Applications of gas chromatography to anaesthesiology 1. Gases and vapours

13 Pages·2008·0.88 MB·English

APPLICATIONS OF GAS CHROMATOGRAPHY TO ANAESTHESIOLOG~
.1 GASES AND VAPOURS*
DRAWOH L. ,REOrtAZ ,.nL,~ ,.O.HV fAsn SIUOL R. O~:Ix, M.n. I
EHT EVITArrTN_aUQ NOITANII.~ltIETED of anaesthelic vapours contained in a gas
mixture has, until recently, usually been accomplished by chemical analysis. The
classical oroeedures, of which the Orsat and it~ modifications 1 are oerhaos the
most representative, consist of metering a pefinite volume of a gas mixture into
a burette, bringing the gas into contact wi~'h a reagent which will react with rO
absorb the particular constituent being determ{ned, and finally measuring ~e
change in volume resulting from removal of the cqnstituent vapour. This procedur et
is then repeated with different reagents for successive removal of each of the com-
ponents of the mixture until the entire compositidn is determined. Analysis by this
techmque (cid:12)9 s~. quite obviously slow and tediouo. ~ In addition, several common I
errors are possible. These include those due to ransfer of sar0ples, gas leaks in
the apparatus, and incomplete removal of the egnstituent of the mixture by the
appropriate reagent. "Alternatively, one may en)ploy a "eon,qersion technique."
This involves conversion of the particular consfitttent to a non-volatile by-product.
which is then determined quantitatively by some physical or chemical method.
The conversion technique most commonly employed for analysis of the ana~s:/
thetie agents utilized the iodine pentoxide train In its original or modified forn?.-
Here a sample containing a gas or vapour is d~awn ?ough a tube filled with
asbestos, which in turn is impregnated with eni~co~ p~mtoxide. The organic su b-
stances are oxidized to carbon dioxide and ware.i-by tae iodine pentoxide. In the
process, free iodine or hydrogen iodide or both are liberated. These substances
are then quantitatively titrated with a sodium b~sulphite or thiosulphate sol~-
tion. The disadvantages of this method eombi#e those of the Orsat technique
with the disadvantages of organic analysis and finally of quantitative titration of
the by-products.
Because of the obvious disadvantages of the chemical methods of analys~s.
many investigators have, in recent years, turned o~ physical means of assay. Th(tse
methods measure some physical property of a ga~ or mixture of gases and thereby
give information as to the quantitative composition. Physical methods may be
divided into two groups: specific and non-specific. The non-specific methods,
which include analysis of density, ~" thermal conduetivitw, 4 refractive index, 5 the
velocity of sound through gases, 6 and paramagnetism measure a property which
is possessed to varying degrees by all gases. The specific methods, including
absorption of radiant energy in the infra-red s an d invisible 9 regions of the spectra,
morF*l~6 the Department of ,ygoloisehtsenA Albert E~stein College of Medicine, weN kr6Y
, N.Y. Supported inpart by Contract .oN AD 49-007-MD ,26Q Research and Development
,dnammoC Ottlee of the Surgeon General, United States Alany, and United State~ Public
Health Service Grant .oN .2672
822
Can. Anaes. Soe. ,.J vol. ,01 .on ,3 May, 3,691
ZAUDER & OILKIN: APPLICATIONS OF GA SY I-IPk~u~GOTK2~v/,~fRI-IC 229
gas chromatography, ~1 and polarography, n measure a prope~-ty possessed only
by the gas being analysed. The relative advantages and disiadvantages of the
physical methods have recently been reviewed by Maple~on?2 /
Anaesthesl01oglsts, and others concerned with quantitation )f ~ag and vapour
mixtures, have long sought an analytical technique that huoWf pem~t rat~id and
accurate quantitative analysis of these compounds. Ideally, th, technique should
not only be quantitative, but should permit.qualitative sepan noiL of the agents
from each other as well as from the fixed respiratory gases. It wohld be advan-
tageous if the method was. universal, i.e., applicable for all vapours and gases
that are or may be employed in anaesthesia, and ff it could e iminate costly and
time-eonsltrning physical and chemical methods of analysis.
Gas chromatography appears to satisfy all the~e criteria. This technique, intro-
duced approximately a decade ago, has found wide industri~d a~pplieation. 31 Its
potential as a tool in medical research has been recognized only t ,ylta~ec'er 4a The
first mention of its use in anaesthesiology was by Fabian ,na 1 Games in 19607 o
It has since been used by HalP s for qualitative separation of eh_- halothane-ether
azeotrope, and by Glover and Hodgson s~ for determination of the purity of a
series of fluorinated hydrocarbons of anaesthetic interest. Several other investi-
gators have used this technique as a tool in the quanttation of inspired
atmospheres of anaesthetic gasesJ z There is some indication that gas chromato-
graphy will be applicable to ~e determination of anaesthetic levels: in blood and
other body fluids) * ':' '"-" The basic principles of gas ehromatogr ~phy were reviewed
by Summers and Adriani. ~-
In adapting the technique to the quantitative analysis of anaesthetic drugs,
we found several unanswered questions. How accurate is the method? How
reproducible are the results? What temperatures and carrier gas flows should be
employed for a particular anaesthetic? Which column should be erfiployed for
,}na particular agent? In this and the subsequent reports, we shall attempt to
answer these questions in the hope that others wall (cid:12)9 be saved i considerable (cid:12)9 time
ni putting this technique into routine use.
~ELPICN1RP FO NOITATNEINU/ITSNI
Basmally, gas chromatography si an adsorption technique utilizing a physical
method of separation in which the components to be separated are distributed
between two phases. One phase consists of~ stationary bed of large surface area
while the other phase is a gas or vapour that percolates through or along the
stationary bed. As with other types of chromatography, resolution of a mixture
si based on the affinity of the compounds for the stationary phas,e. == Two types
of stationary beds were employed. A solid phase in the form of a molecular sieve
was used for quantitation of rare gases of anaesthetic interest: the commonly
employed anaesthetic agents were separated on a stationary phase consisting of
an organic liquid of high boiling point coated on inert granules (diatomaceous
earth). The gases or vapours being quantitated are swept into the column by
the streaming of an inert carrier gas. Those gases which are strongly attracted
230 NAIDANAC '~TSITEHTSEANA YTEICOS LANBU)(J
to the stationary phase move through the cohm~.n at a slow rate and appear at
the end of the column later than the componenys' which have demonstrated an
indifference towards the stationary bed. Substanqes are thus concentrated by the
column and eluded in a pure form. The ~rate ~f adsorption alad elution, e.g.,
retention time, is dependent upon the tempemthre of the column, flow-rate of
the carrier gas through the column, and finally upon the nature of the packing
or active material of the column.
When a surge of gas appears at the end of la column, it may be detected in
several ways. Detection by thermal conductivity ~vas employed in the analysis of
the gases and vapours of anaesthetic interest. Detectors of this type (thermistors)
are constructed of ceramic materials of a high thermal coefficient. Small changes
in the composition of the gas surrounding the de~ectors alter the thermal conduc-
tivity of the mixture. Small changes in thermal ~onductivity markedly' alter the
electrical resistance of the thermistor. This alteration in electrical resistance is
measured by upsetting the balance on a Wheat~tone bridge which incorporates
the reference and sensing detector in opposing limbs. The unbalance is amplified
and recorded on a potentiometrie recorder. As Bach component of a mixture is
detected a characteristic curve or chromatogramJis produced. )Eluded substances
can be readily identified by comparison with a lqaowh or predetermined chroma-
togram. In addition,-the area of the inscribed c~rve is related to the concentra-
tion of the substance passing through the detector chamber. The apparatus is
depicted schematically in Figure .1 It should be noted that in addition to a pres-
Fmve~ .1 Block diagram of the chromatograph. Provision is made for regulation of cartier
gas flow, maintenance of constant temperature of the column and detector block, and constant
volume sampling.
sure regulator, pressure gauge, and flow-meter to regulate carrier gas supply, the
hamber is thermostatically controlled so that the temperature of the column a~d
etectors is maintained within 1.0--+ C. The apparatus is also provided withl a
constant-volume sampling valve for gases and ;~;ith an injector block for liquid
samples. Helium is most commonly employed as the carrier gas.
REDU.A.Z & :NIK_BO APPUCATIONS OF SAG CHBOMATOG~'kPFIY 231
SUT~AVVA
There are several highly satisfactory gas chromatographs available commer-
cially. Alternatively, one may, if he is particularly ambitiot~s or mechanically
gifted, fabricate his own apparatus. 30- The instrument used inbur laborat6ry was
a Perkin-t~!mer 154-C Vapor Fractometer. The bridge c~cuitacross the sensing
and reference thermistors was coupled to a Leeds-Northrup Type G Speedomax
Strip Chart Recorder. The five-millivolt full-scale respofise time ob this recorder
was one second. Signal output from the bridge could be attenhated by factors of
two down to 1/512 of maximum sensitivity. This permits adjustment of peak-
height signals from the detector block. A valley sensor and Gansmitting poten-
tiometer of a Perkin-Elmer Model 194 Printing. Integrltor v~as aft:ached to the
recorder pen drive shaft. The areas of the peaks produfced b~ the eluded gases
as they flowed through the thermal conductivity cell were thus aut"omatically
integrated and recorded. The current across the bridge circui I wals kept constant
at 8 volts.
OPERATING CONDITIONS
The rapidity with which a sample of gas can traverse the e+lumn is dependent
upon three factors: the flow-rate of the carrier gas through thel column, the nature
of the column itself, and finally the temperature of the columla. In si_~t investiga-
tion, each of these factors was varied independently until the optimum combina-
tion was determined. The effect of increasing flow-rate of earlier gas upon reten-
tion time using a ~ in., 2-metre DC 200 silicone oil (dimethy I siloxane polymer)
column and holding the temperature constant is shown in Figure .2 Obviously
retention time can be reduced by increasing the rate of flow of carrier gas. A flow
DIETHYL ETHER 9.9%
2 METER SILICONE OIL COLUMN
45~ X-16
90 cc Imin 45 cc Imin 22 cc Imin
S ~ 2 mm '
FICHE .2 Decrease in retention time of diethyl ether whcn flow-rate of
carrier sag si .desaercni The initial peak stneserper room air.
232 CANADIAN ANAESTHETISTS t YT;EICOS JOURNAL
HALOTHANE 5.5%
HELIUM FLOW 90~r
C"57 X 61-
C B A
f
(- I
2 rain
! __
~ut~mF .3 Effect fo column packing retention time. A = di-isodecyl-
phthalate, B := di-2-ethylhexyl sebacate, C = dimethyl silo'xane polymer (DC
002 silicone off).
of 90 c.e./min, was found to be at the optimum for anaesthetic agents. At higher
flow-rates there is incomplete separation of the components. Fusion of the peaks
made rapid identification and quantitation difficult. At lower flow-rates read-out
time was needlessly prolonged. The effect fO varying the column packing with all
other factors being held constant is shown in Figure .3 Retention time is shortest
with a silicone oil (C) column and longest with a di-2-ethylhexyl sebacate
column (B). The di-isodecylphthalate column z. (A) occupies an intermediate
position. In each instance the eomnn was ~} in. in diameter and 2 metres in
length. When a 1-metre column was employed, separation was often incomplete.
The effect of temperature upon retention time with the three differently packed
columns is shown in Figure 4. The speed with which a given substance is eluded
from the column is inversely proportional to the temperature of the column.
Except for nitrous oxide, retention time is lowest on a silicone oil column (C). It
should be noted that a column temperature at least 10 ~ C. above ambient is neces-
sary to give stability to the system. It was seldom necessary to operate the svstem
above 75 ~ C.; liquid anaesthetic agents of high boiling point could be elucied at
this temperature in less than five minutes. The higher the temperature employed,
the shorter was the useful life-span of the column. At high temperatures the
organic liquid of the stationary phase boils off, requiring periodic renewal of ihe
column. For routine analysis a temperature (75 ~ C. ) that was consistent with
rapid determination of the compounds in question was employed.
Although any of the three columns can be used, the silicone off column was
chosen for rapid analysis at a temperature consistent with a long colurrkn life. It
is recognized that a "tailing-off" effect is seen when divinyl or diett~yl ether
1L~ I I~t,
I00 "C ETHER I TOO "C varying
ETHER 75 VINYL 1 7~
L
5O ~... .50
OlVll-,Iv B C 25 TEMPERATURE RLFLUR~ETHYL 01 A C I Z5 TEMPERATURE anaesthetics,
20 20 15 0 os
u,I t.d I-- Z -s o ~
series
A t25 1~.5 a
~ e~;, "C I00 "C s
ETHER i 75 7~ time
I, 5,0 .5~
D~ETHYL I ;'5 TEMPERATURE TRICHLOR~TNYLENE Z5 TEMPERATURE
Ec A
retention
20 15 ~ t- 0 2QC 14C lZO i "~0BO 40 20 0 mx

1 I~, J 12~5
packing c.c./min.
I 100 ! 1o0 "C
"C 90
I 75 OXIDE ~1 75' at
column
CYC LOPROPANF 9 A c I r ~5 50 TEMPERATURE NITROUS B C A I . I 25 SO I'EMPERATUR and constant
_
4 3 k- Z 0 -- z ~ t was
temperatore
',Z5
1Z3 , oF
flow-rate
~OO "C 1 io~ "~
I
RIE "t,~
T5
OFORI, Influence Helium
HLOR 50 M PERATU HALOTHANE -%Q TE.MPEKATURE 4.
C TI;" 1 z5
ZS
column.
200 tSO t60 120 40 20 100 E,s "/E m 60 z ~E 40 ~0 2~ IC FzcuaE the
234 NAIDANAC ANAESTHETISTS' YTEICOS JO~A.L
(Figure 2) is quantitated on this column. ~ Th~ practical aeettraey of the results is
not affected by the asymmetry of the curve.
~REPARATION OF SDRADNATS
Standard gas mixtures were made in gas burettes. Following preparation they
were transferred to gas-tight bottles. S~anda~ds of volatile agents were made by
vapourizing a known amount of liquic~ anaqsthetic, delivered from a calibrated
syringe, in a bottle of known volume~The gas mixture was agitated constantly
by a magnetic spin bar in the bottle. Volgmes were corrected for the space
occupied by the bar. Aliquots for standardiSation of the apparatus were drawn
into 50-c.c. syringes previously rendered air-tight by spraying with teflon aerosol.
It was found that the preparation of the standard was the greatest potential
source of error in the entire procedure. With some practice, however,, it was pos-
sible to prepare standards with a great deal of reproducibility.
CALCUI~TION FO RESULTS
The chromatogram (i.e., the curve inscrib d by the recor~er when the detector
bridge is unbalanced) may be related to ~_he concentration of gas or vapour
passing through the detector cell in one of two ways. The first entails determina-
tion of the area under the curve. This area, in turn, may be found in several
ways: (1) by electronic integration, (2) gravimetrieally, (3) by planimetry,
(4) by mathematical integration, or (5) it may reasonably be approximated by
-r
q._
HEIGHT
Fmv~. 5. The chromatogram ean be
related to concentration by determining
the area under the curve (stippled) or
by measuring the peak height. In some
instances the area can b~ approximated
by calculating height timds width at one-
half the height.
235
ZAIIDER & ORKIN: APPLICATIONS OF GAS CI:ROMATOGILA-PHY
multiplying the height os the peak by the width of the peak at one-half the
height (Fig. 5). Another method of handling the data is to relate the peak height
of the chromatogram to the concentration of the gas or ~apom'. This, the simplest
and least e~pensive way of handling the data, is quite accurate in the case so
these peaks (Table I)
The relation os the area to the percentage concentration so eight commonly
employed anaesthetic agents is shown in Figure 6. The relatio ashi, p is linear over
the entire range examined. In each case a 2-metre silicoze oil column was
employed; operating conditions for the individual anaesthe ics aIe given. The
* :r ' I t
plots are the averages of three determinations os at least six s :andards. The fit os
the curve to each os these points is a reflection so the Ihigh egree os accuracy
obtainable with this method. Those points which lie somewhat off the plotted line
CYCLOPROPANF~
CHLOROFORM
COOI 5OO :
102"C X 4
muld~H Flow 90co/rain
C08 'tOO
(4 003
C06
~ 4oc (cid:12)9 c 002
C0I
2oo
i i l 1 1
2 | 4 I I 6 I 1 lfI 0 20 40 60 081 100
PER CENT PER CENT
DIETHYL ETHER DIVINYL ETHER HALOTHANE
-- 0521
I000 45"C X 61 45"c x 61 ,4 loc~ 75"C X 4
-- , w , mu,lJcH rlo,, 85cc/,n.~ Hei,u.~ Flow 85cc/e,.
000I
OOB 80(
L
~, 006 ~ 750 ~06
w
004 ~ 5OO
C02 250
4 8 21 6 20 t 4 8 21 61 20 I I0 1 1 lo 1 1 1 1 02 1 1 30 04
PER CENT PER CENT PER CENT
NITROUS OX "FDI
TRICHLORETHYLENE TRIFLUFIOETHYL VINYL ETHER
30~ X128
r07 005
500 muheH FI.~ 125"C X 1 I0001-- 75"C X 61
I
4OO O04, i ~ 008 -rF Hehum Flow 261r09
3oo .r 3O0
J~
~r
2oo 200
IO0 t00 OO2
/I 1 I ~ I I I I 1 J k ~ l I I I
2O 4O 60 8O OQI 0 tO 2.0 0 25 5 75 I0
p(a C(~T P(R C(NT P(R CENT
T
FiGurm 6. Relation os area to concentration utilizing a 2-metre silicone oil column. Tempera-
tures, flow-rate, and attenuation are given for each anaesthetic agent.
236 NAIDANAC STSITEI-ITSEANA ~ YTEICOS LANRUOJ
TABLE I
/bOITARTNECNOC DETALUCLAC YB CINORTG~IE ,NOITARGETNI )rr-IGI,_FH
~/Ta.rw~xonPPA *~,~m4.
Approxmaate I (cid:12)9
,noitartnecnoC % Ar~a* ~ *thgieH ~ *gerA w
Diethy ! ether 6.8 40.01_+ ---0.05 7.2___
Divmyl ether 5.3 30.0___ _+0.03 2.2___
Halothane 2.0 ~o.oi Io.o=__ _0.8
Cyclopropane 20.0 ~0,09 80.0=_+ +_4.4
Triehloroethylene 2.4 ~0.02 10.0+_ 7.0-*-
Chloroform 4,0 -+0,03 40.0+__ 0.2___
*Standard deviation, 20 samples each.
fBy electronic integration.
~Measured m cm.
w from measurement of peak height ~nd width at one-half the peak b_eight.
are less likely to represent inaccuracy of th~ analytical method than an inability
to prepare accurately standard gas and vap6ur mixtures.
The linearity and accuracy as reflected yb these curves greatly simplify the
method of analysis in that only one standard is needed to tconstruct the calibra-
t.ion curve for any particular agent. Using a knowaa concentration that is approxi-
mately in the range of the unknowns and by adjusting the attenuation oS that
this concentration gives approximately full, scale deflection on the recorder it is
possible to increase the resolution of the analytical method to a point not ordi-
narily achieved by other methods of physical or chemical nnnlysis with any
degree of ease. The concentration of an unknown sample can be determined from
a calibration curve or since the concentration is directly proportional to the peak
area or peak height, an unknown can be de~ermined from the following equation:
C.~ -- C1A1j/A2,
..z
where 1C --- concentration of known,
zC -- concentration of unknown,
1A = area or height of known,
2A ~ area or height of unknown.
REPRODUCIBILITY FO RESULTS
Twenty runs of the same concentration were successively made to de~ermine
the reproducibility of the method. These results are summarized in Table I. It
will be observed that the degree of reproducibility for twenty samples is virtually
the same when analysis of the chromatogram is made by integration of he area
under the curve and by measurements of peak heights. When one appr( ximates
the area under the curve by multiplying height of the peak by the width at one-
half the height, the method becomes quite unreliable. This is, no d ~ubt, a
reflection of inability to measure, with suflleient accuracy, the width ( f) these
narrow peaks. For this reason this method of interpretation of the dl tta was
abandoned early.
ZAUDER & ORKIN-" APPLICATIONS OF SAG CI-IRONIATOGPtAPHY 237
O ~ SNOITACILPPA
Separation of mixtures of several anaesthetic agents is possible utilizing this
technique. This includes resolution of the fluo-ether azeotrOpe, 'which is not
separable,by other physical or chemical methods (Fig. 7). When this figure is
FLUO -ETHER
2 METER SILICONE OIL COLUMN
HELIUM FLOW 90cc/min
30"C X16
HALOTHANE ETHER
:~mcrcIF .7 Resolution of fluo-ether.
'~railin~-otr' of the ether peak is well
demonstrated.
2 METER SILICONE OIL COLUMN
C"001 X 61
HEUUM FLOW 90co/rain
3.2 % 4 8%
CH 3 - O- CF t- CHBrF EM THOXYFLURANE
Ficvrm .8 Chromatograms of methoxyfluorane and of a fluori-
nated hydrocarbon of potential use as an anaesthetic.