747
Anal. Chem. 1981, 53, 747-748 (18) Schutten, H. R.; Beckey, H. D. Messtechn/k(Braunschweig) 1973, 87, 121. Chem. Abstr. 1973, 79, 71402. (19) Winkler, H. U.; Neumann, W.; Fassbender, B.; Hilt, E.; Beckey, H. D. Messtechnlk (Braunschweig) 1974, 82, 119. Chem. Abstr. 1974, 87, 97017. (20) Neumann, G. M.; Rogers, D. E.; Derrick, P. J.; Peterson, p. J. K. J . Phys. D 1980, 13, 485.
(2 1) Remy, H. “Lehrbuch der Anorganischen Chemie”; Akademische Verlagsgesellschaft: Leipzig, 1961; Vol. 2, p 68.
for review Oct&er 24, 1980. Accepted December 22, 1980.
Multichannel Peristaltic Pump with a Pneumatic Nebulizer for Atomic Absorption or Emission Spectrometry
L. R.
Layman,* J. G. Crock, and
F. E. Llchte
US. Department of the Interior, Geological Survey, Box 25046, MS 928, DFC, Lakewood, Colorado 80225 Pneumatic nebulization is the most frequently used method for introducing solutions into flame atomic absorption or plasma emission spectrometers ( I ) . It has gained this widespread usage because of its ease of operation and general reliability. It can suffer, however, from the fact that sample uptake rate is dependent on nebulizer gas-flow rate and sample viscosity (2). These effects can be magnified by misadjustment of the nebulizer (3). Recently, peristaltic pumps have been developed that can produce reasonably smooth sample-delivery rates. The use of such a pump with pneumatic nebulization makes the sample-delivery rate an independent variable ( 4 , 5 ) . Other variables in the nebulization process, such as droplet size distribution, are not controlled by the addition of the pump, however. This addition to the analytical schemes has been given very little attention in the literature. One or more dilution steps are often required to get the elemental concentration into the optimal calibration range of the atomic spectrometric method. Many elements also require the addition of an ionization buffer or chemical releasing agent. A multichannel peristaltic pump can simultaneously deliver sample, buffering or releasing agent, and/or diluent to the nebulizer, thus reducing sample preparation time (6-8). This communication describes procedures used in our laboratory and some of the results obtained. EXPERIMENTAL SECTION Apparatus. The pump used in this work is a Gilson Medical Electronics (Middleton, WI) Model Minipuls 2, four-channel, variable-speed peristaltic pump. The nebulizers used were a concentric type purchased from Perkin-Elmer Corp. and a cross-flow type purchased from Jarrell-Ash, Inc. Pump tubing is FISHER brand standard manifold tubing as supplied by Fisher Scientific Co. All other tubing used is Teflon. Perkin-Elmer Models 306 and 5000 atomic absorption spectrophotometerswere used to collect the data. Procedures. The pump is used as shown in Figure 1. Samples are pumped from an autosampler tray to the nebulizer. A second channel of the pump is frequently used to deliver a buffering reagent and/or diluent to the nebulizer simultaneously. The delivery rate is proportional to the pump rpm and to the square of the tubing internal radius. The dilution factor is determined by the relative tubing diameters used for the sample and diluent channels. For the analysis of a group of samples using an autosampler, the pump is left on continuously, eliminating the need for operator attention during the run. R E S U L T S AND DISCUSSION Current nebulizer systems (both cross flow and concentric) used for flame atomic absorption, flame emission, and plasma emission are often unable to maintain a constant sampledelivery rate. With a pump, solution delivery is unaffected by changes in nebulizer gas flow (i.e., with changes in flame stoichiometry) or in sample viscosity (caused by changes in salt or acid concentration or solvent composition). Therefore,
Table I. Analytical Conditions for Some Elements in Example Solution sample buffer flow flow element rate, rate, dilution determnd mL/min buffer compn mL/min factor Na
0.50
lOOOpg/mLK
K Ca
0.50 2.50
Sr
7.50
lOOOpg/mLNa 5000pg/mLLa none
7.00 7.00 5.00 none
1:15 1:15 1:3
none
long-term precision can be improved in some cases with a use of a peristaltic pump because the pump can maintain a constant sample delivery rate. In addition to holding the delivery rate constant with changes in other parameters, the pump system also allows the delivery of solution to be varied independently. This allows selection of a sample-delivery rate to maximize the desired signal under various circumstances. For example, as Figure 2 shows, the maximum absorbance signal is achieved for a concentric nebulizer-flame atomic absorption system, with a pumping speed of about 6-11 mL/min. However, as shown in Figure 3, a much higher efficiency of nebulization occurs a t low flow rates (9). A 3 times greater integrated absorbance-time signal can be obtained by using a flow rate of about 1 mL/min, rather than the usual 7 mL/min. In this case, the absorbance signal is reduced by about half and a given solution volume will yield 7 times the signal duration. This higher efficiency can become important when only a small volume of solution is available. On-line dilutions or reagent additions are accomplished, as shown in Figure 1, by pumping the sample through one channel and the reagent (e.g., ionization buffer or chemical releasing agent) through another and mixing the solutions before they reach the nebulizer. Dilutions up to 20:l can be successfully handled in this manner. Larger dilution factors frequently create problems with precision due to the magnification of any very small irregularity in the pump. A typical example that demonstrates the utility of this process is the analysis of a set of samples that contain on the order 100 pg/mL Na+, 100 pg/mL K+, 10 pg/mL Ca2+,and 0.1 pg/mL Sr2+in the presence of phosphate. Normally, four aliquots of each sample solution would be taken. Each aliquot would then be quantitatively diluted with the reagent required by one of the elements and then that element determined. With a multichannel pump, however, the original solution can be sequentially analyzed by using the second channel to deliver the required reagent or diluent. Adjusting the relative flow rates of the two channels and the pump rpm allows the proper dilution factor and total delivery rate to be chosen. The reagents and dilution factors used for the above samples are
This article not subject to US. Copyright. Published 1981 by the American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 53, NO. 4, APRIL 1981
748
NEBULIZER
AIR
BURNER
SAUTOAMPLER
1 FUEL
gL 1 \-;I TRIGGER
,
,
L A K H ENABLE
&yJ D4TA LATCH
AUTO-SAMPLER TRAY 7+
TO
CHANNEL 2 REAGENT (BUFFER OR DILUENT)
PHOTOMETER SPECTRo-
i
,
T I Y E DELAY
L.-A p~
o”riNIBLE_-p-
__
.-. .
A
-Ai
Figure 4. Functional diagram of the timedelay module inserted in line between the PE AS-50 autosampler and the PE 5000 spectrophotometer. The SIX signal lines delayed are as follows: Read, Auto-Zero, Std. 1, Std. 2, Std. 3, and Reslope.
Figure 1. Multichannel pump system diagram. 8.6
A
J/J/J/L
P E R I S l A L T I C PUMP
I
-~
-
a timing problem when using an autosampler. For example, the Perkin-Elmer Model 5000 atomic absorption spectrophotometer with the Perkin-Elmer AS-50 autosampler system allows a maximum of 15 s delay time for the sample to reach the nebulizer before initiating a read cycle. When large dilution factors are used (i.e., low sample flow rate), the time required for the sample to reach the nebulizer can be as long as 30 s. To obtain the required delay times, we added an outboard module between the autosampler and the instrument to delay the “ R e a d pulse as required. Figure 4 shows a block diagram of this delay module. Referring to the diagram, one e n L~-.LL--u sees that when the trigger circuit detects a pulse on any of s 2 8 48 ~e ~e ~ n e 120 IIB ~ i s iae ae the six-signal lines from the AS-50, it enables the data latch to store the signal and starts the time-delay circuit. After the simE WIYER~ *if (nm) selected delay period (from 6 to 60) the delay circuit enables Figure 2. Absorbance of a 10-ppm Cu solution at 324.7 nm as a the line driver to transmit the signal stored in the latch to the function of sample delivery rate. El = optimized nebulizer without P E 5000 which in turn starts the read cycle. The use of this Pump. simple time-delay module lowers the sample throughput rate but allows completely automated operation of the P E 5000 system. A similar time-delay module can be implemented for other spectrometer-autosampler systems if needed. Any spectrometric instrument using pneumatic nebulization can be modified to include a multichannel peristaltic pump to deliver sample solution to the nebulizer. Precision is improved through a more constant sample-delivery rate. Sample preparation time is shortened by using a second channel of the pump to perform on-line dilutions and/or reagent additions. The problem caused by increased sample transit time in conjunction with the autosampler can be solved, as with the Perkin-Elmer AAS-5000-AS-50 system, by the addition of a simple time-delay module. e 2 8 18 ne 88 1e8 120 118 ine 180 288
---
SURE
mrvei RATE (ncRIm)
Figure 3. Relative nebulization efficiency as a function of sample delivery rate. Vertical axis is (absorbance units)(minutes)/(milliliter of
solution) for a 10-ppm Cu solution at 324.7 nm. 0 = optimized nebulizer without pump.
shown in Table I. Several additional benefits are also derived. Fewer manual manipulations of the sample lead to a reduced risk of human error and less chance of contamination. A greatly reduced volume of reagent is required. The dilution factors are automatically compensated by running standards with the same dilution factor as the samples. The more concentrated standards used with this method are more stable with time and less likely to become significantly contaminated than are the more dilute standards normally used (10). The relativity long time required for samples to reach the nebulizer through the pump and associated tubing can cause
LITERATURE CITED (1) HieftJe,G. M.; Copeland, T. R. Anal. Chem. 1076, 50, 300R. (2) Greenfield, S.; McGeachin, H. McD.; Smith, P. B. Anal. Chlm. Acta 1976, 84, 67. (3) Kolrtyohann, S. R.; Wen, J. W. Anal. Chem. 1973, 45, 1986. (4) Garbarino, J. R.; Taylor, H. E. Appl. Spectrosc. 1979, 33, 220. (5) Beadecker, D. R.; Williams, L. L. Jarrell-Ash News/. 1978, 1 , 5 . (6) . . Robinson, J. W.: Wolcott, D. K.; Rhodes, L. Anal. Chlm. Acta 1975, 78, 285. (7) Fuller, C. W.; At. Absorpt. News/. 1978, 15, 7 3 . (8) Niemojewski, S. Tech. Posyuklwan Geol. 1977, 16, 52. (9) Willis. J. D. Soectrochim. Acta, PartA 1067, 23A. 811. ( i O j LaFleur, P. D.; Ed., “Accuracy in Trace Analysis: Sampling. Sample Handling, Analysis. Proceedlngs of the Seventh Materials Research Symposium;” NBS Spec. Pub/. U.S.1976, No. 422.
RECEIVED for review August 22, 1980. Accepted December 22,1980. The use of trade names is for descriptive purposes only and does not imply endorsement by the US.Geological Survey.
