irig condition.. Although changes in gasoline base stocks rnay slightly alter the retention time., t i e relationship of each alkyl to T E L remains constant. Since the T E L retention time may be determined by ana1y:ls of a series of effluent samples, i t is not necessary to use a conventional recording chromatograph at any point in the new procedure. The first step in the calibration procedure is to obtain a steady temperature and a constant carrier gas flow through the unit. Remove thtb septum from the injection port and insert a dial thermometer until it touches the column packing. Adjust the voltage to the heating tape to give a steady temperature of 70" i 2 " C. Remove the thermometer and replace the septum and injection port nut to form a tight seal. Adjust the flow of carrier gas (helium) to 7 5 cc. per minute b y means of the pressure regLlator. The unit is now ready for sample introduction. T o a portion of t h t gasoline sample to be analyzed, add T E L antiknock fluid to bring the con1:entration of lead
as T E L to about 1 pg. per pl. (3.8 grams per gallon). (If the sample contains T E L equivalent to at least 0.2 pg. of lead per pl., the T E L addition is unnecessary.) If antiknock fluid is unavailable, add some commercial gasoline known to contain TEL. Inject a 50pl. sample into the column, starting a stop watch a t the moment of injection. Scrub the exit gas for 1 minute in each of six absorption tubes containing 5 ml. of methanolic iodide, as described in the original publication. The l-minute samples are collected at 3-minute intervals beginning with the 23rd minute after injection and terminating at the end of the 38th minute. Analyze each scrubber for lead according to the published procedure. A411 lead found will be as TEL, for the other lead alkyls will have passed through the column long before the first scrubbed sample was taken. The absorption tube containing the maximum amount of lead will indicate the retention time (5") of TEL. From this, the retention times
for the other lead alkyls can be calculated from the relationships: Retention time for ;Ile4Pb = 0 039 T LIejPbEt = 0 090 T IIelPbEtt
MePbEt,
=
0 200 T
= 0 470 7'
To analyze the original sample, establish the same column conditions as used for the retention time calibration. Introduce 50 pl. of the untreated sample and collect the individual lead alkyls as they are eluted. Change absorption tubes midway betiveen the retention times of adjacent lead alkyl.. Allow the last tube to remain in place 15 minutes beyond the T E L retention time. Finally, analyze the absorption tubes for lead a. described in the original publication. LITERATURE CITED
IT.W., Smith, G. Z., Hudson, R. L., A x . 4 ~CHEX. . 33, 1170 (1961).
(1) Parker,
A n Automatie, Continuously Variable Attenuator for Gas Chromatography Kenneth Abel' and m'm. B. Dabney, Research Division, Melpar, Inc., Falls Church, Va.
UTOJIATIC
\
ation and continuously variable attenuation. Each of the four sensing elements has a cold resistance of 44 ohms, the coarse zero is a 12-position switch with a total resistance of 2.5 ohms, the fine zero is a 10-turn, 25-ohm potentiometer with a 30-ohm resistor in parallel n i t h the winding. and the bridge current adjustment consists of a 250-ohm rheostat. The attenuator is a variable resistor across n-hich the bridge current produces a millivolt output, the millivolt output being inversely proportional to the resistance. A 6-poleJ double throw switch enables the use of either attenuator. The continuously variable attenuator is mounted within the recorder (Brown Electronik, 1-mv, with l-second full scale response time) and is geared directly to the pen servo motor so that full scale pen travel rotates the attenuator wiper arm to provide a continuously varying attenuation from 1 to m . A 1000-ohm (is%) single-turn linear (+0.2%) potentiometer with 8.5-inch wiper travel is in use in our laboratories at the present (Series 54-48JAJ Clarostat Mfg. Co., Dover, S.H.). The resistance was chosen to match the stepfunction attenuator resistance at its most sensitive setting; accordingly, this value applies only to the F&M instrument. For other chromatographs, a different resistor may be required. This potentiometer provides the response (relative to a step-function attenuator) as shown in Figure 2.
12 v
Present address, Laboratory of Technical Development, National Heart Institute, Bethesda, &Id.
attenual ors have demon-
A strated usefulness in both analytical
and preparative gas chromatography. Such attenuators provide chromatograms in which both minor and major peaks are presented without operator attention ( I , 2 ) . Ess:ntially the same end result can be obtained b y utilizing a recorder with a nonlinear response if only small attenuation ranges are required. A4nonlinear eesponse provides good sensitivity for minor components while the peaks of major components
remain on scale. Fonlinear recorders are available but are not in general use in gas chromatography laboratories.
A simple, low cost method of converting a linear recorder to provide a widerange nonlinear response is illustrated in the electrical schematic shown in Figure 1. This schematic shows the four-element bridge circuit used with F&M Scientific Corp. thermal conductivity detectors modified to provide both conventional step-function attenu-
7 FINEZERO
COARSE
Figure 1. Schematic of continuously variable system for 4-element thermal conductivity detector
ZERO
STEP ATTENIJATOR
".
9Q,
VOL. 35, NO. 9, AUGUST 1963
1335
1moo 1000 6000 4000
moo ,000 800
e00
p
400
3
p
200
12
100
L
6
80
so 40
20 10
a e 4
2 1
I
I
5'
10
20 ,J
LFOL
80
40
.res
$Lo
320
140
NJECTED
Figure 2. Comparison of nonlinear response with conventional linear attenuator response Upper curve: step-function attenuator Lower curve: continuous attenuator
Mathematically, this attenuator provides a system such that when z = 2", then y = 1 - 1/2n. For this situation, the response curve follows Equation 1: (1)
2-1=zy
where y is the recorder response (fullscale response being equal to 1) and z is the input signal. With this system the first 407, of the recorder response is nearly linear, allowing satisfactory quantitative determinations of minor components. Above this the response decreases rapidly, providing an equiva-
Table I.
Precision and Accuracy
Av. Sample response size ( % full(relative) scale) 1 0
3 0 5 0 10 0 20 0
12 S 25 4
34 1
5x 2i ( I
Std. Accuracy6 (%)
de^.^ i z 0 07
1 0 08 f 1 1 1
O 0 0 0
06
08
06 10
0.54
0 93 0 85 1 30 1 40 3 21
93 9 30 0 a Standard deviation = [ Z (deviation from average response)2/(n - 1)]1'2. b Accuracy = [(standard deviation)/ (average response)] X (relative sample size).
1336
ANALYTICAL CHEMISTRY
0
4
8
12
16
20
24
28
32
36
40 0 4 8 TIME IN MINUTE6
12
16
20
1
1
24
~
1
21
I
32
I
36
I
I
I
40
Figure 3. Typical chromatograms of methyl esters from impure oleic acid utilizing ( A ) continuous attenuation and ( 6 ) step-function attenuation
lent to more than 50 chart-widths in slightly less than full-scale pen movement. Due to limits of the potentiometer resolution, injections which normally require attenuations of lOOX or greater result in some pen servo oscillation, but this does not adversely affect the quality of the chromatograms. This oscillation is related to the very low resistance in the circuit at maximum attenuation and the step-change in resistance inherent in wire-wound resistors. For example, the first 50% of pen travel changes the attenuation from 1x to only 2 X while movement of the pen from 97 to 100% of full scale results in an attenuation change from 30X to m. The maximum stable attenuation obtainable is, therefore, directly related to the resolution of the potentiometer used as the variable attenuator. By using higher resolution potentiometers or multiturn, nonlinear potentiometers, greater stable attenuation ranges are possible. Base line drift with this instrument is less than 3% per 8-hour day with a completely stabilized column and detector block. Table I indicates the level of precision obtainable with multiple injections of ethylene glycol monobutyl ether. The standard deviation in this case include potential sampling deviation and is equivalent to that obtained with conventional attenuation systems. For quantitative analysej, a calibration curve is
required. The accuracy of analysis is a direct function of sample size; accordingly, we have found this system to be more useful for preliminary qualitative analysis. llfter the preliminary analysis, reference is made to the calibration curve and step-function attenuator settings are chwen which d l provide the desired response without an attenuation change during the analysis of additional portions of the sample. An example of the use of this attenuation system is shown in Figure 3 in which a chromatogram obtained using automatic continuously variable attenuation is compared against a chromatogram obtained with the same size bample using conventional step-function attenuation. I n addition to simplifying interpretation of the chromatograms, incompletely resolved peaks are easily seen in the continuouqly variable system but not in the step-function system. Furthermore, the continuously variable system can be utilized directly with presently available automatic preparative gas chromatographs. LITERATURE CITED
(1) Baumann, F., White, F. A., JohnBon, J. F., ANAL. CFIBM. 34,1351 (1962). ( 2 ) Darling, D. J., Miner, F. D., Bartsch, R. C., Trent, F. M., Ibid., 32, 144
(1960).
