: _ _ _ . _ CIIAIlT SPE:E:D.III ....11/... :

'coi COl.UWSS • DE:HSl M.le."l. .5... : IlE:CORDE:R • I ....11".1. :

---

.~.:.;.

.-

-~ I=-+-fi- -

-_

.

.

__

.

._.. - I

--- -~ I=~·-fl· _. . . . . •_.. !I...

_--,--.-

tt-§.

_{-_.}

=:/..:=.

-:-==

.~~ ~ ._=1-.:.1-:I=:-':i .-·:J~.""'I:~:

..=-..:.. .

_··:r.

-':f ...

o 4 ,

,

fig. 4.3" togram A typical

chroma-

-;0-adsorbtion flow fast. At the end of the column the chemicals can be detected one by one. In gas chromatography different kinds of detectors are in use, classified in two groups: thermal conducti-vity detectors and ionisationdetectors. In literature I could only find systems which use thermal conductivity detectors, so called catharometers.

A combination gas chromatograph is described by [4.3'], using both a liquid and a solid stationary phase. By this method a dual column/dual thermal conductivity detector is used to separate and quantitate Oz, Nz , CO, CO z and CH. in gas samples. After drying, the sample passes successively through two columns. The f i r s t stationary phase contains a polar stationary liquid, the second is packed with molecular sieve 13-X. At the ends of both columns detectors are placed. The f i r s t column retains CO z , which is eluted after passage of the rest of the mixture (the composite peak). The f i r s t detector thus records two peaks, one correspond-ing to the unresolved gases and the second to CO z . The gases are swept into the second column which separates all the components. The second detector records these gases. CO2 is irreversibly adsorbed on the molecular sieve 13-X and does not elute, which implies changing of this column after saturation.

Peak heights are used in conjunction with calibration plots for quantitative measurements. Alternatively electronic integration of peak areas may be used. Fig. 4.3~ and 4.3~.

,,,,,,,,,,1 CO"lCokItI,w,t'f Ctl' fig. 4.3b _Schematic dual column/dual chromatograph. of the detector

Selectivity: Argon is not separated from Oz, and is present in natural air at 0.9 volume percent. For samples with low oxygen concentration a correction may be necessary depending upon the preparation of the calibration standards.

Polar compounds including retained on both columns will not interfere. Heavier

acid gases are strongly at ambient temperatures and hydrocarbons than methane

11

-are retained somewhat by the polar column and are separated in order of increasing molecular weight.

Sensitivity: With a hot wire detector and helium in a 1 ml sample, in a carrier gas ml/min: Co z 250 ppm (v/v), oz 300 ppm, 500 ppm and CH4 300 ppm.

as carrier gas, flow rate of 50

N

z 300 ppm, CO

Accuracy: Depends upon the standards. Also introduction. Gas of

=

0.3%.

availability of accurate depending on the mode

sampling valves provide

calibration of sample a precision

Stability: Is affected by detector control of carrier gas flow ture. Temp. control within ± control to ± 1% is adequate.

drift, depends upon the rate and system tempera-0.5 °C and flow rate

Response time:

=

O. 1

For CO2 , min.

delay time is

=

1.7 min. rise time is

Calibration: A standard curve of peak height or peak area vs. volume percent is prepared for each component of interest, by analyzing the calibration standards. The calibration plots should bracket the sample concentra-tions. The standard curves should be checked periodi-cally, from daily to once every week dependent on the accuracy required.

If the sample source is natural air, the may need correction for argon present in not separated from the O2 ,

result for O2

the sample but

Fail conditions: A severe loss in resolution of one or more pea~s might happen, due to loss of stationary liquid phase

through volatilization. These vapors are adsorbed on the molecular sieve along with CO2 , which leads to slow deterioration of that column. This will normally happen over a long period, with an order of magnitude of months.

The drying tube which is installed between sampler and f i r s t column will saturate also, this is indicated by a color change.

The columns must be carefully matched, to ensure that the retention times of the components allow separation of the CO2 from the Oz.

Strong points: Simultaneous analyse of five gases. Good sensiti-vity and accuracy. Simple apparatus and easy to

-12-operate. Commercially available thermal conductivity detectors, sampling valves, recorders and integrators can be adapted.

Weak points: Deterioration of molecular sieve and drying tube, which means regular replacement. Volatilization of the polar stationary liquid. A complete analyse takes; 8.5 min. About 30 min. are required for instrument stabili-zation.

Literature: [4.32 ] , [4.33 ] , [4.34

] , [4.3 e ].

4.3.1 Thermal conductivity detector. (Catharometer)

In a thermal conductivity detector the gas under investigation is passed by a filament. The filament cools down dependent upon the

th~rmal conductivity of the gas and thus in case of a mixture

dependent upon the composition of the gas. Because the resistance of the wire is dependent upon its temperature , the measurement can easily be performed when this filament is part of a bridge of Wheatstone. It has the advantage, when used for analysis of COz in air, of a linear direct reading output with constant sensiti-vity. COz can only be distinguished from anesthetic gases by comparing two samples of the same gas, from one of which the COz has been absorbed.

fig. 4.30 _{Schematic of a} catharometer.

4.3.2 Ionisation detectors.

Ionisation detectors can be distinguished detectors (FID), halogen and phosphorus electron capture detectors (ECD).

in flame detectors

ionisation (HPD) and

In FID, the gas from the stationary phase of the chromatograph is burned in an electric field. If the gas contains organic compo-nents ions will be produced by the burning process, and will cause a current between the electrodes. As carrier gas N_z is

- ,

3-used, which after passing through the stationary phase is mixed with H~ and Oz to get a flammable mixture.

The HPD operates similar phosphorus compounds.

to the FID but detects halogen arid

The ECD has one cylindrical shaped electrode and another rod shaped electrode in the central axis of the cylinder. The cylindrical electrode is a radio-active tritiumfoil. The carrier &as from the stationary phase flows through this cylinder and is ionized by the radiation from the tritiumfoil, which causes a certain base-current in the electrodes. If the carrier gas contains molecules or atoms with an affinity to electrons this current will decrease. The ECD is especially sensitive to halogen compounds.

4.4 THE LIGHT SPECTROMETER. [4.4'J

The principle of this multiple gas analyser is based on the measurement of the intensity of light emitted by excited atoms or

i~ns in a direct current glow discharge. When a gas is introduced into a low pressure chamber and in an electric field under the proper conditions of pressure, voltage and chamber geometry a glow discharge is formed. Selected spectrel bands exist near the cathode region of this discharge where the intensity of emission is proportional to the concentration of the component of interest in that gas sample. Monitoring the intensities of these bands provides a continuous measure of the component gas concentra-tions.

[

CATHODE DAIII: S'ACE NEGATIVE GlOw

'AIADAY I'OSIlIVE

,,'~·-i.~

ill

J'

'NO~

VOLTAGE

V

l

LUMINOUS aNUNSlTY AT 219ft",

fig. 4.40 _{Major luminous zones} (shaded) in a &low discharge with corresponding axial varia tions of voltage and luminous intensity.

Fig. 4.40 _{shows the major features of the &low discharge and the} corresponding variation of voltage and luminous intensity along the axis of the discharge tube. The discharge is sustained by a complex exchange of energy among atoms, molecules, ions and

-14-electrons. Most of the ionization (production of ions and electrons) in the discharge occurs in the negative glow and the cathode dar~ space. Some of the ions produced will accelerate toward the cathode. Although they are slowed down by collisions, many of the ions s t i l l have sufficient energy when colliding with the cathode to produce secondary electrons. The collision of electrons and photons, at the cathode can also generate elec-trons. These secondary electrons accelerate toward the anode, exchanging energy with the discharge in the process. They are the main source of energy required for ionization and excitation in the negative glow. Excited states are also created during the recombination of ions and electrons. The radiation seen in the negative glow and elsewhere in the discharge arises from the relaxation of molecules, atoms and ions from these excited states to ground state in one or more transitions. The spectrum of radiation from the discharge consists of a series of narrow emission lines. Each gas component in a gas mixture will have a characteristic emission spectrum that will contribute to the total emission spectrum of the gas mixture. If the pressure and the current through the discharge are held constant, then, in general, the intensity of anyone of these spectral lines is proportional to the partial pressure of the gas component corre-sponding to that line.

The luminous intensity is much greater in the negative glow discharge than in the positive column, whereas the air bac~grcund signal is the same in both regions. Hence, for purposes of gas analysis the negative glow region is only observed. See fig. 4.~o for CO z , the same can be found for Oz, NzO and He when they are miMed with Nz . A schematic diagram of the analyser is show~ in fig. 4.4b _• the of fig. 4.4b _Schematic light spectrometer. / OPTICAL INTEartatPllCE FaTER a.ICON D£T£C10R 1OVM:UJM QUAlm - __-.l1"""l _ _._1I1

Selectivity: The computed gas concentration from the detector output is transformed to an accurate gas concentration by an algorithm that corrects for a.o. background interference from other gases, including water vapor. This correction algorithm was restricted to include

-15-only the more clinically relevant conditions and did not yet include the effects of Ar and Ne, but these can be included also.

The effect of acetone and ethanol was evaluated by bubbling a known gas mixture through a mixture of either acetone or ethanol in water. 10 mg acetone in

100 ml water showed a

=

0.02 mol% increase in CO z . 0.3 g ethanol in 100 ml water showed a

=

0.1 mol% increase in CO z .

Sensitivity: 0.02 mol% for gases with range restricted to 0 - 10% otherwise i t was 0.05 mol%.

Accuracy: For CO z the maximum error error was ± 0.04 mol%.

was ± 0.15 mol% , the mean

Response time: 0 - T90~ is line. 215 ±15 msec. 100 ±10 msec. with a 25 m. for a 2.5 m. sampling line. sample

Lag time for all gases from point of sampling to point of detection was = 0.3 sec/m.

Calibration: Dark current and amplifier zero offsets are subtrac-ted from the detector signals by a microprocessor.

The calibration procedure is automated and uses nine lecture bottles of accurately prescribed mixtures of the five gases of interest (Oz, Nz , NzO, CO z and He) . The bottles are installed in the analyser and will last 1 year or more with 3 calibrations per day. The procedure takes about 12 min.

The calibrated signal is transformed to an accurate gas concentration by an algorithm that corrects for nonlinearities and subtracts background interference from other gases, including water vapor. The CO2 algorithm operates in the range of 0 -10%.

Reliability: The light spectrometer operates at a slight vacuum of 0.6 Torr and does not use a filament, both points make the analyser more reliable.

Likely cause for failure: When an unknown, unmeasured gas is introduced into the light spectrometer that computes the percentage or partial pressure of each gas by referencing i t to the total sum of all gases measured, all the readings will be in error.

Strong points: Patient application. Multi gas analysis. Conti-nuous. Suitable for rapid shallow breathing patterns

-1£,-(neonates) .

Wea~ point: The number of gases that can be analysed is limit8d to those for which a discrete spectral line can be found.

4.5 RAMAN SPECTROMETRY. [4.5']

The Raman effect relies on the interaction of monochromatic light with the vibrational/rotational modes of molecules to produce scattered light which is frequency shifted from that of the incident radiation by an amount corresponding to the vibrational-/rotational energies of the scattering molecules. Since these energies are species specific, an analysis of the frequency components present in the Raman scattered light provides an indirect chemical identification of the gases present in the scattering volume.

A sample is drawn into a sample cell and exposed to laser light. Raman scattered light is collected by a lense end projected onto a detector after passing filters which select particular frequen-cies corresponding to gases of interest. A microprocessor acquires the analog data and provides a data display. Fig. 4.5e _.

LASE> I

C>

rOCUSINGLEoS fIL TE> '.OlO"JL nPlIE> n.:BlO'" OT.EO SENSOR OEvICl SIGN" '>~CESS:NG "Jill!)~;rIER P!'lCiO-' COU-:E> fig. 4.50 _Schematic Raman spectroscope. of the

Selectivity: There is rarely any spectral overlap between the Raman spectra, provided that narrow (1 nm FWHM, although not stated by the autors, this seems to stand for Full Width at Half of Maximum) band pass f i l t e r s with high (>10 e dB) out of band rejection are used. The table

frequency cm-' (the lenght) .

with results shows that CO2 and NzO have one shift in common, stated with wavenumber 1285

wavenumber is the reciprocal of the wave-This seems to give problems in respiratory

and anesthetic

the authors.

-1;-gas monitoring, but is not mentioned by

Water vapor does not interfere.

Sensitivity: Better than 1 ppm (v/v).

Accuracy: The detected signal levels for the diatomic molecules,

Nz , Oz and CO are approximately 1020000 counts/ sec.

This magnitude is also valid for CO z . The noise level

is 20000 counts/sec. with no gas. Thus the signal to

noise ratio is approximately 50:1. The counting

s t a t i s t i c s error, associated with any count level

currently limits the precision to 0.05 volume percent.

Stability: The system is not affected with drift.

Calibration: The relative sensitivity to the different gases

remains absolutely fixed, eliminating the neec for

frequent calibration.

Strong points: Simultaneous detection of many different gases.

Laser power only

=

4 Watts. In future possible

replace-ment of the photo multiplier by photodiodes, thus low

costs. Breath by breath monitoring possible.

Weak points: Application of narrow

lenses could be expensive.

sampling volume precision.

band pass f i l t e r s and of

Accuracy dependent on

4.6 THE FLUERIC GAS ANALYSER. [4.6'J

The chief component of this gas sensor is a miniature flue~ic

oscillator of the jet-edge resonator type. This particular

flueric component consists of a jet-edge oscillator located

between two resonating cavities. Jet-edge oscillations occur when

a sharp edge solid body (wedge) is placed in a jet of gas that is

issuing from a nozzle, see fig. 4.68

• The jet will oscillate;

move back and forth laterally across the sharp edge. At a given

nozzle to wedge distance and jet velocity a uniqe frequency will

be observed, called edge tone frequency. In addition to the

jet-edge, the oscillator utilizes two resonating cavities, see fig.

4.6b

• Flow continues into one cavity until sufficient mass has

been accumulated with adequate energy to create a pressure

gradient large enough to cause the stream to reverse direction

into the opposite cavity. Depending upon shape of the cavities

and the pressure gradient inside, a second frequency, known as a

resonator cavity-characteristic frequency, is produced. This

second frequency is independent of the edge tone frequency, but

specific heats (Cv/Cp ) , weight and temperature.

the

-18-universal gas constant, moleculer

Exhaust Nozzle

To Transducer

- Inlet Nozzle

Sample In

fig. 4.60 _{Schematic of the jet} edge oscillator.

fig. 4.6b _{Schematic of the} flueric jet-edge resonator oscilator.

Oscillations are measured as oscillator cavity pressure pulses with a pressure transducer, and the signal is processed electro-nically. Thus for a given geometry, temperature and jet flow rate, the output in this sensor is a function of the physical properties (molecular weight, specific heats) of any gas flowing through the oscillator. A change in composition changes frequen-cy. Since there is not a unique frequency for a given gas component in a mixture, analysis can only be applied when one component is changing and the background gas is held constant. One method for analyzing a component in a changing multicomponent mixture is by using two flueric oscillators in parallel. The original gas mixture is drawn through one oscillator; through the other the same gas is drawn, but the interesting component is chemically or physically removed. The difference in frequency between the two oscillators will be caused by the removed component and will depend upon concentration difference. This me tho d i s use d i n [4. 6'] .

Sensitivity: The frequency for pure COz is approximately 25000 Hz Referenced against air the sensitivity is: 65Hz/%COz ,

-19-Accuracy: Measuring the percentage Coz against the frequency. For 0 - 10% Co_{z ,} frequency ranging from 27.48 to 27.12 kHz, SEM of frequency was ± 3.7 Hz (n 3). Lineair regression analysis yielded r

=

0.998 (P<0.001).

Stability: Over an 80 hours interval the drift ranged from -0.02% to +3% CO2 in air.

Time response: 0 - T~o_ is 450 ±10 msec. lenght of the sampling lines. This mechanical sampler, lower response with improved design.

Dependent upon the system included a times are possible

Calibration: The frequency is also a function of temperature. By using two oscillators (reference and sample) at the same temperature, temperature fluctuations will affect the two frequencies equally and the effect will be removed after subtraction. Hence the measurement is unaffected. In the same manner the affection by water vapor is solved.

The frequency is also sensitive to inlet jet pressure. A pressure compensator has to be used.

Likely cause for failure: The long drift of the system appears to be a function of temperature. During the determination of drift, temperature was not closely controlled. By control and feedback of room temperature drift might be within closer limits.

Strong points: Portable. Continuous analysis. Wea~ point: Single gas application.

4.7 PHOTOMETRIC ANALYSER. [4.7']

The method is based on highly specific color reactions. Coz is determinated by i t s color reaction with a solution of Fuchsin +

Hydrazine. The basic experimental equipment is that of the AutoAnalyser, [4.7Z ]. The basic components for continuous determination ore a proportioning pump, mixing coil, flow meter, photometer ond a recorder. Additionally an injection chamber and an integrator for discontinuous determination ore needed. See fig. 4.7a _{• In the case of continuous determination, a stream of} the gas to be analyzed is pumped into the mixing coil, in which i t reacts with the color-developing reagents. In the case of discontinuous determination the gaseous or liquid sample is injected into the injection chamber. This chamber contains 4 to 5 ml of a fluid suitable for the elution of the gas to be analysed. (0.01 N CH3 -CooH in the case of CO2 ) , The gas is eluted by a

-20-stream of nitrogen which is finely distributed into small bubbles after passage of a glass filter, thus producing a large surface. In the _{case of Co z (wavelenght} & 545 nm), solution 1 is hydrazine and solution 2 is fuchsin. Fuchsin is decolored almost completely by Hydrazine, the resulting faintly pink mixture reacts with Co z yielding a red product.

Selectivity: No cross reactivity between Oz and CO z .

Sensitivity: Continuous determination, in a gas flow of 0.5 ml/min. a _{concentration of 0.025% CO z can be detected.}

In a flow of 5.0 ml a concentration of 0.0025% CO z can be detected.

Accordingly, a sensitivity of 100 ppm (v/v) for CO z . Discontinuous determination, 0.2 ~l. for CO z . A sample volume of 5 to 10 ~l is sufficient .

...0'

P"ot-,.,

'KO'~' . ",teg,ato,

SOlutIOn 2

allt,gh, ~b'a...

--~_ - : . ja::w(

sy''''g·

fig. 4.?a Flow diagram photometric analyser.

of the

Stability: There is no drift in baseline.

Response time: Continuous determination, 0 - T,oo_ is : 2 min. for a stepwise concentration increase of _{CO z} of 0.5 volume

%.

T,oo_ - _{To _} is : 3 min. for a stepwise decrease of CO z of D.5 vol.

%.

Discontinuous determination, after injection of B ~l of CO z the peak is reached in : 0.2 min. After injection with 30 ~l of _{10 mM Na z CD}3 the peak is reached in : 0.5 min. The baseline is reobtained in : 3 min.

Calibration: Calibration gases were used, mixed from purified _{Nz ,} O2 and CO2 •

Atmospheric O2 eliminated from the recorder to

-21-and CO2 diffusing into the apparatus is the results of the analysis by setting zero before performing a measurement. Likely cause for failure: Especially in the case of CO z a precise

combination of the concentration of the components, i . e . fuchsin and hydrazine is required. If not correct there is no linear dependence between the amount of gas determined and the concentration of the gas in the sample.

Strong points: Discontinuous and continuous determination possible. Most components of the apparatus are usually present in a laboratory, so no big investment has to be made. Gaseous and liquid samples can be injected. With the same equipment at least three different gases (Oz, CO z and CO) can be detected.

Weak points: When different constituents is necessary to change the gas tions. Not portable and due probably fragile apparatus.

4.8 CONDUCTOMETRY.

have to be measured, i t specific reaction solu-to all the pipework a

The principle is very simple: measurement of the electrical conductivity in liquids. When analysing gases this means that the component of interest has to be dissolved in a liquid and the change in conductivity of the liquid must be monitored. Quite a number of gases can be detected in this way, by just matching the dissolving liquid.

A method of registration, arranged in such a way that the decline in the level of the curve of electrical conductivity is strictly proportional to the amount of CO z taken up, is shown in fig. 4.8°. A, and Az are spiral absorbers (fig. 4.8°). Az is provided with electrodes for conductivity measurements. P is a magneti-cally driven pump for circulating the air in the closed circuit of the apparatus. S, - S4 are gas pipettes containing the air samples to be analysed. C is a flask containing a solution for calibration purposes. M is a magnetic stirrer. By means of four way stop-cocks the spiral absorber A" the sampling pipettes and the calibration flask can be in- or excluded from the closed circuit. After starting the pump 1 ml strontium hydroxide (which is the absorber for CO z ) is injected through the wall of a rubber tubing into spiral A,. A, is included in the circuit and some liquid will flow with the gas through the helix, see fig. 4.8°. After absorption of CO2 initially present in the circuit in A" this absorber is shut off from the circuit. Now an exactly measured quantity of 1 or 2 ml of strontium hydroxide is injected

-22-into ebsorber A_{z .} The sensitivity of the recorder is edjusted to indicete full scele deflection. After e few minutes e nearly horizontel 2reph si2nifies thet the apperetus is eirtight. Now an eir semplin2 pipette cen be included into the circuit for enelyzin2 or flesk C for celibretion purposes. [4.8'].

Accurecy: Quentities of COz up to 8 ~mol cen be determined with en error less then 0.02 ~mol.

Accordin2ly quentities of 2 to 8 ~mol conteined in semples of et leest 150 ml of fresh etmospheric eir or 1 ml of expired eir cen be enelysed with en eccuracy better then ± 1.0% of their COz content.

Stability: There is a certain background error, indiceted by the slope of the curve in the time between enalyses. In a 10 min. period this error corresponds to about 0.01

~mol COz .

Response time: 0 - T,ccM is about 10 min. for a sample volume not exceeding 20 mI. It tokes about 5 min. to remove the initial amount of CO_z from the circuit. It takes another 5 min. to check whether the apparatus is airtight. £ a

1

p 8 Sc,..

fig. 4.8D _{Flow diagram of the}

conductometric analyser.

Colibration: A quantity flosk C. When recorded curve CO 2 token up by

-23-of NaHC03 solution can be correctly adjusted, the will be proportional to the hydroxide.

injected into drop of the the amount of

Likely ceuse for feilure: The concentretion of the Sr(OH)2 reegent kept in e bottle unevoidebly decreeses with time, beceuse of conteminetion of etmospheric CO 2 . The 5 mmol/l Sr(oH)z solution used for the CO 2 ebsorption wes kept in e bottle with gless stopper, i t wes opened ebout 20 - 30 times e week for removal of semples for the enolyses, end ebout 1% of the Sr(OH)z was generolly lost during one week.

Strong point: A simple detector.

Weak points: A fragile set up. It needs some skill to inject the regent. Not continuous.

4.9 PARAMAGNETIC CO 2 ANALYSER.

The behaviour of gases in a magnetic field can be used as a means for analysis. Especially for the detection of oxygen i t is a very much used method. Matter having a permanent magnetic moment is called paramagnetic. The paramagnetic property of oxygen is much stronger than other gases and thus provides a good basis for detection of Oz.

The paramagnetic Co z enalyser [4.9'J depends upon observing the change of oxygen concentration in a gas mixture before and after ebsorbing the carbon dioxide component. Fig. 4.90 _{shows the} method. A gas sample is directed through either one of two glass

'RIOX't'CEN ___ ANAL\"SER EUCT1lOM AGSETIC ~-w.n\'ALVE CALCIUM CHLORIDE SODA-LIME

fig. 4.9- Method of direction of the gos semple through one orother of two absorption tubes before entering the paramagnetic analyser.

-24-absorption tubes before entering the paramagnetic analyser. One

tube (A) contains successively soda-lime and calcium chloride,

and the other tube (8) calcium chloride only. Thus after passing

absorber (A) the gas is free of carbon dioxide and is dried, and

after passing (8) the gas is only dried leaving the carbon

dioxide unaffected. The gas flow is switched from one tube to the

other at 1 min. intervals by means of an electromagnetic valve.

From the electromagnetic valve the gas flows to a paramagnetic

analyser. the electrical output of which is recorded by a digital

voltmeter and a chart recorder. The used oxygen analyser had

three ranges of sensitivity giving a full scale deflection for

either 5. 25 or 100 % oxygen.

80th oxygen and carbon dioxide can be measured:

Fcoz CO z fraction in the gas sample.

eFoz

=

Oz fraction in the gas sample before removing CO z .

AFo z = Oz fraction in the gas sample after removing CO z .

Then AFo z = eFOz / ( 1 - Fcoz ),

so that Fcoz = ( AFo z - eFOz ) / _{AFo z}

Selectivity: Interference of the paramagnetic

diamagnetic susceptibility of other

nitrous oxide, helium or argon) is

allowed for. or can be eliminated

instrument to electrical zero output

flowing in the correct concentration.

analyser by the

gases, such as

small and can be

by setting the

with these gases

Sensitivity: The analyser and recording

change of 0.0005% oxygen.

system can detect a

Accuracy: Results recorded had a maximum coefficient of variation

of 0.55% with a digital voltmeter.

Y

=

0.996X - 0.001. ( r

=

0.9999 ± 0.04 SEM of estimate; (P<0.001) with a digital voltmeter. X is the Haldane

number, Y is the paramagnetic number.

Response time: A complete analysis can be made

vals.

at 2 min.

inter-Calibration: The sample gas flow was standardized at 60 ml/min.

and the output was set to zero output with the carrier

gas flowing. The span was adjusted on the 25% range

using room air (20.95% oxygen).

Likely cause for failure:

while.? g. of

containing 8

%

which an abrupt

The soda-lime will be saturated after a

soda-lime absorbs _{CO z} in a gas sample

CO z for approximately 4 hrs .• after

decline takes place.

-25-analysis of oz and Coz. Wea~ points: Not continuous,

exhaustion.

4.10 THE Pcoz ELECTRODE.

replacement of soda-lime after

Measurement of CO2 concentration with an Severinahause electrode is an adaptation of a pH measurement. Diffusion of Coz in a hydroaen-carbon solution results in the formation of carbonic acid, aivina a chanae in pH which can be detected with a alass electrode (pH- electrode). The reaction room is separated from the sample by a membrane so that only aas molecules are able to diffuse into the hydroaen-carbon. The electrode is temperature dependent. With decreasina temperature, the displayed Coz concen-tration increases.

In [4.10'] the intraelectrode pH chanae is proportional to the change in COz concentration. The resultant sianals from the electrode are fed into a recording pH electrometer. The entire housing containing the electrode is mounted horizontally in a small water bath connected to a high quality circulating thermo-stat which serves to maintain the assembly at a constant tempera-ture of 37.0 °C. To allow temperature equilibration of the gas entering the electrode chamber. the inlet tubing is coiled inside

the bath. See fig. 4.10"'.

Selectivity: The COz electrode is totally specific for CO~ and unaffected by anesthetics [03 ] .

Sensitivity: Changes of 0.5% CO_z can be detected.

G

H

fia. 4.10'" Schematic of the electrode assembly.

A electrode, 8 electrode housina, C reaction chamber,

o

waterbath, E sample in, F sample out, G water in, H water out.

-26-Accuracy: Maximum accuracy with this apparatus is within the physiological range of 3 5% CO2 concentration. Accuracy and stability are within ± 3% of full scale. Response time: 30 sec. (no specification given about delay time

nor about the intercept between zero and final value where this time was measured).

Calibration: At the beginning of each new experiment a calibra-tion should be performed by injecting a series of gasmixtures with known CO z concentration.

Likely cause for failure: Each injection has to be followed with introduction of air to allow return to baseline.

Strong points: The pH electrometer and Pcoz electrode are commercially available. Rapid and easy construction. Weak points: Maintenance, changing electrode membrane

electrolyte solution, dependent upon use every weeks. Single gas capability.

and 2 - 4

Alternative: Beckman Cl/C02 analyser [4. 102J Another alternative is the determination of CO z with optical sensors, called optodes, [4. 103

J.

_{The measuring principle is}

also based on pH changes within a bicarbonate indicator buffer solution.

4.11 THE EINSTEIN CO2 DETECTOR. [4.11'J

The technique is based on the fact that expired gas contains between 4% and 6% CO2 , whereas swallowed atmospheric air or gas from a properly functioning anaesthetic circuit contains neglegi-ble amounts of CO2 . The device is based on the chemical a t t r i -butes of Cresol-red and Phenophthalein. Both of these indicators change colors in the presence of an increased concentration of hydrogen ions, resulting from carbonic acid.

Selectivity: No color change due to Oz, NzO and the commonly used volatile anaesthetics.

Sensitivity: A CO z content of less than 0.3% does not color change.

Response time: 3 - 5 sec. Calibration: None.

Strong points: Reliable, rapid, cheap.

-27-Wea~ points: Detects only one gas. prior preparation.

Fragile and is dependent on

Alternative: TriMed model 510, electronic, lightweight and small CO2 detector [4.112 J. Another alternative is [4.1F'J,

in which alveolar gas is dissolved in a transparent liquid. After reaction with the CO2 the liquid gets more or less opaque, which is a measure for CO2

concentration.

4.12 REFRACTIVE INDEX MEASUREMENT.

The refractive index of a gas is the ratio of the velocity of light in vacuum to the velocity of light in the gas. All gases and vapours have a specific refractive index. The velocity of light in a gas will depend upon the molecules present, and thus upon its density and therefore varies with pressure and tempera-ture. Absolute measurement of the velocity of light in a medium is difficult, but i t is quite easy to compare the velocity of light in one medium with the velocity in another. A suitable instrument for this purpose is the Rayleigh interference Refrac-tome t e r, [4. 12'

J .

This article from Edmonson, dated from 1957, is the only article I could find about this subject. Probably this method is obso-lete, and is therefore only addressed briefly.

-28-5 DISCUSSION.

In the preceding chapter twelve different methods of carbon dioxide detection are described. It is quite obvious that methods, in which the carbon dioxide has to react with matter,

have a slow response time and don't seem to be useful in breath by breath analysis. The methods ment ~re; chromatography, photometry, conductometry, the Pcoz electrode, the paramagnetic

~n~lyser ~nd the Einstein analyser. Since they are usu~lly cheap

and simple to handle, they do have their own application.

The rest of the methods are applicable for breath by breath analysis. The infra-red absorber and mass spectroscope have been in clinically use now for many years, and have set some kind of standard. New methods that are competitive with the infra-red and mass analysers are the light spectroscope and the Raman spectro-scope. Whether they will set a new standard depends upon their size/weight and price, criteria that could not possibly be included in this report. The flueric gas analyser has also breath by breath capability, but has only one gas applicability and cannot yet compete in response time with the spectroscopes.

This report only reviews methods found in medical literature. A subsequent study should be performed on methods of CO2 analysis found from other fields of literature (the output of a computer search in Chemical abstracts and in Inspec is already enclosed in this report), to be continued with an investigation of possible application in respiratory gas analysis.

The values given under, e.g. accuracy might be too optimistic, since most experimenters don't like to publish bad results. When working with ~ system one might never again reach the same results. Interesting from this point of view is the article [5'J in which is stated that; although instrumentation has improved there is a deterioration in analytical standards in lung function laboratories.

-29-6 REFERENCES.

[ 0' ] Smidt,U., Niedine,G.von,

del' Atemgasmessung. Biomed.

Lolleen,H., Methodische Probleme

Techn. 1976;21(4): pp. 102-114.

Eberhart,R.C., Wiegelt,J.A., Respiratory monitoring: current

techniques and some new developements. B. Eur. PF1ys. 1985;

21 ( 3): pp. 295-300.

Severinghaus\J.W., Methods of measurement of blood and gas

carbon dioxlde during anesthesia. Anesthesiol. 1960;21(0):

pp. 717-726. Automatie (Dutch) 1983; 27( 5) : [04 ] Industriele gasanalyse (1). pp. 176-180. [1'] "Metingen in de geneeskunde 2".

University of Technology (Dutch)

chapter 1, pp. 1.2-1.64.

Syllabus of the Eindhoven

nr.5.615.1 issue 1986,

[12 _{] Smalhout,B., Capnografie, Dissertation (Dutch), issue 196'i'.}

Publisher A. Dosthoe~'s uitg. mij. N.V. Utrecht.

Introduc-tion and Chapters 1 to 4, pp. 3-24.

analysis. J. Clin.

gas with

Ca~~inS~J.M.( Reduced ris~

Monlt. 1~86;2 2): pp. 77-78.

Smidt U., Nieding,G.von, Lollgen,H.~ Methodische Probleme

del' Atemgasmessung. Biomed. Techn. 19/6;21(4): pp. 103-106.

[ 2' ] "Elektl'ische University of chaptel' 5, pp. meettechniek". Syllabus of Technology (Dutch) nr. 5.574 5.1-5.6. the Eindr-,oven issue 19BL<, [23

] "Electronica 2, digitale gedeelte". Syllabus of the

Eindhoven University of Technology (Dutch) issue 1984,

chapter 2, pp. 48-49. Automatie (Dutch) 1983;2'7(5) of Haldane's gas-analysis 1958;143: pp. 5P-6P.

.

,...,

,

I U ) 1983;2'?[ ( Dutch) Automatie [ 24 ] Industriele gasanalyse (5) pp. 367- 368. [3'] Industriele gasanalyse (1) pp. 1'76. [4.0'] Lloyd,B.B., A development

apparatus. J. Physiol. (London)

[4.02 ] Camp~ell~E.J.M., Simpl~fic~tion of.Haldane's ap~arat~s.for

measurlng ~02 concentratlon In resplred gases ln cllnlcal

practice. Br. Med. J. 1960; 1: pp. 457-45B.

[4.03

] Cormack,R.S., Eliminating two sources of error in the

Lloyd-Haldane apparatus. Respir. Physiol. 1972; 14: pp.

382-390. [4.04

] Owen-Thomas,J.B., Meade1F., The estimation of carbon

dioxide concentration in the presence of nitrous oxide,

using a Lloyd-Haldane apparatus. Br. J. Anaesth. 1975;4'7:

pp. 22-24.

[4.1'] Olsson,S.G., Fletcher,R., Jonson,B., e t . a l . , Clinical

studies of gas exchange during ventilatory support - a

method using the Siemens-Elema CO a analyser. Br.J.Anaesth.

1980;52(5): pp. 491-499.

[4.1 2

] Blackburn,J.P., Williams,T.R., Evaluation of the Datex

CD-101 and Godart capnograph mark 2 infra-red carbon dioxide

-30-[4.13 ] Cormac",R.S., Powell,J.N., Improving the performance of

the infra-red carbon dioxide meter. Br. J. Anaesth. 1972;44:

pp.131-141 [4.14

] Carbon dioxide monitors. Health Devices 1986;5ept.-Oct.:

pp. 255-285.

[4.1 e ] Long,C.L., Carlo,M.A., Schaffel,N., e t . a l . , A continuous

analyzer for monitorin~ respiratory gases and expired

radioactivity in clin~cal studies. Metab. Clin. Exp. 1979;

28( 4): pp. 320-332.

[4.1 e ] Kalenda,Z., Equipment for capnography. Br. J. Clin. Equip.

1980;5(5): pp. 180-193.

[4.17 _{] Ammann,E.C.B'}

L Galvin,R.D. I Problems associated with the

determination o~ carbon diox~de by infra-red absorption. J.

Appl. Physiol. 1968;25( 3): pp. 333-335.

[4.1 e ] Severinghaus,J.W., LarsonIC.P., Eger,E.I., Correction

factors for infra-red carbon d~oxide pressure broadening by

nitrogen, nitrous oxide and cyclopropane. Anesthesiol. 1961;

22( 3): pp. 429-43~'.

of nitrous oxide on an infra-red

Anaesth. Intens. Care 1978;6(2):

The effect

analyser.

[4.1 9_{] Russell,W.J.,}

cal"bon dioxide

pp. 167-168.

[4.1'°] Ramwell,P.W., The infrared analysis of carbon dioxide

during anaesthesia. Brit. J. Anaesth. 1957'29: pp. 156-159.

Walt,W.H.van der., A

methods for the

determi-Int. Z. Angew. Physiol.

[ 4 . 1 " ] Benade,A.J.S., Strydom,N.B'

i

comparison of physical and chemica

nation of respiratory quotient.

1970;28: pp. 193-196.

[4.2'J Sodal,I.E., Swanson,G.D., Micco,A.J., e t . a l . , A

compute-rized mass spectrometer and flowmeter s~stem for respiratory

gas measurements. Ann. Biomed. Eng. 198:3; 11: pp. 83-99.

Modified ion-source cage for

Med. BioI. Eng. Compo 1981;

[4.22

] Beatty,P., Greer,W., l<.ay,B.,

a clinical mass spectrometer.

1 9( 6}: pp. 770-774.

[4.23

] Gl"avenstein,N., Theisen,G.J., Knudsen,A.K'

i

mass spectrometer reading caused by an aero so

Anesthesiol. 1985;62( 1): pp. 70-72.

Misleading propellarit.

[4.2Q

] McCleary,U., Potential effects of an un"nown gas on mass

spectrometer readings. Anesthesiol. 1985;63(6): pp. 724-725.

[4.2 e ] Gillbe,C.E., Henegan,C.P.H., Branthwaite,M.A., Respiratory

mass spectrometry during general anaesthesia. Br. J.

Anaesth. 1981;53: pp. 103-l09.

[4.2 e ] Stout,R.L' _L Jewell,E.D., Wessel,H.U., Paul,M.H' _L Mass

spectrometer ~or multiple respiratory gas analysis. ~iomed.

Tech. 1975;20(5): pp. 165-171.

[4.27

] Hallback,I./ Karlsson,E., Ekblom,B'

1 Comparison between

mass spectrometry and Haldane technique ln analysin~ O2 and

CO2 concentrations in air gas mixtures. Scand. J. Cl~n. Lab.

Invest. 1978;38( 3): pp. 285-288.

[4.2 e ] Schoeller,D.A., Klein,P.D., A microprocessor controlled

mass spectrometer for the fully automated purification and

isotopic analysis of breath carbon dioxide. Biomed. Mass.

monitor- spectro-

-31-[4.29 _{J Ozanne,G.M., YOLJng,W.G., Mazzei,W.J., Sever·inghaLJs,J.W.,}

MUlt~patie~t ane~t~etic mass spectrometry. Anesthesiol.

1981 ,55( 1). pp. 6.:::-/0.

[4.2'OJ Fo~ler,K.T., The respirator~ mass spectrometer. Phys. Med. B101. 1969;14(2): pp. 185-199.

[4.2"J Turney,S,Z'l Computerized multibed respirat~ry

ing. Camp. Cr t . Care Pulm. Med. 1983;3: pp. 9-25. [4.2'I2J Smidt,U., Niedin£.1..G.V., Today's respiratory mass

meter. Pneumono1. 19/5; 151: pp. 241-295.

[4.2'3J Henderson,A.M., Mosse,C.A., Forrester,P.C., e t . a l . , A system for the continuous measurement of oxygen uptake and carbon dioxide output in a r t i f i c i a l l y ventilated patients. Br. J. Anaesth. 1983;55(8): pp. ?91-800.

[4.3'J Tentative method for gas chromato~raphic analysis of O2 , N2.l CO, CO 2 , and CH4 • Health. Lab. SC1. 19?5;12(2): pp.

1?3-1?b.

[4.3 2 ] Johansson,U.B., Determinationof respiratory gases (C02 , 02t Ar ~nd N2 ) _With g~s so~id chromatography. Scand. J. Clln. Leb. Inve:::>t. 1983,43( 1). pp. 91-94.

[4.3~J Jay,B.E., Wilson,R.H., Adaption of the gas adsorption

chromatography for use in respiratory physiology. J. Appl. Physiol. 1960;15: pp. 298-302.

[4.34 _{J Barnikol,W.K.R., Dohring,W., Ein unabhagiges verfahren zur} messLJng der zusammensetzung von gasgemischen mit hilfe der

~aschromatographie. F'flLig. Arch. ELJr. J. Physiol.19'7'1';

369( 1): pp. 91-94.

[4.3~] MLJysers,K. Siehoff,F., Worth G., BILJt- und

Atemgasanaly-sen mit Hiife del' Gaschromatographie. Klin. WSCfil-. 1961; 39(2): pp. 83-8?

[4.3~J Dumas,T., Determination of carbon dioxide at the ppm

level: a statistical comparison of a single-filament and a four-filament thermal conductivity detector. J. Chromatogr.

1984; 299( 2): pp. 432-435. [4.4'J Fraser,R.B.,

analyses: lii:ht pp.1001-100i.

Turney,S.Z. ,

spectrometer. New methodJ. App. Physiol.of respiratory1985;59(..:»:~as [4.5'J VanWagenen,R., Westenskow,D.R., Benner,R., Respiratory gas analysis by Raman scattering. Suppl. Anesthesiol. Equip. Monit. Eng. Techn.1 1985;63(3A): pp. A163.

[4.6'] Calk.ins,J.M., Waterson,C.K., Saunders,R.J., Samoy,V.J., A flueric respiratory and ane~thetic gas analyser. Ann. Biomed. Eng. 1982; 1 O( 2): pp. 8:';-96.

[4.?'] Lang,W., Wolf,H.U., Zander/R., A sensitive continuous and discontinuous photometric determination of oxygen, carbon dioxide and carbon monoxide in gases and fluids. Anal. Biochem. 19?9; 92( 2): pp. 255-264.

[4.?2] WolftH.U., Zander,R., LangtW., An automated continuous determlnation of oxygen with R1gh sensitivity. Anal Biochem. 19?6;?4: pp. 585-591.

[4.8'] Holm-Jensen,I., Conductometric determination of CO 2 in air samples. Bcand. J. Clin. Lab. Invest. 1976;36(5): pp. 493-495.

-32-(4.9'J Meade,F., Owen-Thomas,J.B., A paramagnetic metrlod for

measuring the carbon dioxide concentration oT a gas. Br. J.

Anaesth. 1975;47( 10): pp. 1037-1042.

[4.10') StrahlendorT,H.K., StrahlendorT,J.C., Johnson,G., An

easily assembled and inexpensive apparatus Tor continuous

monitorin2 oT expired carbon dioxide concentrations in

artiTicially ventilated animals. Pharmacol. Biochem. Behav.

1979;10(4): pp. 619-621.

[4.102

_J

_{Hicks,J.M., Aldrich,F.T., JoseTsohn,M., An evaluation of}

the Beckman chloride/carbon dioxide analyzes. Clin. Chern.

1976; 22( 1 1): pp. 1868-1871.

[4.103

] Opitz,N., LUbbers,D.W., Compact CO2 2as analyzer witr.

Tavourable signal-to-noise ratio and resolution using

spe~ial Tluorescence sens~rs ~optodes) illuminated by blue

LED s. Adv. Exp. Med. 1984,180. pp. 757-762.

[4.11'] Berman,J.A., Furgiuele,J.J., Marx,G.F., The Einstein

carbon dioxide detector. Anesthesiol. 1984;60(6): pp.

613-614.

[4.112 _{) Bashein,G., Cheney,F.W., Carbon dioxide to veriTy}

intratracheal placement oT a breathing tube. Anesthesiol.

1984;61(6): pp. 782-783.

[4.113 _{) Smith/R.H., Volpitto,P.P., Simple method oT determining}

CO2 conLent OT alveolar air. Anesthesiol. 1959;20(5): pp.

702-703.

[4.12') Edmondson,W' _l Gas analysis by reTractive index

measure-ment. Brit. J. Anaestrl. 1957;29: pp. 570-574.

[5'J Chinn,D.J., Naruse,Y., Cotes,J.E., Accuracy oT gas analysis

-33-APPENDIX

A.

THE INPUT OF THE COMPUTER SEARCH IN THE DIFFERENT DATABASES.

The combination of words, ~iven

in the titel, the keywords and in

?-sien at the end of a word

extended with additional letters.

Database: Inspec.

Input:

under "Input:" was searched for

the abstract of articles. The

means that this word mieht be

.-

and

..

,

anal? concentration?

or Carbon dioxide deter'minat? _{e as} partial pressure?

~

CO2 measur? gases Traction?

capnograph?

In the period 1977+ we Tound

In the period 1983+ we Tound

tr,is report.

Database: Chemical abstracts.

Input:

434 articles.

142 articles, these are enclosed in

4 - and

--

I

t

anal? concentraticn?

!

or RN = _{124-38-9 determinat? gas? partial pressure?}

..

mea sur? fraction?

capnoli;raph?

I

RN = 124-38-9 re~resents all the synonyms for carbon dioxide.

In the period 197/ to 1982 : 22 art1cles.

In the period 1982 to 1986 : 27 articles. The articles of both

periods are enclosed in this report.

Database: Medline.

Input:

4 - and

-..

t

CO2 anal? concentration?

or car,bon dioxide detel~minat? li;as? partial pressure?

RN = _{124-38-9 measur?} fraction?

~

capnograph?

In the period 1980· : 930 articles were in consideration, this

was too much to have them enclosed in this report.

Database:

EM

Base.

Input:

.-

and

-t

CO2 anal? fraction

Ol~ car,bon dioxide _{determinat? respirat? 2 as ?} concentration

+

measur?

In the period 1976+ :

=

1300 articles were in consideration, this