Automated reaction-rate methods of analysis - Analytical Chemistry

AUTOMATED REACTION-RATE METHOD8 OF ANALVSIS HOWARD V. MALMSTADT EMlL A. CORDOS COLLENE J. DELANEY School of Chemical Sciences Roger Adams Laboratory University of Illinois Urbana, Ill. 61801

Reaction-rate methods of analysis often offer advantages in selectivity and accuracy. New automated instrumentation encourages the development of specific procedures and the likely adoption of these methods

26A

THE

FIRST “ A U T O ~ I A T E D ” reactionrate (kinetic) methods and instrumentation for the rapid selective determination of glucose and other constituents werc reported over a decade ago (1-6). Through the 1960’8, analytical methods based on the measurement of initial rates of chemical reactions grew in importance as many new sensitive and selective procedures were reported, and the instruments were improved. The frequent journal review articles and recent books attest to the interest in kinetic methods (7-12). I t seems now that it yill be the middle of the 1970’s \Then reaction-rate methods become generally accepted and widely used on a routine basis. It is certainly not coincidence that the conversion from research interest to general routine use of reaction-rate methods corresponds to the development of a new generation of elegant, completely automated, and computer-controlled chemical instrumentation. The rather recent concepts and advances in analytical instrumentation make it just as easy to obtain sensitive quantitative chemical results Tvith reaction-rate methods as with conventional stoichiometric (equilibrium or endpoint) methods. By elimination of the barriers imposed by difficult laboratory rate techniques, it nolv becomes worthwhile to consider the inherent advantages of reaction-rate methods. I n this report the inherent advantages and possible limitations of reaction-rate methods as compared to equilibrium methods are reviewed, and the general concepts of

ANALYTICAL CHEMISTRY, VOL. 44, NO. 12, OCTOBER 1972

encoding reaction-rate information are presented. Hon-ever, the major discussion is focused on the automated systems that make it possible to perform hundreds of accurate, sensitive, and selective quantitative determinations per hour via rate data and to develop nen- methods more rapidly. ADVANTAGES AND LIMITATIONS OF REACTION-RATE METHODS

With rate methods it is often possible 1 o measure, immediately after mixing the reactants, the rate of change of some parameter P of the reactant whose concentration is to be determined, or other reactant or product of the reaction, and not wait for the reaction to go to completion (equilibrium). This is illustrated in Figure l. The saving in time may or mag not be significant depending on the specific reaction, but there are good examples (1-12) of obtaining quantitative rate results in seconds for selectivc reactions that would have required many minutes or hours to go to completion. This is especially true for many of the highly selective enzymatic reactions. Because it is possible t o obtain quantitative rate data shortly after the reagents are mixed, the measurement may be completed before interfering side reactions begin. This can be a distinct advantage in providing higher accuracy for some determinations. One of thc most important characteristics of the reaction-rate method is that it involves a relative

ANALYTICAL CHEMISTS measurement. The absolute value of thc parameter (i.e., absorbance, cell potential, fluorescence) chosen to monitor the rcaction does not have t o be measured accurately, as shon-n in Figurr 1. It is only necessary to measure t h r parameter's time rate of change with high precision and accuracy. Hence, even for cxtremely rapid reactions, the reaction-rate method can offer freedom from those interferences which contribute to the absolute value of the parameter (turbidity, dirty cells, junction potentials, and other fluorescing materials) but do not enter the chemical reaction and do not contribute to the rate of change of the paramcter n-ith time. Of course, there are those applications nherc it is the rate per se, rather than the absolute concentration of a specific species, which is the important quantity to be determined. Most noteworthy in this category is the determination of eiixyme activity. Another possible advantagc of kinetic methods is that they sometimes provide a means of determining the concentration of two or three constituents of closely related chemical properties without physical separation. XZa rule of thumb, the successful development of differential rate methods requires that the first-order rate constants of the individual components differ by a t least a factor of 10. For example, silicate and phosphate in mixtures have been determined by a differential rate procedure based on the formation reactions of the heteropolymolybdate and the reduced heteropolyblues ( I S ) . There are some limitations in the general application of reaction-rate methods. The most important is that imposed by the reaction rate itself. The half-time of the reaction must be greater than the mixing time of the instrumental system available. Considering the other extreme, very s l o ~reactions with half-times greater than a fex hours are not too practical for routine analyses. Also, the accuracy and precision of the measurement depend upon good reproducibility (although not necessarily good accuracy) for all experimental conditions such as temperature, pH, ionic strength, size, and shape of reaction vessels.

GENERAL CONSIDERATIONS I N ENCODING CHEMICAL REACTION-RATE I N FORMAT1ON

Determination of Glucose. X few practical examples of reactionrate methods are presented here to illustrate the general considerations involved in the encoding of rate information. The first example is the quantitative determination of glucose by use of the wdl-lmown selective oxidation of glucose in the presence of the enzyme, glucose oxidase, as illustrated by Equation 1. Glucose

+ O2

Glucose oxidase

Gluconic acid

+

+ H2O2

(1)

Encoding Indirectly V i a Coupling Reaction. In our own laboratory n-e first became heavily committed to the development of rate methods more than a dozen years ago by devising an automated rate measurement system for the determination of glucose in blood serum (1). For expediency we first chose to determine the rate of change of glucose by indirectly following the rate of formation of H?O?. Since the H202did not have a physical parameter that could be easily and directly measured, the relatively fast coupling reaction of H202with an organic dye v a s used. This was conventional practice at the time for the end-point methods

because the colored reaction product, as illustrated in Equation 2 , provided high sensitivity by photometric measurements. Organic dye A

+ H2O2 xy-Peroxidase

Colored product B

(2)

Unfortunately, the peroxidase enzyme that catalyzed the coupling fast reaction was rather unstable and expensive, and the automated two-point, reciprocal time, ratemeasuring system was a t first relatively crude. But from this example, it became apparent that sensitive and precise quantitative measurements could be realized lvith the aforementioned advantages over end-point methods if suitable reactions could be selected and controlled; sensitive devices for converting (transducing or encoding) the concentration change for one of the reaction species to a measurable electrical signal could be developed; and reliable electronic systems for the rate measurement could be developed. The glucose reaction then became a test system for demonstrating nelv encoding systems and rate measurement systems. By use of a different fast coupling reaction, as illustrated in Equation 3, it was possible to determine the rate of change of H20s

FilZUl'e 1. Initial rates of rc2acticIn propoirticinaI I to C'01icentra.tion of so ught,.fl 3r speci es


27A

Report for Analytical Chemists Molybdate

A Non-Destructive Method for Separating Cells, Viruses, Cell Particles and Macromolecules

THIN LAYER COUMTER CURRENT DISTRIBUTION

H2°2

+ 31- Catrtlystc 1,-

CIRCLE 25 ON READER SERVICE CARD


(3)

by following the 1 3 - nith potentiometric (a),spectrophotometric ( S ) , and amperometric (14) encoding (transducer) systems, and to eliminate the troublesome peroxidase enzyme. I n Equation 1 thc rate of formation of gluconic acid could be used to obtain the initial rate information. Therefore, a sensitive digital pH-stat system (15 ) was devised \\-herein the p H \\as held constant by adding small increments of S a O H during the glucose reaction. The number of increments of base added during a short fixed period of time proved to bc directly proportional to the glucose concentration. Howver, the system is not as sensitive as the methods based on the H202coupling reactions. Encoding Dwectly T’ia Piimary Reactio?i. The obvious dpsire to eliminate any type of sccondary or coupling reaction also lcd t o the investigation of direct methods for following the rate of change of 02. Yoltanimctric encoding systems that could measure O2 directly (16-18) I\ erc developed, although again the method \\-as not as sensitive as the indirect H102 color methods. The conversion of glucose or gluconic acid rate information to a directly measured physical parameter has not been reported. Deter rn inat io n of Phosphate. Et!cotliii q I n dzrectl y T7zu Seco t i c1ai.y Reactton. The determination of phosphate by a rat(. method illustrates similar considerations ab thosc for the glucose determination. l‘irst, the classical molybdcnum bluc procedure could bc dcvclopcd into a sensitive rate method involving the primary reaction of phosphate with Jlo(V1) to form 12-molybdophosphoric acid (12-AIPA) and it3 subsequent reduction to form the heteropolyphosphomolybdcnum blue, YMB (19), as sho\\-n in Equations 4 and 5 .

12-:\IPA

28A

+ 20H-

+ 9H+

(4)

Report for Analytical Chemists

RAW S A M P L E

REAGENTS

PREPARATION

PRE P A RAT I O N

c I

I

I

tively large choice of methods is now available (1-21). The method of choice for encoding the rate information from either the primary or coupling reactions depends, of course, on several factors including sensitivity, freedom from interferences, simplicity, and dependability.

I

ALlQUOTlNG Et MIXING

INSTRUMENTATION SYSTEMS FOR REACTION-RATE METHODS

REACTANTS

CON T R 0 L SYSTEM

DATA D O M A I N CONVERTER

4

Figure 2. Block diagram of complete instru m enta tion system for reaction-rate methods

RATE MEASUREMENT DATA MANIPULATION

OSCILLOSCOPE

RECORDER

I OUTPUT I DATA

DlSPL AY

DIGITAL

In this case, the measured reaction rate depends on the nature of the reductant, order of adding reagents, as well as acidity, etc. (19).

Encoding Directly Via Primary Reaction. It was again obvious that it would be advantageous to eliminate the reduction reaction (Equation 5) and convert one of the species in the primary reaction to a physical parameter that could be readily measured. Fortunately, the 12-1IPA has a relatively high absorptivity at readily available wavelengths. However, for the first time in our development of quantitative reaction-rate procedures, there was a confrontation with a reaction whose half-life was so short that it was necessary to make rate measurements in milliseconds rather than seconds. A rather long development program was thus required to develop the new automated equipment that would provide rapid mixing and readout of rate information. After the automated system was developed, it was then possible to determine

phosphate directly (20), utilizing the primary reaction shown in Equation 4. It is feasible to make as many as 3000 determinations of phosphate concentration per hour with this type of system. Geneyal Conclusions. It is seen from these examples for glucose and phosphate that a sequence of chemical reactions is often used to obtain a chemical species that is related to the desired rate information and readily converted to a sensitive measurable signal, or to obtain a readily measured chemical species whose rate of formation is slow enough to be measured by available equipment. illthough the methods utilizing multiple sequential reactions can be dependable under carefully controlled conditions, almost inevitably they will be less reliable than methods using measurements on one primary reaction. Much research effort has been invested in providing more sensitive and selective encoding systems and high-speed automated ratemeasuring instruments, and a rela-

The work on automating reaction-rate methods in the past decade primarily provided improved rate measurement devices (as described in this month’s Instrumentation feature, page 79 A) and sensitive data domain transducers for converting the chemical rate information to measurable electrical signals. Today, the attention is being focused on complete systems that start with the raw samples and reagents and end with a formatted printed readout in the desired quantitative units. A general system for obtaining quantitative data is illustrated by the block diagram in Figure 2 . By preliminary treatments (e.g., dissolution, dilution, filtration, ion exchange) the sample and reagent solutions are prepared as required for the specific procedures. Predetermined volumes of sample and reagents are then introduced, mixed, and transported to a vessel which serves as the reaction cell. The chemical reaction is monitored by a suitable transducer lvhich converts the rate information about a specific species inherent in the reaction to a measurable signal in the electrical domain. Frequently, several interdomain conversions are required to obtain data in the preferred form. The control and rate measurclment systems can be hardwired for specific applications, or they can be incorporated in a minicomputerinterfaced system that can provide through softiyarc much versatility in control of the measurement sequence and thc processing and readout of data. Readout is visually displayed with digital lights or printed out on serial teletype or a high-speed parallel printer. When desired, a servo recorder or storage oscilloscope can display the parameter vs. time and rate curves.


29A


*

TO W A S H O U T

~~

TO

DILUENT

DRIVEN S T I R R E R UNIT B

MOTOR

FOR COMPOSITE

ASPIRATOR PUMP

TR44SDUCER

C0,VTROL

i NTERFACE

SYSTEM

DIGITAL

RE4DOUT

STEP I MAGNETIC

STIRRER

Figure 3a. Block diagram of automatic potentiometric reactionrate analyzer (5)

The weakest links in reactionrate instrumentation systems in the recent past have been the sample and reagent preparation and aliquoting and mixing systems. I n fact, it becomes apparent that the greatest differences in completely automated reaction-rate instruments will probably be the methods of sample and reagent preparation, aliquoting, and mixing. These operations arc not only tedious, repetitious, and time demanding when done manually but are subject to human error and bias. It is not unusual that the sample-handling procedures require much more time than the rate measurement itself. This, of course, is not a problem exclusively associated with the development of reaction-rate methods; rather, it is an important consideration in the development of all automated analytical met hods. One of the first reaction-rate instruments that incorporated automated pipetting of sample and reagents and provided rapid deproteinization and filtration of serum samples n-as presented by llalmstadt and Pardue ( b ) . The instrument u-as designed specifically for glucose in serum determinations. A potentiometric concentration cell, Figure 3a, n-as designed as the transducer to encode the initial rate information. One automated injection pipet (22) \vas used to deliver a 1-ml aliquot of composite reagent to the sample compartment. h second automated pipet delivered a measured 36A

STEP 2

STEP

3

STEP 4

Figure 3b. Manipulations for filtering, measuring, and delivering deproteinized serum or plasma samples. Step 1: picking u p polyethylene sleeve. Step 2: picking u p filter paper. Step 3: filtering deproteinated sample. Step 4: delivering sample into reaction cell

aliquot of diluent to n-ash out the calibrated delivcry tip that contained an accurate aliquot of deproteinizcd serum. The rate information for glucosc \\-as automatically measured n ithin about 30 see. The reaction mixture \\-as then removed through an aspiration tube, and the cycle repeated. The deproteinization, filtration, and pipetting of an accurate aliquot of samplc into the reaction cell is schematically represented in F'g '1 ure 3b. The serum sample and deproteinizing reagents are added to a small cuvette. Another sampling pipet immediately d r a m an accurate aliquot of the dcproteinized sample solution through a glass-fiber filtcr \\-hich retains the precipitate. When the calibrated tip is filled with sample filtrate, the filter is knocked off, and the thrreway valve is turned to connect the diluent injection pipet to xach out the sample and diluent into the reaction vessel. -411 of the operations illustrated by Figure 3b can be performed automatically. iilthough it is, of course, preferable to develop a procedure that does not require filtration, the method does illustrate how classical deproteinization, filtration, and pipetting operations can be automated in discrete steps. Many of the newest commercial instruments operate in similar discrete steps. Considerable research, dcsign, and engineering activities are currently being directed to automation of aliquoting and mixing of reactants, and many commercial


models have only recently appeared on the market. Automated Stopped-Flow ReThe action-Rate System. stopped-flon- apparatus is widely used for kinetic studies (25, 24) of fast reactions, and it has been modified (20, S I ) for rapid automated reaction-rate methods, as illustrated in Figure 4. l l a n y sample and reagent handling systems arr only suitable for utilizing relativcly slow reactionrate systems. As pointed out earlier, there are advantages in selecting a fast reaction for analytical purposes. The limitation in making rate measurements on a fast reaction is that the time taken for mixing and observation must be shorter than the half-life of the reaction. Today, observation time can be reduced to the millisecond range by use of fast-responding detector-readout systems (25, 26) so that the real limitation becomes the time rcquircd for physical mixing of reactants and initiation of reaction. To minimize this tinic, mechanical systems arc used hich drive reactants rapidly enough to promote turbulent flon through the system and thereby insure rapid mixing and uniformity of solution composition in thc observation cell. Although a number of flow methods have been used for studying fast reactions, thc stopped-floiv method has been most n-id(>lyapplied for analytical purposes. This technique consists of rapidly mixing reactants by forcing the solutions through a mixing chamber and into

ReDOrt for Analvtical Chemists

as a control system for a number of operations. At the end of a measurement cycle, the logic circuit provides a signal which activates the syringe drive circuit, thereby injecting sample and reagent automatically. The number of rate measurements per injection can be preset on the logic unit. Likewise, the number of injections per sample cup is programmed. A preselected number of injections are used to insure complete flushing

an observation cell. The flow of solution is abruptly stopped, creating a back pressure which completes the mixing, and t h r rate measurement is rapidly made. Once experimental parameters have been selected, and samples and composite reagent have been loaded on the sample trap, the modified stopped-flow measurement system (31) operates automatically and continuously until reagmts are exhausted. A logic circuit is used STOPPING

BAR&

TO DRAIN VALVE

VALVE

4f

OBSERVATION CELL

I

I M I X I N G CHAMBER I I

SAMPLE SAMPLE CELLS SAMPLE PICKUP SYSTEM MU

M M

DOUBLE 3-WAY

I

Figure 4. Schematic diagram of automatic stopped-flow system

(PO)

1-4

I REAGENTS

I I

I

I 2 - M L SYRINGES TURNTABLE AND SAMPLE DIPTUBE MOTOR

CIRCUIT

-

of the flow system of previous reacted sample. During this flushing interval, the readout system is locked by the control circuit. After the preset number of injections and measurements have been made, the logic circuit activates the sample tray, and the next sample cup is brought into position. A teletype is used to log the data and to generate a paper-punched tape. The paper tape is subsequently used to input thc data to a small computer (Digital Equipment Corp. PDP-SL). Results for standards and samples are computer averaged and corrected for blank if necessary. iileast-squares routine is used to provide printout of concentration of the samples. Automated System with Prepackaged Reagents. A novel method for automation of reactionrate and equilibrium techniques minimizes the problem of laboratory reagent preparation and contamination by prepackaging all required reagents in stable form in a disposable plastic pack. The same pack also serves as the reaction chamber and observation cell. This approach has been initiated by D u Pont (E. I. du Pont de Semours 8: Co., Wilmington, Del.) in the development of the automatic chemical analyzer (aca) system. As shown in Figure 5 , the reagents are contained in plastic bubbles near the top of the pack. The pack header contains the name of the test and a binary code which is interpreted by a small, R O N

@---L

DRIVE SYRINGE MOTOR

MICROSWITCH

CHROMATOGRAPHIC COLUMN

TO OUTLET VALVE

NEEDLE CARRIAGE RAIL /

POSITION A SAMPLE INTAKE

DILUENT INJECTION TEMPORA R Y SEALS

Figure 5. Courtesy

PERMANENT SEAL

REAGENT COMPARTMENTS

SAMPLE CUP EXIT TRAY

Analytical pack for Du Pont automatic clinical analyzer

of E. I.

du Pont de Nemours 8. Co., Wilmington, Del.

Figure 6.

POSITION C SAMPLE & ELUANT INJECTION

Filling station for Du Pont automatic clinical analyzer

Courtesy o f E. I. du Pont de Nemours & Co., Wilmington, Del.


31 A


Hungrs

(read-only-memory) computer. The code specifies experimental conditions such as sample volume, reagent dilution, and processing cycle in addition to providing for automatic setting of specified instrumental parameters and for manipulation of the data so that the final printout is in desired units. Since the code is embossed on the top of the header, the packs may be loaded in any order to provide the desired sequence of analyses for a given sample. An additional feature is the inclusion of a gel filtration or ion-exchange column in those packs for determinations which require the elimination of interfering components. Initially, the sample is placed either manually or by automatic sampler into a plastic cup to which an identification card is clipped. The cup assembly and test packs, after being manually loaded on the input tray of the analyzer, are automatically transported into the filling station shown in Figure 6. Here, the binary code on the individual test packs is decoded by the RON computer, and the filling needle moves sequentially to perform the operations of flushing the delivery needle (position B in Figure 6), withdrawing the specified volumes of diluent (or elution diluent if a chromatographic column is used) and sample (position A in Figure 6), and injection of sample and diluent into the pack (at position D or C in Figure 6). The fluid metering system consists of a series of reagent reservoirs which are connected to a piston pump through a series of precision microvalves. The pump is driven by a stepping motor. The volume of liquid drawn and delivered by the pump is proportional t o the number of impulses applied to the stepping motor by the built-in computer. The operation of the microvalves is also computer-controlled. At the conclusion of the filling cycle, each pack is automatically moved into the main processing section of the analyzer. As the pack proceeds on a transport chain, it is heated to and maintained at a constant temperature, and reagents are released and mixed at two breaker-mixer stations by rupture of the reagent bubbles and oscillation of the pack.


32 A

ANALYTICAL CHEMISTRY, VOL. $4, NO. 12, OCTOBER 1972

Upon completion of this cycle, the pack advances to the photometer station where a set of mechanical jaws close around the pack generating hydraulic pressure within the pack to form a precision cuvette between quartz windows. The initial rate is measured by use of the fixed-time technique, and the computer automatically supplies an appropriate conversion factor so that concentration or activity is printed out on a report sheet along with the sample identification number. The ROM computer is also used for surveillance of all operations required for processing the sample. Should a malfunction be detected, the system also prints out a code on the report sheet indicating the source of error. Parallel Vs. Sequential Mixing of Samples and Reagents. I n the sample and reagent handling approaches described in the above methods, the reactants are aliquoted and dispensed into the mixing chamber, and the process repeats sequentially for a series of samples. I n most cases, the reagents are directly expelled from a pipet into the mixing chamber where the mixing operation is immediately carried out. An alternate approach can be taken by using a time separation of these trTo operations and then the simultaneous (parallel) mixing of reactants for a group of samples. All reactants are premeasured and introduced into reactant wells. The mixing operation takes place a t a later time. This principle of parallel mixing is applied in the unique GeX’ISAEC analyzer developed by Anderson (27) and Coleman et al. (28) a t Oak Ridge National Laboratory (Oak Ridge, Tenn.). The latest system allows parallel mixing of reactants for up to 42 samples. Centrifugal force is used t o transfer reactants into separate mixing chambers which also serve as the observation cells. Samples and reagents are manually or automatically pipetted into individual holding compartments of the transfer disk as shown in Figure 7 . The transfer disk is then manually locked into place in the rotor which also insures proper indexing of the samples. The holding compartments are arranged in two or more concentric circles in the CIRCLE 38 ON READER SERVICE CARD

----)

ow you can save time and eliminate the complicated set-ups normally required to make low level dc measrements. And do it t h e sweetest way. . . with state-of-the-art precision. The time-shaver is our allnew, all solid-state Model 180 Digital Nanovoltmeter. Simply connect it directly to your unknown potential. And . measure. All the way down to 100 microvolts, with 4%.digit resolution. What puts it all together is a sweet combination of features. Like automatic ranging. Remote programming.

..

Analog output. Floating input isolated from output. Zero suppression. And a systems-oriented BCD output for automating data acquisition that makes the 180 totally computer compatible. Plus specs that sweeten its amazing sensitivity. Like +0.05% accuracy, full scale. 3 0 megohms input impedance. Less than 30 nV noise, p-p. Normal mode rejection greater than 90 dB at line frequency, 75 dB a t twice line frequency. And a c o m m o n mode rejection ratio greater than 160 dB at dc, line frequency or twice line frequency.

Add these features and specs and you have a hands-down winner for labs, production work and systems applications at only $2190, including BCO output; $1995 without. See how sweetly the Model 180 puts it all together. And get our latest "How Sweet.. ." button. Call your Keithley Sales Engineer or contact us direct. Keithley Instruments, Inc., 28775 Aurora Road, Cleveland, Ohio 44139. In Europe: 14, Ave. Villardin, 1009 Pully, Suisse. Prices slightly higher outside the U.S.A.

KEITHLEY

Tte svCren eT;esT;

wyeve~

TO TAP 30 NANOVOLTS OUT OF100 MICROVOLTS

The Model 180 Digital Nanovoltmeter. . . latest in the Keithlev tradition of sensitive instrumentation

CIRCLE 1 1 8

34A

0

ON

READER SERVICE CARD


It's either or

TWICE AS MUCH for your money,,,

1.

2.

3.

4.

The multi-speed chart drive is the equal of a n y competitor on the market today. Further, the OmnlScrlbe,,, is field adjustable. precisely for English/ Metric scaling. Quick, fast, convenient.

Houston Instrument h a s perfected sprocketless paper drive. It's accurate. It's selfaligning. It works!

A unique operating

Did we mention the price? Just $395 for single pen; $595 for dual pen. With options to $1400. What's more, all models are available off-the-shelf for immediate shipment.

feature is the noncontact re-balance potentiometer. It eliminates cleaning a n d trace "glitches" forever. It's beautiful!

THE FULL PRICE IS $39500 'For single pen model

I

Write today for your complete brochure and price list. The OmnlSarlbe, is either half the cost or twice a s much for your money. You decide.

1

Instrument 4950 TERMINAL AVENUE, BELLAIRE, TEXAS 77401 (713) 667-7403 CABLE HOINCO.



e

35A


I

S4MPLE BEFORE TR4NSFER

4x1s OF

ROTATION

& 1 I

/

,REAGENT BEFORE TRANSFER

//

PHOTOMULTIPLIER TUBE To

LREMOV4BLE TRANSFER DISK

I

2

3

i

5

6

7

8

9

$0 !,

12

: 3 , 4 IS

CUVETTE NUMBER

Figure 7. Sections of the GeMSAEC fast analyzer. Upper left: loading of transfer disk. Upper right: transfer disk and centri. fuga1 analyzer section (27, 28)

Figure 8. Oscilloscope traces from photometer output of 15cuvette GeMSAEC analyzer. Upper: signal for a single cuvette. Lower: display for water blank and duplicate sets of protein standards ( 2 8 )

Courte~yof International ScientiBc Communications, Inc.. Green Farms, Coon

C o ~ l t e s yof International Scientific Communicationr, inc., Green Farms, conn.

transfer disk, with each circle containing the same number of compartments. The compartments are aligned radially and are angularly arranged so that samples and reagents are dumped through a transfer cavity into the corresponding reaction chamber (which also serves as a photometric cuvette) when the rotor reaches a sufficient speed (typically 400 rpm). To prevent lateral splashing of the solutions in the transfer disk, the rotor is gradually accelerated, and then during a breaking interval, a pulsed vacuum is applied which draws a stream of small air bubbles through each cuvette. The resultant turbulence produces complete mixing of the reactants. The air bubbles are removed by rapid acceleration followed by deceleration to the normal operating speed of about 500 rpm. I n present designs this entire process requires only a couple of seconds. The cuvettes rotate past a stationary light beam completely interrupting it between cuvettes. 36A

The signal from the photomultiplier is continuously displayed on an oscilloscope uvhich is synchronized with the rotor to provide an immediate visual monitor of the signals for all of the cuvettes. The photomultiplier signal is also passed through a buffer amplifier, directed to an analog-to-digital converter, and subsequently to a P D P 8/I computer. During each rotation of the rotor, the dark current value is obtained between each peak, and a blank peak value equivalent to 100% transmittance, i n , addition to a series of standard and experimental peak values, is determined, as shown in Figure 8 for a 15-place rotor. After the spectrophotometric rate or stoichiometric measurements are completed, the samples are dumped through drainage siphons located in each cell by application of an air stream to the center cavity of the rotor which extends radially to each cell. The rotor is then stopped, and wash water is added from a wash bottle to both the sample and reagent

ANALYTICAL CHEMISTRY, VOL. 44,NO. 12, OCTOBER 1972

compartments of the transfer disk. The rotor operational sequence is repeated, allowing the wash water to be transferred to the cuvettes which are drained as described abovc. To increase sample throughput rate, rcsearch is being conducted to provide automation of sample and reagent loading and of cleaning and drying of thc rotor between runs. To extend the utility of the analyzer to more complex methodology, two additional research objectives reported (W) arc incorporation of ion-exchange and gel filtration columns into the system and provision for posttransfer self-decanting of supernatant from a precipitate. Also, throwa!vay disks already loaded with reagents are being considcred. When an end-point dct.crminatiou is made, the readout is obtained by averaging the results for a number of rotor passes. For a reaction-rate proccdure, a fixed-time technique is used. Two or more sets of digitized absorbance readings

An entirely new breed of GC The HP 5700 is not just another n e w model but an entirely new k i n d of GC. None can outperform it, even at t w i c e the price. None is more easily or exactly set-up, even b y inexperienced technicians, because of its all-digital design.


for each cuvette are stored simultaneously, and the difference is calculated and expressed as AA/ min. The commercial versions of GeMSAEC are the Centrifichem (Union Carbide, Tarrytoivn, K.Y.), the Rotochcm (American Instrument Co., Silver Spring, l l d . ) , and the GcllSAEC (Electron’ucleonics, Fairfield, K.J.). ELLA. The use of a small computer with a spectrophotometric rate system can be extended to mechanistic investigations by programmed decision-making. Automation of such investigative experiments enhances the collection of data required for the development of new, reliable reaction-rate methods. This approach to automated instrumentation has been incorporated into a system called ELLA (39). The hardlvare used for ELLA (Experimental LIXC Laboratory Analytical System), excluding the computer 1/0 devices, is shown in Figure 9. The reactants are placed 011 the sample tray in the order of desired addition to the system. For example, for an enzymatic experiment the order is buffer, enzyme, buffer, substrate, buffer. The initial volume of enzyme to be aliquoted and incubation time required prior to rate measurement are specified via a teletype by the experimenter in response to the system’s programmed request. The thermostated digital pipet sequentially d r a m reactants through a dip tube

r

38A

ACCUMULATOR BUFFER

from the turntable and into a holding coil. The volumes of solutions to be picked up and delivered are determined by the number of clock pulses received from the accumulator buffer of the computer (DEC PDP-12). These impulses activate a stepping motor which allows delivery of 5 pl from the pipet per pulse. The reactants are expelled from the pipetting system into a thermostated mixing chamber, and the reaction mixture is transferred by vacuum into the flov cell where the spectrophotometric measurement is made. The software used for ELLA permits performance of a kinetic experiment to a dcsired end point, making all necessary decisions and controlling all instrumentation via a special-purpose time-sharing system. This softn-are consists of four subprograms called the initialization routine, the on-line operating system, the updating system, and the cleanup routine. The initialization routine requests the experimenter to enter the volume of enzyme and the incubation time to be used, sets up a blank reading tube and the first series of reaction tubes containing varying amounts of substrate, calibrates the rcactionrate reading system, and activates the sample-reagent introduction system automatically. The on-line operating system takes readings of the rate of reaction, controls the reaction-rate interface, and performs a least-squares fit to the data. When a sufficient

SEQUENCE TIMER

RATE METER


number of data points have been obtained n-hich exhibit an error of estimation below the level programmed by the experimenter, a flag is set indicating the experiment is completed. The updating program runs after each tube to process information obtained from the previous tube and to set up and activate the hardware for the next run. The program performs bookkeeping tasks and calculates the rate. The calculated initial rate is corrected for blank and printed along with other desired information. If the expcriment is not completed, control is returned to the on-line operating system. If the experiment is completed, the cleanup routine is called which stores the data in the file on magnetic tape for successive experiments. The system then prints out this material upon command. Multimode instrumental System. Recently, Deming and Pardue (50) have developed a computercontrolled instrumental system for characterization of chemical reactions. The system consists of four distinct elements: a programmable digital computer (controller), a spectrophotometer used to convert the chemical information into electrical information, an analogto-digital converter, and a device for initiation and control of the experiment. The initiation and control device is composed of a number of electrical and mechanical components which enable the system to introduce into the spectrophotometer cell the proper volume of reagents and diluents. Four stepping motors are used to drive four calibrated syringes which pull the reagents from a stock reservoir and dispense them. The volume of delivered solution is controlled by applying a corresponding number of pulses to the stepping motors. The system is also provided with the ability t o remove the contents of the cell, rinse the cell, and mix the reagents. The automated instrumental system has been evaluated for three types of operations: routine operations, experimental design, and data interpretation. For routine operations, the experiments are preprogrammcd into the information base, and data

Now here is the improved Aquatest II ...

Solids as easy to measure as liquids Much larger sample capacity.. .Accuracy, reproducibility and sensitivity improved to 2 1 0 micrograms of water . . Improved electrode system for longer chemical life and lower operating cost.

,

.

As before.. . always ready for use. Just inject

4 9 PHOTOVOLT CoRpoRAT/oN 4

.

the sample.. . flip the switch.. .step back.. read the answer. It still features digital readout., . no standardization.. . no drying system.. , no burettes., . no measuring.. . no computations. Ask us to demonstrate Photovolt’s AQUA-

TEST 11. 1115 Broadway, N e w Y s r k , N.Y. 10010 * (212) 989-2900



WeWkeyou to turn us on...

are plotted on the storage scope and printed out on the teletype. I n the experimental design operation, the concentration of reagents and the limits over which they must be varied are introduced into the computer. Then the instrument plots the reaction rate vs. concentration of the species. Data interpretation type of operation is accomplished by using the results of precision experiments. These results are interpreted, and new experiments are automatically designed to complete the characterization. The automated instrumental system allow the time required for the operations involved in the experiment preparation to be decreased. The efficiency of the experimenter is greatly increased by the rapid availability of the results of a characterization and by more free time to accomplish other tasks that are more difficult to automate. CONCLU S O N S

matic volumetric dispensing methods will be a thing of the past. Flip the switch on our ADP-30 series AUTOMATIC DIGITAL PIPETTE and compare its unique features to any other dispensing unit on the market. Our ADP-30 dispenses liquids with better than .5% accuracy with a repeatability of 0.1 O/o different fluids simultaneously ery capacity from 0.02 ml. to range of cans, bottl new FLEAKER (Reg. chronized conveyor tr can select the n automatically after f posed to fluids ar DuPontTM)to assur a least count down to 0.02 pl through titrating. The Grumman ADP-30 series is ideal for aseptic filling of biologicals, serums, antigens and pharmaceuticals; for dispensing aqueous solutions, oils, suspensions, liquified ointments and corrosive chemicals. Turn our ADP on. It's built to reciprocate. For a demonstration, call collect (516) 575-3282 or 3888

GRUMMANDATA SVBTEMB GaRBEIRATOON BETHPAGE,NEW YORK

11714

IINSTRUMENT SYSTEMS PRODUCTSI CIRCLE 87 ON READER SERVICE CARD

40A


A new generation of automated and computer-controlled instrumentation makes it possible to perform reaction-rate methods of analysis at times per sample that are equivalent to the fastest equilibrium methods. The rate methods often have advantages in selectivity and accuracy that will encourage their adoption as the new automated instrumentation becomes available in laboratories, and more methods are consequently developed. The present interest in automating basic chemical reaction-rate studies should make it possible to develop specific ncw procedures many times faster than in the recent past. Because of space, no attempt was made to review all of the new automated methods or the well-known Technicon continuous-flow analyzers. The methods presented illustrate that the major differences in chemical reaction-rate analyzers are in the sample and reagenthandling and mixing procedures. The functional units of the instruments are similar, and with suitable modular designs, it should be possible to easily modify analyzers for various general approaches to either reaction-ratc or cquilibrium analytical methods.

At present, only the stoppedflow analyzer has been applied for rapid quantitative procedures based on measurement of initial reaction rates for extremely fast reactions. REFERENCES (1) H. V. Malmstadt and G . P. Hicks, Anal. C h a . , 32,394 (1960). (2) H. V. Malmstadt and H. L. Pardue, ibid.,33,1040(19611. (3) H. V. Malmstadt and S. I. Hadjiioannou, ibid., 34,452 (1962). (4) H. V. Malmstadt and T. P. Hadjiioannou, ibid., p 455. (5) H. V. Malmstadt and H. L. Pardue, Clin. C h a . . 8.606 (1962). ( 6 ) W. J. Blakdkl and G. P. Hicks, Anal. C h a . , 34,388 (1962). (7) G. A. Rechnitz, ibid., 38,513R (1966); 40,455R (1968). ( 8 ) G. G. Guilbault, ibid., 42, 344R (1970). ( 9 ) K. B. Yatsimerskii, “Kinetic Methods of Analysis,” Pergamon Press, New York. N.Y.. 1966. (IO) H.’Mark and G. Rechnits, “Kinetics in Analytical Chemistry,” Wiley, New York, N.Y., 1968. (11) G. G. Guilbault, “Enzymatic Methods of Analysis,” Pergamon Press, Oxford. Eneland. 1970. (12) H. L. Pardue, Rec. Chem. P I O ~27, .,

Howard V. Malmstadt isProfessor of Chem,istry at the l..i,iuersity of Illinois in Urbana. After receiving a BS degree in chemistry f r o m the University of Wisconsin ( 1 94.9) and serving as a radar ojicer in the Pacific, he returned to Wisconsin and earned his MS and P h D degrees in 1948 and 1950. H e jdined the faculty at the

151 (lBfiR1. ...~ ....,

(13) J. D. Ingle, Jr., and S. R. Crouch, Anal. Chem., 4 3 , 7 (1971): (14) (1963). , , H. L. Pardue. ibid..,35.1240 . . , (15) H. y., Malmstadt and E. H. Piepmeier, zbzd., 37,34 (196.5). (16) H. V. Mslmstadt and A. C . Javier, unpublished internal report, 1965. (17) T. Kajihara and B. Hagihara, Rinaho Bum’. 14,322 (1966). (18jKadish, E. Litle, and J. Sternberg, Clin. C h a . . 14. 116 (196x1.

Emil A. Cordos i s aii nssistant professni. in the Aiialytical Division of (21) H. V. Malmstadt, C . Delaney, and E. Cordos, CRC, Critical Rev. Anal. C h a . , 2,559 (1972). (22) H. V. Mdmstadt and G. P. Hicks, Anal. C h a . . 32.44.5 (1960). (23) B. Chance, J . Franklin Inst., 229, 455. fila. 737 (19401. , (24) Q . H. Gibson and L. Milnes, Bioehem. J., 91,161 (1964). (25) E. Cordos, S. R. Crouch, and H. V. Malmstadt, Anal. C h a . , 40, 1812 ~~~

~~~~I

~

~

the University of Cluj, Romania, and i s also a member of the Directing Council of the Center for Research

~I

(14RXI I _.,, . ~

I

(26) H. V. Malmstadt, C. J. Delaney, and E. Cordos, ibid., 44 (121, 79A (1972). (27) N. G. Anderson, Amp. J . Clin. Pathol..,53.77R (19701. , , (28) R. L. Coleman, W. D. Shults, M. T. Kelly, and J. A. Dean, Amcr. Lab., 3, 7.26 (19711. , , , (‘29) A. A. ISggert,

Automated reaction-rate methods of analysis - Analytical Chemistry

Recommend Documents