lnstrumentation
J. L. Glajch J. J. Kirkland E.I. du Pont de Nemours and Company Central Research and Development Department Experimental Station Wilmington, Del. ‘19898
Optimizationof Selectivity in Liquid Chromatography Mobile phase, stationary phase, and pH and other ionic effects are most useful in optimizing an LC separation Use of modern liquid chromatography (LC) has increased so dramatically in the past decade that in many laboratories it is now a routine method for solving many practical analytical and preparative problems. Along with this surge in application, advances have been made in the fundamental understanding of various LC separation mechanisms. Despite this, most routine LC analyses are still being developed in a nonisystematic manner, and very often the results are not as good as might be expected from this powerful technique. Optimizing the separation is an especially important goal for any routine LC method. Although one should , not attempt to optimize all separations-such as one-time-only samples or separations that are trivial to perform-the advantages of an optimum separation for quality control or other repetitive analyses are dear. Increased accuracy and precision, as well as savings in analysis time, often justify the effort spent in dcveloping optimum LC separating conditions. In LC, as with any technique, the primary question is, What should be optimized? Generally, separation quality is the most useful criterion for all components in a mixture or for certain components of interest in the sample. Often of interest is a related 0003-2700/83/0351-319A$O1.50/0 0 1983 American Chemical Society
optimization factor-separation time-especially when a large number of samples is to be analyzed. Sampleloading capacity for the system also can be important in preparative work, and detectability is especially crucial in trace analysis. However, these latter two factors are of less general interest and, therefore, will not be discussed in this treatment.
Effects Influencing Separation
There are two major effects that influence the LC separation mechanism: physical and chemical. Physical effects relate to factors such as column efficiency, generally measured by plate count, N , and column capacity, described by the capacity factor, k’(I). The well-known resolution equation describes these effects:
(-y
R, = 3 B. 4 1+k’
(a- 1)
the retention time of the solute peak, and t o is the retention time of an unretained peak. Both the efficiency and capacity factors of the resolution expression in Equation 1 are well understood from a theoretical standpoint (1).With respect to physical effects, well-packed, small-particle (3-10 pm) columns operated to give solute h’ Values of about 1-10 ( I ) provide excellent resolution. Chemical effects, as described by a, the selectivity factor in Equation 1, are not as well understood, nor are they as predictable as physical effects. It is usually not possible from basic principles to predict the influence of mobile-phase composition or temperature on the relative retention characteristics of different compounds. However, changes in chemical selectivity are the single most important variable leading to adequate separations, as predicted in Equation 1.
(1)
efficiency capacity selectivity factor factor factor where, N is the column plate count calculated by ( t ~ / a k’ ) ~is, the capacity factor calculated by ( t -~t o ) / t o , CY = k’z/k’l for peaks 1 and 2, a is the standard deviation of the peak, t R is
Chemical Selectivity Effects
A large number of operating variables can be used to change the chemical selectivity in an LC system. Although not inclusive, the five most useful variables are mobile-phase composition, stationary-phase composition, temperature, pH or other ionic effects, and secondary chemical equi-
ANALYTICAL CHEMISTRY, VOL. 55, NO. 2, FEBRUARY 1983
319A
(b) 8
X
PH 7
::
6
X 30
"
x
(
(0.333.0.333, 0.333)
3
2 (0. I . 0)
6
( 0 . 0 . 1)
" 35
k
40
45
50
Temperature
(0.5.0. 0.5)
Figure 1. (a)Mixture design experiments used in mobilaphase optimization strategy. (b) Factorial design experiments used for examining temperature and pH effects libria. These five factors are listed roughly in the order of general applicability to all LC systems. although in certain specific instances the latter two parameters can be the most imDortant in Droducinn- the desired separation. Effects of mobile phase on chemical selectivity have been well documented for reversed-phase ( 2 4normal bonded-phase ( 6 ) ,and adsorption (7, 8 ) chromatographic systems. The solvophobic theory of Horvath ( 9 ) strongly suggests that mobile-phase effects are the most important in the separation mechanism of reversedphase chromatography. The systematic approaches initiated by the solvent selectivity triangle concept (IO)have
been instrumental in the development of optimization strategies based on mobile-phase composition. Stationary-phase effects are also important in 1.C separation selectivity, as exhibited by the development of bonded-phase packings with a variety of functional groups. especially for reversed-phase separations. Recent studies ( 1 1 , 1 2 ) have adequately demonstrated these stationary-phase effects. However, the number of stationary phases needed in 1.C is relatively small compared t u those used in gas chromatography, due in part to the fact that selectivity effects provided by changing the mobile phase are not significant in gas chromatography. Changes in temperature have not
been used extensively to change selectivity in LC. However, Snyder (13) and others ( 1 4 ) have pointed out the potential for changing separation selectivity by varying temperature. primarily because of effects on the extent of ionization with certain molecules. Optimization uf pH and other ionic effects has also been carried out by simple statistical approaches and shown to be quite useful in specific analyses ( 1 5 ) . Secondary chemical equilibrium effects have also received increased attention (16).This approach usually involves addition of a small amount of a particular selective agent into the mobile phase to alter the chemical equilibria that exist for solutes in the mobile and stationary phases.
Prulon Acceptw
n
Proton m o r
nn
-
-
Dipole lnleractwn RweMPha NDTmal P h w
pure 3. Preferred modifying solvents lor use in reversedand normal-phase LC optimization. X.. X,. and X. are seleclivity effectsfor proton acceptor, proton donor, and dipole interaction. respectively. Adapted from Reference 5. with per-
mission 320 A
ANALYTICAL CHEMISTRY. VOL. 55. NO. 2. FEBRUARY 1983
m re 4 PlUl Switches LC columns without wrenct I
W
W
~
Rheodyne's newest valves make it unnecessary to disconnect the plumbing when you want to sub13m ,n,er.or stitute one LC column for another. You need only turn a pair of tandem valves to select any column from as Switching columns frequently is becoming a generally-accepted practice for several reasons. Often, you must switch to a different column to make a different analysis. But. even when switching columns is not
W
.-.
L
essential, dedicating a separate column to each analysis eliminates column equilibration delays, reduces interferences,and prolongs column life. Rheodynek Model 7066 Tandem Column Selector connects any of up to five columns (or a bypass line) between the sample injector and the detector. Off-line columns are sealed at both ends, and no column need ever be exposed to a solvent intended for another. OurTechnical Notes 4 tells all about column seleclion with Model 7066 and other valves. To receive a copy, along with product literature, contact Rheodyne, Inc., PO. Box 966. Cotati California 94928, U.S.A. Phone (707) 664-9050.
Optimization Strategies 1-2
5-6
2-3
6-7
A
Rs
/
3-4
\
7-8
4-5
lure 4. Resolution maps for all eight peak pairs in a mixture of substituted naphilenes. White areas represent solvent mixtures that will result in R. > 1.5. Rented with permission from Reference 5
/Optimum
ACN
THF
Figure 5. Overlapping resolution map (ORM) obtained from resolution maps in Figure 4 322A
ANALYTICAL CHEMISTRY, VOL. 55. NO. 2, FEBRUARY 1983
To optimize any LC system, three factors must first he specified 1)the type of variables important in effecting the separation, 2) the optimization strategy that is the most appropriate to those variables, and 3) an accurate method to measure the performance of the separation for the quantitative comparison of different separation approaches. The chemical parameters described above that are the most important in LC can be conveniently grouped into two categoriesrelated variables and discrete variables. Related variables are those that by their nature directly affect each other. For example, mobile-phase composition must total 100%for all components at all times, so individual mobile-phase solvents are related variables. A similar situation exists for stationary-phase composition. By contrast, temperature, pH, and secondary equilibrium effects can he considered discrete variables since they have little direct effect on each other or on the other mobile-phase or stationary-phase components. It is possible, for example, to carry out LC separations in a temperature range of 30-70° C and an apparent pH range of 3-8, with all possible combinations of these two variahles. On the other hand, it is not possible to carry out separations with methanol ranging from 0-100% with acetonitrile also siniultaneously ranging from 0-100% in the mobile phase. Definition of these two types of variables provides two distinct optimization strategies of experimental design. The mixture design statistical approach for mobile-phase optimization is illustrated in Figure la. This approach is well-known in the statistical literature (1 7). Seven experimenta are employed to fit experimental retention (k') data to a second-order polynomial equation with respect to the three mobile-phase modifiers. In the case of reversed-phase LC, the mobile-phase carrier, water, is modified with methanol, acetonitrile, and tetrahydrofuran (5).Figure l b illustrates a factorial design for combined temperature and pH effects in LC that employs nine experiments to fit the h' data to an equivalent secondorder equation. It is also possible to combine the mixture design and factorial approaches for both related and discrete variables such as mohilephase composition, pH, and secondary equilibrium effects. However, in many cases. adequate separation selectivity can he obtained hy utilizing only organic mobile-phase modifier effects. This approach is usually simpler and more easily predicted than the other chemical effects altering selectivity.
How Many Pumps DoVou Need? IY:(11I Be the IIIIIW ewwuv.
Some HPLC companies want you to buy a multi-pump gradient system because that's what they make. Other companies insist that you buy a single-pump gradient &! system because that's what they make. You alone are the best judge of what &&!& should be in your lab. Hitachi makes both, so take a look, and take your pick. ~~~
MODEL 655 GRADIENT CONTROLLER
I "*
OUT
COntrd'er in
nbination with either pumping system. I
Includes a built-in thermal printerlplotter multiple method storage, and memory protection. It completely controls your HPLC-flow rates, solvent composition, column temperature, pressure limitseverything.
\I
Single-pump gradient system Select our low cost, single pump, ternary gradient system with solvent proportioning valves for highest precision when using a low percentageof a mobile phase comment.
Both systems are modular, so you can buy what you need when you need it. And both systems are built
down to 10 ht/min) version. It is easily upgradable, with no costly modifications necessary. Both systems are
,
Select a multi-pumpsystem when ability to spllt the system into two isocratic systems is desirable.
ty is our most imporlent
I
for
Chromatography Active Silicas Active Aluminas for
cc
analytical preparative pilot plant TU Dry-Column C
HPLC high & med. pressure Available in Australla
H. B. Selby & Co. Pty. Ltd. 352-368 Ferntree Gully Road Notting Hill Victoria Tel.: 5444844
Austria
C. Richter & Co. K.G. Feldgasse 19 4600 Wels Tel.: 4404
Belgium
N. V. Pieuger S. A. Turnhoutsebaan 511 21 10 Wijnegem Tel.: 538551
Canada
Universal Scientific, Inc. 2070 Peachtree Industrial Court Atlanta, Georgia 3034 1 Phone: 404-455- 1140
Denmark
MEDA AS Linde All6 48 2720 Vanlase Tel.: 744400
Finland
Lllketukku Oy Vattuniemenkuja I Helsinki 20 Tel.: 6731 91
France
Prolabo 12, rue Pelbe 75011 Paris Tel.: (1) 355-44-88
Germany
Woelm Pharma GmbH & Co. D-3440 Eschwege Tel. : 0 56 51-65 81 Great Britain Koch-Light Laboratories Ltd. Haverhill, Suffolk CB9 8PU Tel. : 0440-2436i7 India Mis. Sarabhai M. Chemicals P.O. No. 89 Gorwa Road Baroda-390007 Tel.: 8621 Israel M. Kochn 12, Shamai Str. Jerusalem Tel.: 2228 62 Dr. Amato & Figii r.s.i. Italy Via Livorno 25 00162 Rome Tel.: 420506 Wako Pure Chemical Japan Industries, Ltd. 10 Doshomachi 3-chome Higashi-ku Osaka 541 Tel.: (06) 203-3741 General Scientific Company Malaysia SDN. BHD. 7 Jalan 222 Section 10 A, Petalina - Java . Seiangor Tel.: 775433
GmbH & Co. D-3440 Eschwege West-Germany Tel. 0565118581 CIRCLE 341 ON READER SERViCE CARD,
Mexico
Netherlands
New Zealand
Portugal
Sweden
Swltserland
USA
Universal Scientific, Inc. 2070 Peachtree Industrial Court Atlanta, Georgia 3034 1 Phone: 404-455- 1140 Pleuger Nederland Postbus 44 Amstelveen 1134 Tel.: 433651 Selby-Wiiton Scientific, Ltd. 410 Hutt Road P.O. Box 30556 Lower Hutt Tel.: 697099 Concessus S.A.R.L. Apartado 1455 Lisboa 1 Tel.: PPC 652406/07 Flow Laboratories Svenska AB tjstervagen 3 17123 Solna Tel.: 08i7305575 Instrumenten-Gesellschaft AG RaffelstraOe 32 8045 Zurich Tei.: 663311 Universal Scientific, Inc. 2070 Peachtree Industrial Court Atlanta, Georgia 3034 7 Phone: 404-455- 1140
SIEMENS
New norizons in GL
Take the twinTovenversion of the SiCHROMAT? and you have two gas chromatographs for little more than the price of one - a laboratory tool that does far more than just double your analytical With it, you can N n two working temperatures in the same device, in three different wi - bothchambers equally or differer isothermic -one chamber with a temperature program, the other isothermic , -both chambers with a temnerature.. program 0
add a new dimension to w I u I I I I I:.~. switching, e.g. using cotuhnsof different polaritie8at different :: temperaturesfor hah ~
~
~~~
detection limits double the column out the pactition, f~ columns and a total
with the twin-ovei model of the Siemens SiCHRClMd CIRCLE 288 ON READER SERVlcE
,
Optimization in LC differs from approaches used for other analytical methods in that data are taken a t specified values of the key variables, fitted to a polynomial equation (generally second-order), and interpolated to search for an optimum response region. Such a simple method is feasible since the response of LC separations to variable changes is usually continuous and regular. Other analytical techniques (e.g., atomic spectroscopy) rely on different optimization routines (e.g., simplex algorithms) to handle
data that are less regular in form. Perhaps the most crucial step in optimization is deciding upon a measure of system performance. Various methods have been proposed to assign a single value of separation quality to a chromatogram (18).However, these methods have the inherent disadvantage that much of the chromatographic information is either lost or ohscured. An extension of the “windowdiagram’’ approach (19) based on minimum separation factor plots has been used (20-22) to provide information
Figure 6. Experimental design for combined mobiikphase and stationary-pha& Optimization
on separation performance. However, this approach can be limited since iuformation based on a (at constant column plate count) does not take into account the important influence of the capacity factor k’ values. We prefer to use the resolution function (Rs)as a separation performance criterion, since this is directly .related .to the actual separation required. A representation of the difference in performance information provided by a and Rs is shown in Figure 2. A column of plate count = 10 000 and a = 1.04 for two components would give an R8 = 1.8 at k’ = 10. However, the same a value would provide only R, = 1.0 at k‘ = 1.Thus, practical separation performance is inadequately described by a values alone, since the ability to separate one component from another is a function of how long these compounds are retained on the column as measured by capacity factor k‘ values, and peak sharpness as measured by column plate count values, N . Previous use of a values to measure separation quality is probably due, in part, to the simplicity of having to measure only k’ (or corrected retention time) values, rather than having to determine both the retention time and peak widths that are required for calculating accurate resolution values (see Equation 1). However. measurement of Ra can be simplified by using the expression (23)
Re = [N1’2/2][(kz- ki)/ (2 + ki
+ kz)l
(2)
Equation 2 can be rigorously derived from the fundamental resolution expression (Equation 1)by making the reasonable first-order assumption of constant plate count for all peaks in an LC chromatogram (I ). We currently use Equation 2 for all our isocratic mobile-phase optimization calculations. Once the separation variables, optimization strategy, and system performance criteria have been selected, one can quickly define the proper set of experimental conditions necessary for an optimum separation. Examples of various LC optimization approaches will now be described to illustrate the potential for each method.
Time (min)
Figure 7. lsocratic separation of 20 PTKamino acids. Experimental details in Reference 26 926A
ANALYTICAL CHEMISTRY, VOL. 55, NO. 2, FEBRUARY 1983
Mobile-Phase Effects Several studies have shown that optimization of mobile-phase selectivity is prohably the single most useful effect in LC. One of the first systematic studies of mobile-phase optimization was illustrated by the reversed-phase separation of nine substituted naphthalenes on a C-8 hydrocarbon bonded-phase column (5).To obtain
EbuNeedA LiquiidChrom.abgrapl1 Chamber, ThenYou NeedA
uia,
Revco's new top-of-the-line Chromatography Chambers deliver a dependable 4°C temperakure uniformity and have these standard features: .Four Temperature Controls for Maximurn Safety and Uniformity. *Fouslkmperature Alarms for Complete Protection. *ContinuousTemperature Display in All Quadrants. .An Alarm System that Automatically Tests Itself Every 24 Hours. *Audibleand Visual Alarms for Low Voltage or Power Failure, Door Ajar or Unsafe lkmperature Conditions.
*AlarmDelays with Automatic and Manual Re-set Modes. *ComputerConnections for Automatically Collecting, Storing and Monitoring lkmperature Data and Test Functions. *TheIndustry's Only 13Month Full Warranty on both Parts and Labor. All Revco Chromatography Chambers are equipped for optionslike these: an Auto-Dialer Alarm System (it will
call your home or nine other numbers), a 240 Day Electronic Temperature Recorder, and a Remote Audio Alarm.
Available in three sizes for convenient
Rheem Manufacturing CompanyiScientific Products Division, Aiken Road, Box 275, Asheville, N.C. 28804. Tel.: 7041658-2711,Telex: 67-7449. CIRCLE 276 ON READER SERVICE CARD
ANALYTiCAL CHEMISTRY, VOL. 55, NO. 2, FEBRUARY 1983
327A
useful selectivity changes leading to improved resolution, modifiers of the carrier mobile-phase solvent were selected using the Snyder solvent selectivity triangle approach (IO). This organization of solvent selectivity describes the major effects of proton acceptor, proton donor, and dipole interactions of solvents with compounds to be separated. Preferred modifying solvents for both reversed- and normal handed-phase systems are shown in Figure 3. In reversed-phase LC, the carrier solvent is water, while n-hexane or 1,1,2-trifluorotrichloroethane (FC-113) (24) is the preferred carrier solvent in normal-phase systems. Methanol, acetonitrile, and tetrahydrofuran are effective proton acceptor, proton donor, and dipole solvents, respectively, for reversed-phase separations, while methyl tert-butyl ether, chloroform, and methylene chloride are effective counterparts in normal handed-phase LC. The experimental design used to select the optimum solvent system (Figure la) is based on the report of Snee (25).Retention (k') values are measured for each compound in the mixture using the seven mobile-phase solvent mixtures depicted in Figure la. These data are then fitted to a secondorder polynomial to obtain retention or k' value maps for each solute. Usii Equation 2, the resolution of each peak pair in the system also can he mapped in this way. Once the resolution maps are obtained, the solvent compositions that allow a certain minimum resolution can be determined for each peak pair. Figure 4 shows this for eight peak pairs of Reference 5, where the white areas in each small triangle represent solvent compositions that will resolve the particular peak pair to a resolution of at least 1.50. These resolution maps of all constraining peak pairs are then overlaid and the resulting overlapping resolution map (ORM) can he used to define the optimum solvent mixture (or solvent mixture region) to obtain a certain minimum resolution for all peak pairs in the mixture, as shown in Figure 5. In this case, the optimum mobile phase corresponding to the x in the figure is 32% acetonitrile, 16%tetrahydrofuran, and 52% water. This produces a minimum resolution of 2.5 for all peaks in the mixture. Stationary-Phase Effects
Optimization of bonded stationary phases in LC can he carried out in a manner similar to that for mobile phases, since both phases involve related or dependent variables (Le., the sum of all components of a stationaryphase system must also total 100%). Differences in selectivity are usually 328A
a
Phenylacetic Acid
T
oCH2-cmH
Cinnamic Acid
A
Figure 8. Retention lime maps lor phenylacetic acid and cinnamic acid as a function of pH and Ion-interaction reagent IIIR). Adapted from Reference 21. with permission
not as great for the common chemicalIy bonded stationary phases as for mobile phases. However, the experimental design just described can he used as well for stationary phases in LC columns to provide selectivity changes over and above those possible with only mobile-phase optimization. An example of this approach involves the use of both mobile- and stationaryphase effects for optimizing the iso. cratic separation of the 20 common phenylthiohydantoin (PTH) derivatives of amino acids (26). The experimental design uses the strategy shown
ANALYTICAL CHEMISTRY, VOC. 55, NO. 2, FEBRUARY 1983
in Figure 6 to examine the selectivity effects of three mobile-phase modifiers in an acidic (pH = 2) aqueous solution together with three different stationary bonded phases. A total of 21 experiments (seven mobile phases for each of the three stationary phases) is required. The calculated optimum column and mohile-phase composition predicted by the ORM technique produced the chromatogram in Figure 7. This separation was completed in 22 min with a minimum resolution of 1.1 for all components, using a single mobile phase. This separation time com-
I
The ChromiumMechaylism Thefirst comprehensive explanation of electrochemical activity during the plating of chromium has recently been formulated at the General Motm Research Laboratories. This understanding has aided in transformingchromium plating into a highly e f i e n t high-speed operation.
Complex Concentration
7
0 f
v-H-0-5-0-
0-Cr-0-Cr-0-Cr-OH
I
0
N
0
I
8 7
o-H-O-s-0-
F
OR MANY industrial applications, chromium coatings of more than 0.2 mil thickness are required for wear and corrosion resistance. But the conventional method of plating chromium is neither fast nor efficient. Nor, until the recent work of a GM researcher, had the steps involved in the century-old plating process been explained in detail. Through a combination of theory and experiment, Dr. James Hoare has devised the first comprehensive mechanism for chromium plating. This increased understanding has helped electrochemists at the General Motors Research Laboratories develop a system that plates chromium sixty times faster than the conventional method, while improving energy-efficiency by a factor of three. The electrolyte for plating is
a chromic acid solution which contains various chromate ions: chromate, dichromate and trichromate. From a series of steady-state polarization experiments, Dr. Hoare concluded that trichromate is the ion important in chromium deposition. Sulfuric acid has been recog nized as essential to chromium plating and has been assumed by some to be a catalyst for the process. In this strongly acidic solution, sulfate should be mostly present asthe bisulfate ion (HSOJ. Dr. Hoare found, contrary to expectations, that the addition of sulfuric acid to the plating bath decreased the conductivity of the solution. Combining these findings with the results of previous investigations, Dr. Hoare concluded that the electroactive species was a trichromate-bisulfatecomplex (see Figure 1). From equilibrium considerations, he theorized that the maximum concentration of this species occurred at a 100-to-1chromic acid/sulfuric acid ratio. The observation that the maximum rate of chromium deposition also occurred at this ratio supports the conclusion that this trichromatebisulfate complex is the electroactive species. During the plating process, the complex diffuses from the bulk solution toward the cathode (see Figure 2). Electron transport takes place by quantum mechanical tunneling through the potential energy barrier of the Helmholtz double layer and the unprotected chromium in the complex (Cr atom
on the left in Figure 1) loses electons by successive steps, going from Cr+6to Cr+? Decomposition of the resulting chromous dichromate complex takes place by acid hydrolysis to form a chromousoxybisulfate complex: 0 ‘Cr-0-H-0-S-0-
II
6
The positive end of this complex is adsorbed onto the cathode surface. Electrons are transferred from the cathode to the adsorbed chromium ion, forming metallic chromium and regenerating the (HS04)- ion. Thus, Dr. Hoare’s mechanism explains how sulfuric acid, in the form of the bisulfate ion, participates in the plating process.
I
shortening the diffusion path increases the speed of chromium deposition. A high rate of relative motion between the electrolyte and the cathode will shorten the path. This can be accomplished by rapid flow or by agitation of the electrolyte. Dr. Hoare found that the rate of chromium deposition increased with electrolyte flow until the process was no longer diffusion-controlled. He also found that the use of dilute electrolyte significantly increased plating efficiency. “This project is an excellent example,”says Dr. Hoare, “of how basic research and engineering principles can be combined to develop a new, successful process. Now, we’d like to take on the challenge of plating successfully from Cr ‘3, which would be an even more efficient way to provide corrosion and wear resistance.”
THAS long been known that
chromium cannot be plated from a solution when initially present as Cr+3because of the formation of the stable aquo complex, [Cr(H20)J+4Yet chromium can be plated when initially present as Cr+6 even though it must pass through the Cr +%tatebefore being deposited. Dr. Hoare’s mechanism handles this paradox by explaining that the chromium ion being deposited (on the left in Figure 1)is protected by the rest of the complex as it passes through the Cr + 3 state, so that the stable aquo complex cannot form. The diffusion of the electroactive complex apparently controls the rate of the process, so that
THE MAN BEHIND THE WORK Dr. James Hoare is a Research Fellow a t t h e General Motors Research Laboratories. He is a member of the Electrochemistry Department. Dr. Hoare served as an electronics technician in the US. Navy during the Second World War. In 1949, he received his Ph.D. in physical chemistry from the Catholic University of America. After an assistant professorship at Trinity College in Washington, D.C., he joined theUS Naval Research L a b oratory as a physical chemist. He became a staff member at General Motors in 1960. Dr. Hoare’s sustaining interest has been in electrochemical kinetics and the mechanisms of electrode processes. He is best known to the scientific community for his basic studies of hydrogen and oxygen electrode mechanisms. His book, The Elecfrochanistty of Oxygen, published in 1968, is considered a work of primary importance to the field. In addition to his work on chromium plating, he is responsible for the fundamental research that helped make electrochemical machining a precision process. ~~~~~~~~~~
~~~~
~CIRCLE 128 ON R E A M SERVICE CARD
pares favorably with current gradient elution analysis methodology; however, resolution by the optimized isocratic system is superior. There are many other advantages of an isocratic separation relative to gradient elution, such as simpler equipment, improved quantitation, and the lack of need for reequilibrating the column with the initial mobile phase.
LDC/Milton Roy offen a complete familj of precision metering pumps for high performance liquid chromatography IHPLC). miniPump@seriesmetering pumps . The first metering pump manufactured and used for HPLC. Priced from $975.00. Single Piston, 0-6000 psi, solvent delivery from 16-920 mL/hr, flow rate stability f 0.3% per hour. miniMetric"series metering Pumps . The newest addition to the LDC/Milton Roy family of metering pumps. Priced from $1752: Fast-refill single sapphire piston. 0.6000 psi, pulseless solvent delivery from 50 .IL - 5 mL/min, flow rate stability f 0.3% per hour, gradient compatible.
.
ConmMebicO series metering pumps . The ultimate in pulseless flow performance. Priced from 52292.' Dual reciprocating sapphire pistons, 0-6000 psi, pulseless solvent delivery from 0.1 to 9.99 mL/min. flow rate stability f 0.3% per hour, gradient compatible.
'When purchased with agradient system.
For more information or to order call: Toll Free: 800-3274182 In Florida: (305) 844-5241 LOC/Milton Roy Analytical Instruments - a tradition of qualii and performance.
LDC/Milton Roy 3661 Interstate Park Ad. N. P.O. Box 10235 Aiviera Beach, F L 33404
*
d
pH and Ionic Effects The use of pH and other ionic effects can be quite helpful in modifying separation selectivity, particularly in reversed-phase systems. Optimized separation of a nine-component mixture of weak acids, weak bases, and zwitterionic compounds bas been demonstrated by this approach (21). In this case, the pH and concentration of octylamine hydrochloride ion-interaction agent were used as discrete variables, and a four-level two-factor (42) experimental design was employed to specify 16 mobile-phase compositions. These consisted of all combinations of four pH values (3.6, 4.4,5.2, and 6.0) and four concentrations of octylamine hydrochloride (0, 1.5,3.0. and 5.0 mM). From separations with these mobile phases, retention time maps were obtained for all nine species as a function of both pH and concentration of octylamine hydrochloride. Two of these maps are shown in Figure 8. A window diagram was used to determine the optimum set of experimental conditions. The resulting optimum mobile phase separated all nine components in 45 min a t pH 3.7 and 0.75 mM octylamine hydrochloride. A similar experimental design has been used to examine the effects of mobile-phase strength, pH. concentration of a phosphate buffer, End concentration of camphorsulfonic acid ion-pairing agent for the separation of four alkaloids ( 1 5 ) . Other Effects Although not as generally applicable, changes in temperature also have been used to achieve selectivity optimization (13).Changes in relative retention values are often observed for ionizable compounds. Temperature optimization also can be combined with other effects, such as mobilephase and stationary-phase characteristics as well as pH effects, for more comprehensive optimization possibilities. Use of secondary chemical equilibrium (16) to affect selectivity could also be utilized in optimization schemes.
ANALYTICAL CHEMISTRY. VOI
Identify virtually any alloy
commonly used by man-steels, nickel alloys, bronzes, aluminum alloys, plated melal coatings-
with the Koslow #1899 Alloy ID Kit. You obtain results ... ON SITE. This 10-lb. lab-in-asuitcase can be carried to the mill, stockroom, warehouse, machine shop, QA lab, construction site, scrap heap. Anywhere metals are used. You no longer have to take your sample to t h e lab. It's now easier to carry t h e lab to your work. OUICKLY. Identification by chemical spot testing can be done in less than a minute per element. On raw stock. Finished parts. On scrap headed for recycling. And the results are positive and unaffected by the shape, size or physical condition of the metal. EASILY. Comprehensive, stepbystep instructions are easy to follow. Kit materials are clearly identified. Even a nontechnical beginner can learn to use the #I899 Kit in just a few hours. INEXPENSIVELY. The ME99 COR-
tains sufficient materials for literally hundreds of tests at just pennies per test. It is t h e most costleffective metal alloy identification kit available anywhere. For full details on the versatile #I899 or any of our specific ID kits, contact Donna at Koslow Scientific Company, 75 Gor e Road, Edgewater. NJ 07028 (201) 941-4484
Conclusion Optimization in LC can be very powerful if care is taken to choose the proper parameters for change, and if a valid experimental design is selected. (continued on p. 336 A )
CIRCLE 180 ON READER SERVICE CARD
332 A
IDENTIFY 98% OF THE MOST COMMONLY USED METALS
NO. 2. FEBRUARY 1983
CIRCLE 175 ON READER SERVICE CARD
SAF-TDATK System Putting chemical safety information where it belongs. Chemical safety information belongs on the bottle where it can easily be identified by anyone who handles hazardous substances. Because of this, J.T. Baker has developed a new concept in chemical labeling-the BAKER SAF-T-DATATMSystem. Each SAF-T-DATA label indicates the potential hazards of the substance, along with recommendations for storage and laboratory protective equipment. The system is a
major part of J.T. Baker's continuing commitment to the health and safety of the people who use our laboratory reagents. The SAF-T-DATA labels will begin appearing on your shelves as they are phased into our production and distributor inventories. You owe it to yourself and your employees to include SAF-T-DATA in your occupational safety program. To learn more, call or write today for our six-page BAKER SAF-T-DATA brochure.
J.TW
J.T. Baker Chemical Co
wn
222 Red School Lane Phillipsburg NJ 08865 (201) 8592151
Baker Chemikalien: 6080 Gross Gerau. W. Germany, Postlach 1661, Tel. (06152) 710371, Telex 04191113 J.T. Baker Chemicals B.V.: Rijslerborgherweg M ,74M) AA, Devenler. Holland, P.O. BOX 1, Tel. (05700) 11341, Telex 49072
'BAKER ANALYZED" Reagent
I'
WARNING!
~
CIRCLE 40 ON READER SERVICE CARD
Introd-
theBooks that Order Themselves! With American Chemical Society’s PACT plan, you’ll never have to order a book again!
We have a plan for book purchasing that can make your job a lot easier. It’s called PACTPrimary Access to Chemistry Titles. And it’s easy. Just tell us your chemical subject categories, and we’ll send you all the new ACS books in those areas . . . just as soon as they’re published.
EnvironmentallHealth & Safety Food/Flavor IndustrialiEngineering Inorganic Medicinal/ Pharmaceutical
Nuclear Organic PolymerslApplied Polymer Science Physical
Know what that means? No more endless paperwork as you order individual books. No more disappointing “out-of-stock”notices. No more wading through scores of book reviews just to find the titles that are relevant to you.
And then be assured of prompt, automatic delivery throughout the year of select books as soon as they’re published . . . of just those books that perfectly suit your needs.
Just one purchase order. . . and the satisfaction of knowing your collections are complete.
No-Worry Guarantee Your PACT is flexible, too. If at any point you wish to cancel or modify this plan, just say the word. No strings attached! And any book sent to you may be returned for full credit within thirty days.
Just Choose Your Subjects Pick one or all of the following subject categories in chemistry from the ACS Advances Series, Symposium Series and Monograph Series: Analytical AgricultureiPesticide Carbohydrate
I I I I I I I I
334A
Cellulose/Paper/Textile Colloidhrface Energy/Fuel/Petroleuml Geochemistry
Want to know more?Call us toll free at (800) 424-6747. Or clip the coupon and send it in today.
Mail to PACT-Primary Access to Chemistr Titles American Chemical Society Sales Ohce 1155 Sixteenth Street, N.W., Washington, D.C. 20036
YES!I want further information about no-hassle book purchasing through PACT. Name Organization Address
ANALYTICAL CHEMISTRY, VOL. 55, NO. 2, FEBRUARY 1983
Business Phone
I I I I I I I I
With the'BAKER-loTMExtraction System, you prepare your samples 6 times faster than tradit liquid-liquid techniques!
Solid-Phase Extraction Makes it Faster Using a solid-phase extraction technique, sample solutions are drawn through prepacked columns which selectively extract compounds of interest. Elution is accomplished with
