enhancement of detection sensitivity for indirect photometric detection of anions and cations in...

Review

Cameron JohnsMiroslav MackaPaul R. Haddad

Australian Centre for Researchon Separation Science,School of Chemistry,University of Tasmania,Hobart, Tasmania, Australia

Enhancement of detection sensitivity for indirectphotometric detection of anions and cations incapillary electrophoresis

This review focuses on the indirect photometric detection of anions and cations bycapillary electrophoresis. Special emphasis has been placed on the sensitivity of thetechnique and approaches taken to enhance detection limits. Theoretical conside-rations and requirements have been discussed, including buffering, detection sensi-tivity, separation of cations, and detector linearity. A series of tables detailing highlyabsorbing probes and the conditions of their use for indirect photometric detectionare included.

Keywords: Buffering / Capillary electrophoresis / Detector linearity / Indirect photometric detec-tion / Review DOI 10.1002/elps.200305446

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 21501.1 Scope of review . . . . . . . . . . . . . . . . . . . . . . . 21501.2 Previous reviews in the field . . . . . . . . . . . . . 21502 Theory and background . . . . . . . . . . . . . . . . 21512.1 The Kohlrausch regulating function and

transfer ratio . . . . . . . . . . . . . . . . . . . . . . . . . . 21512.2 Peak shapes . . . . . . . . . . . . . . . . . . . . . . . . . . 21522.3 Sensitivity of indirect photometric detection 21522.3.1 Baseline noise . . . . . . . . . . . . . . . . . . . . . . . . 21532.3.2 Detection pathlength . . . . . . . . . . . . . . . . . . . 21532.3.3 Probe absorptivity . . . . . . . . . . . . . . . . . . . . . 21532.4 Buffering of electrolytes for indirect

photometric detection . . . . . . . . . . . . . . . . . . 21542.5 Detector linearity . . . . . . . . . . . . . . . . . . . . . . 21552.6 Highly absorbing probes . . . . . . . . . . . . . . . . 21562.6.1 Use of anionic dyes as probes . . . . . . . . . . . 21562.6.2 Use of cationic dyes as probes . . . . . . . . . . . 21572.7 Separation and indirect photometric

detection of cations . . . . . . . . . . . . . . . . . . . . 21572.8 Simultaneous separation and indirect

photometric detection of anions and cations 21583 Tabular summary of the literature . . . . . . . . . 21594 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 21655 References . . . . . . . . . . . . . . . . . . . . . . . . . . . 2165

1 Introduction

1.1 Scope of review

The detection of analytes in capillary electrophoresis (CE)is usually performed by direct photometric detectionwhich involves monitoring a wavelength at which analytesabsorb. The major restriction of this approach is that notall analytes will be sufficiently highly absorbing at theselected wavelength. Indirect photometric detection isan effective alternative detection technique for suchcations and anions which lack a suitable chromophore.Unlike direct detection this is a universal, nonselectivedetection mode in which analytes are detected by theirlack of absorbance. A strongly absorbing ion, typicallycalled the probe, is added to the background electrolyte(BGE). The probe is displaced by analyte ions of the samecharge as the probe, resulting in a decrease in the back-ground absorbance. While indirect detection has a num-ber of advantages over direct detection, including its uni-versality, the sensitivity and the resulting detection limit ofthe technique remains its major restriction. The focus ofthis review is on factors relating to the detection sensitiv-ity and the resulting concentration detection limits for in-direct detection, together with previous approaches thathave been used to improve sensitivity.

1.2 Previous reviews in the field

A small number of reviews in the area of indirect photo-metric detection have appeared. Indirect detection hasbeen reviewed as a subarea of general detection methodsfor CE and particularly when photometric detection isemphasized [1–4]. Doble and Haddad [5] presented a

Correspondence: Prof. Paul Haddad, University of Tasmania,School of Chemistry, Private Bag 75, Hobart, TAS 7001 AustraliaE-mail: [email protected]: +61-3-6226-2858

Abbreviations: HIBA, 2-hydroxyisobutyric acid; ITS, potassiumindigotetrasulfonate; KRF, Kohlrausch regulating function; LED,light-emitting diode

2150 Electrophoresis 2003, 24, 2150–2167

2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 0173-0835/03/12-1306–2150 $17.50�.50/0

Electrophoresis 2003, 24, 2150–2167 Indirect photometric detection in CE 2151

comprehensive review on indirect detection of anions inCE. Factors relating to limits of detection, buffering, anddetermination of inorganic ions, organic acids, carbohy-drates, phosphates and phosphonates were discussedin detail. In the present review this topic is updated andexpanded to include the indirect photometric detectionof cations, with special emphasis on the limits of detec-tion, detector linearity, and buffering aspects.

2 Theory and background

2.1 The Kohlrausch regulating function andtransfer ratio

Transfer ratio (R) is defined as the number of moles ofprobe displaced by one mole of analyte ions. On first in-spection, it would be expected that this displacementprocess would occur on an equivalent-per-equivalentbasis. For example, R between a singly charged analyteand singly charged probe would be one, between a dou-bly charged analyte and singly charged probe the R wouldbe two, and so on. If this situation occurred in practicethe peak areas of all analytes of the same charge and con-centration would be the same. However, this does nothold true. Transfer ratios depend not only on the chargesof the probe and analyte, but also on their electrophoreticmobilities. Ackermans et al. [6] showed a nonlinear rela-tionship between mobility and peak areas which can berationalized using Kohlrausch’s regulating function (KRF)[7]

� ��

i

cizi

�i� constant (1)

where ci, zi, and �i are the ionic concentrations, the abso-lute values of the charge, and the absolute values of theeffective mobilities of all ionic constituents, respectively.

One single � function will dictate the movement of ionsin a capillary filled with a uniform electrolyte. Two � func-tions will determine the movement of ions if a sample plugcontaining one analyte is introduced, the first being asso-ciated with the sample plug, and the second with the bulkelectrolyte. Both � functions must be equal and constant,therefore the concentration distributions of the ions forthe bulk electrolyte and the sample plug remain as theywere prior to the application of voltage. Therefore, theflux of ions into the sample plug is exactly equal to theflux of ions out of the sample plug. It follows that in orderfor this to occur, R is dependent on the mobilities of theprobe, the analyte, and the counterions. This relationshipcan be derived directly from the � function [9, 10].

Let an electrolyte consist of a single ion A, and its corre-sponding counterion, C. From Eq. (1) the � function de-scribing the electrolyte is:

�1 � cAzA

�1� cczc

�c(2)

where cA and cc are the concentrations of A and C in theBGE.

For the electroneutrality condition to hold true:

cAzA � cCzC (3)

� �1 � cAzA1�A

� 1�C

� �� cAzA

�A�C�A � �C� � (4)

Now consider injection of an ionic analyte, BC, disso-ciated into a co-ion B, and counterion C. The samplezone will then consist of A, B, and C. From Eq. (1) the �function describing the sample zone is:

�2 � c�AzA

�A� cBzB

�B� c�

czC

�C(5)

where c�A and c�

C are the concentrations of A and C in thesample zone.

For the electroneutrality condition to hold true:

c�AzA � cBzB � c�

CzC (6)

� �2 � c�AzA

�A� cBzB

�B� c�

AzA

�C� cBzB

�C(7)

� c�AzA

1�A

� 1�C

� �� cBzB

1�B

� 1�C

� �(8)

Now �1 � �2

� cAzA1�A

� 1�C

� �� c�

AzA1�A

� 1�C

� �� cBzB

1�B

� 1�C

� �(9)

cA � c�A

� �zA

1�A

� 1�C

� �� cBzB

1�B

� 1�C

� �(10)

Let �cA = cA � c’A

��cA

cB� zB

zA

1�B

� 1�C

� �

1�A

� 1�C

� � (11)

� zB

zA

�B � �C� ��A � �C� �

�A�C

�B�C(12)

� zB

zA

�A

�B

�B � �C� ��A � �C� � � R (13)

Attempts to validate the applicability of Eq. (13) in practi-cal situations (with samples containing more than oneanalyte) have been made. Predictions made using Eq. (13)by Nielen [11] corresponded well with the response fac-tors of alkylsulfate surfactants, analyzed with veronal asthe probe. Cousins et al. [12, 13] used a number of probes

CE

and

CE

C

2152 C. Johns et al. Electrophoresis 2003, 24, 2150–2167

to experimentally determine R values for a series ofanions. However, experimentally determined values didnot match predicted values very well. Nevertheless, thegeneral trend predicted by Eq. (13), matched the valuesfound experimentally. Doble et al. [14] found Eq. (13) tobe valid for electrolytes containing only a probe anionand counterion. The introduction of another co-ion, forexample, bromide from tetradecyltrimethylammoniumbromide (TTAB) (used to modify the electroosmotic flow,EOF), caused a decrease in R and deviation from pre-dicted values. Maximum values of R occurred when theprobe and analyte had very similar mobilities. It shouldbe noted that although EOF plays an important role in CEseparations, the EOF mobility does not participate in theequations describing KRF.

It is also possible to detect analytes via the displacementof a counterion. This indirect detection mode is based onthe principle that the difference in mobility between theanalyte ion and the co-ion will result in a change in thecounterion concentration. If the co-ion is transparent, anabsorbing counterion can be used for detection, but astrong dependence of the detection sensitivity upon themobility difference between the analyte and the co-ion(with zero sensitivity occurring for matching mobilities)makes this detection mode somewhat impractical. Colletand Gareil [15] performed indirect detection of cationsusing anionic probes (benzoate, anisate). Negative andpositive peaks were observed depending on the relativemobilities of the analyte and co-ion. If the mobility of theanalyte was greater than the mobility of the co-ion, then apositive peak was observed, and if the mobility of theanalyte was less than the co-ion mobility then a negativepeak occurred [16]. This detection mode relies on a non-absorbing analyte being detected in a BGE consisting ofa nonabsorbing co-ion and an absorbing counterion,which acts as the probe. If the mobilities of the analyteion and the co-ion are identical then no peak will be ob-served. It was found that the response factor dependedon the absolute mobilities of the analyte ion, co-ion andcounterion.

2.2 Peak shapes

Peak shapes of analytes will depend on the relative mo-bilities of the probe and analyte. A number of papers onthis subject have been published [10, 17–20]. Mikkerset al. [21, 22] described analyte zone concentration distri-butions using a mathematical model derived from theKRF. Concentration distributions of the analyte bandswere found to depend on the relative mobilities of theprobe and analyte. This can be summarized as follows.Analytes which have a mobility higher than the probe will

have a concentration distribution which is diffuse at thefront and sharp at the back of the sample zone, whichwill result in a fronted peak (observed relative to the vec-tors of electrophoretic mobilities of the analyte and co-ion). The direction of EOF and whether the separation isrun in the co- or counter-EOF mode will then determinewhich side of the sample zone moves through the detec-tor first and consequently whether the peak appears asfronted or tailed in the electropherogram. Similarly, ana-lytes with mobilities lower than the probe will have a con-centration distribution which is sharp at the front of thezone and diffuse at the rear of the zone, causing tailingpeaks (observed relative to the vectors of electrophoreticmobilities of the analyte and co-ion). If the probe and ana-lyte have the same mobilities, a symmetrical peak willoccur.

This zone-broadening mechanism, which is generallyreferred to as electromigrational dispersion [23], also de-pends on another parameter, namely the relative con-centrations of the sample ion and the electrolyte co-ion.Electromigrational dispersion can be minimized by mani-pulating the concentrations of the electrolyte and samplesuch that the electrolyte concentration is more than twoorders of magnitude greater [21]. Therefore, the two mainapproaches to minimize electromigrational dispersion areto match the mobilities of the analyte(s) and the co-ionas closely as possible, and to keep the electrolyte con-centration as high and the injected sample amount aslow as possible.

2.3 Sensitivity of indirect photometric detection

A nonabsorbing analyte will have a limit of detection givenby [3, 24–27]:

CLOD � Cp

RDr� NBL

R�l(14)

where CLOD is the concentration limit of detection, Cp isthe probe concentration, R is the transfer ratio, Dr is thedynamic reserve (the ratio of background absorbance tonoise), NBL is the baseline noise, � is the molar absorptivityof the probe, and l is the detection cell path length. It canbe seen that decreases in limits of detection may bemade through decreasing the probe concentration Cp orbaseline noise NBL, or increasing the transfer ratio R, dy-namic reserve Dr, probe molar absorptivity �, or detectioncell path length l. However, Dr and Cp are not independentof each another. Decreasing Cp also reduces Dr, whichnegates any detection limit decreases. Previous attemptsat sensitivity enhancements have therefore focused onreduction of baseline noise, increases in pathlength andincreased molar absorptivity of the probe.


2.3.1 Baseline noise

There are various factors which contribute to the overallbaseline noise, and these can be divided into instrumen-tal and chemical factors. The first category includes fac-tors such as spatial and intensity stability of the lightsource, output of the light source at a given wavelength,and other instrumental parameters (electronic noise, tem-perature stability, mechanical rigidity of the optics, etc.).The chemical noise is caused by undesirable fluctuationsin the probe concentration leading to changes in thebackground absorbance of the BGE.

Attempts have been made to reduce the baseline noiseby optimal choice of the light source. Light sources forindirect photometric detection are usually operated ateither the emission maximum of the source or the absorp-tion maximum of the probe. Light sources such as deute-rium, zinc and cadmium lamps are used most commonly.Such lamps are typically operated in the UV range wheretheir light output and noise are optimal. An alternativeis the use of light-emitting diodes (LEDs), which offer ahighly stable, nearly monochromatic light source. Theyalso offer the possibility of excellent performance in thevisible region of the spectrum where the light output isrelatively low for the deuterium lamp. LEDs have beenused successfully for indirect detection, and typicallyresult in significant reductions of baseline noise, leadingto improvements in detection limits [28, 29]. However,no LEDs in the low-UV range which are suitable for usewith the typically used strongly UV-absorbing probes areavailable commercially.

2.3.2 Detection pathlength

The very short optical pathlengths involved in CE withon-capillary detection limit the detection sensitivity. Thedetection pathlength will be equivalent to the capillary inter-nal diameter (ID) only in an ideal situation where the opticsare such that all the individual rays of the light beam passthrough the centre of the capillary. In reality, the effectivepathlength will be shorter than the capillary ID. Capillarydiameters ranging from 50 to 100 �m are used typically forindirect detection. Increasing the capillary ID, and hence theoptical pathlength, results in increases in Joule heating andbaseline noise, which counterbalance potential increasesin sensitivity. Ma and Zhang [30] investigated the effect ofcapillary ID on signal-to-noise ratio (S/N) and found thatover a range of 25–75 �m the ID did not exert a critical influ-ence on limit of detection (LOD). They speculated that thiswas due to increased baseline noise caused by increasedJoule heating with larger capillary IDs. However Steiner etal. [31] calculated S/N for capillaries ranging in ID from 10to 10 000 �m and showed that S/N increased with capillary

ID. Thisdebate isclouded by the fact that the authors meas-ured the background absorbance of the BGE in each capil-lary without the application of voltage. This approach there-fore did not consider the effects ofJouleheating onbaselinenoise.

The use of capillaries with extended pathlengths only atthe detection window itself can offer a combination ofextended optical pathlength without increases in Jouleheating. Weston et al. [32] used a capillary of 75 �m IDwith a bubble of 300 �m at the detection point. Despiteincreasing the pathlength by a factor of 4, they reportedonly a twofold increase in S/N, due mainly to a resultantincrease in baseline noise caused by the extended path-length. Doble et al. [33] used a Z-cell arrangement withan optical pathlength of 3 mm. By comparison with a75 �m capillary, S/N was increased by a factor of 3.The less than expected gains can be explained by theincreases in baseline noise. The probe concentrationwas also reduced by a factor of 4 to maintain a back-ground absorbance of 0.127 AU, which additionally nega-tively influenced the electromigrational dispersion result-ing in broader peaks and smaller peak heights. It can beconcluded that unlike direct photometric detection, in-creasing the pathlength for indirect detection will resultonly in marginal improvements in S/N and correspondingconcentration LODs.

2.3.3 Probe absorptivity

The remaining parameter in Eq. (14) that can be optimizedin order to achieve lower detection limits is the absorp-tivity of the probe. Indeed, improvements in detectionlimits have been achieved in this area mainly throughuse of this approach. The majority of work has beenapplied to the detection of anions. Increasing the probeabsorptivity results in an increase in the dynamic reserve.The probe concentration may also have to be reducedto keep the BGE absorbance within instrumental limits(see Section 2.5), although lowering the probe concen-tration is undesirable as it will normally result in in-creased electromigrational dispersion. The mobility ofthe probe should match the analyte mobility as closelyas possible to improve peak shapes and maximize thetransfer ratio and hence sensitivity. Foret et al. [34]observed a 50-fold increase in sensitivity by using sor-bate (� = 24,120 L�mol�1cm�1 at 254 nm) as a probecompared to the use of benzoate (� = 809 L�mol�1cm�1

at 254 nm) for the detection of aspartate, butyrate, chlo-rate, chloride, dichloroacetate, dimethylmalonate, gluco-norate, glutamate, hydroxyisobutyrate, lactate, malo-nate, methylmalonate, and phosphate. Beck and Engel-hardt [35] observed that the most sensitive detection of


organic and inorganic cations occurred when cationicprobes of the highest absorptivity and mobility closestto the analytes were used. Weston et al. [32] reportedthat changing cationic probes from UV Cat 1 to UV Cat 2(both of which are proprietary reagents of unspecified com-position) improved detection limits of inorganic cations bytwo to four times, due to the higher absorptivity of UV Cat 2and its better mobility match to the analytes.

2.4 Buffering of electrolytes for indirectphotometric detection

The necessity for BGEs to be well buffered has beendemonstrated clearly by numerous authors [36–39]. Theapplication of high voltages to conducting solutions andthe resultant electrolysis can result in a pH decrease atthe anode, where hydrogen ions are formed, and anincrease in pH at the cathode where hydroxyl ions form.Kenndler and Friedl [36] have reported that pH changesas small as 0.03 pH units can cause a loss of resolutionand changes in selectivity. As the EOF, and hence also theanalyte migration times, are dependent on the electrolytepH, electrolytes must contain some degree of buffering inorder to provide suitable reproducibility and ruggednessof the method. Unfortunately, many BGE systems usedin CE provide little or no buffering. Investigations into theeffects of pH variations provide compelling arguments forthe necessity of well-buffered systems. Zhu et al. [37]

observed that the use of BGEs with lower buffering ca-pacities caused faster pH changes, while Strege andLagu [38] highlighted the need for BGE replenishment toprevent poor migration time reproducibility caused bypH variations. Experimentally measured pH changesagreed well with calculated values [40, 41]. Visualizationof pH changes around a capillary and electrode andmeasurement of those changes in the capillary havebeen achieved by the addition of pH indicators to BGEsby Macka et al. [39], again emphasising the need forproper buffering.

Buffering electrolytes for indirect detection may be per-formed in four ways. In the first approach, the probe itselfcan be used as a buffer. For example, for anion detectiona weak acid such as phthalate [42] or benzoate [43] isused at or near the pKa of the probe. An example is shownin the separation of 53 anions in Fig. 1 [44], where the BGEconsists only of 20 mM 2,6-pyridinedicarboxylic acid(probe) and 0.5 mM cetyltrimethylammonium hydroxide(EOF modifier). However, there are three major limitationsto this approach. (i) The pH range of the BGE is confinedto about plus or minus one pH unit from the pKa of theprobe. (ii) The buffering capacity of the BGE is dependenton the probe concentration, and (iii) the probe must beonly partially ionized (to provide suitable buffering) andwill therefore have low mobility. This means that thisapproach will be suited only to the analysis of analyteswith low to moderate mobility.

Figure 1. Separation of a standard mixture of inorganic and organic anions, amino acids and carbo-hydrates. Conditions: capillary, fused silica (50 �m ID), 112.5 cm (104 cm effective length); electrolyte,20 mM 2,6-pyridinedicarboxylic acid, 0.5 mM cetyltrimethylammonium hydroxide, pH 12.1; voltage,�30 kV; injection, 6 s at 50 mbar; temperature, 15�C; detection, signal = 350 nm, reference = 230 nm.Concentrations: chloride, 88 mg/L; phosphate, 80 mg/L; arginine and carbohydrates, 200 mg/L each;others, 40 mg/L each. Reprinted from [44], with permission.


A second approach is to add a co-ion as a buffer. Thismethod has been used with addition of phosphate [45],carbonate [46], acetate [47], 2-(cyclohexylamino)ethane-sulfonic acid [48], and borate [42, 49, 53] to BGEs fordetection of anions. Whilst this may provide adequatebuffering capacity, the introduction of a co-ion and theresultant competitive displacement by analyte ions canlead to a loss of detection sensitivity and the occurrenceof system peaks which may interfere with detection ofanalytes. This approach can be used successfully only ifthe mobilities of the probe, co-ion and analyte fall in theappropriate regions so that system peaks will occur atmigration times which do not interfere with analytes.

A third means of buffering in indirect detection is the useof a counterionic buffer as this provides suitable bufferingwithout the addition of a co-ion. This approach has beenused to provide buffering to chromate-based BGEs fordetection of anions by the addition of buffering cationssuch as Tris [26, 54–62] and triethanolamine [63–67]. Theuse of anionic buffers such as acetic acid with cationicprobes for indirect detection of cations has also beenperformed.

The fourth and so far least commonly employed alter-native is the use of isoelectric ampholytic buffers. Anampholyte at its isoelectric point will have zero (or veryclose to) overall charge and will not act as a competingco-ion. Providing the pKa of the functional groups on theampholyte are sufficiently close (within 2 pH units) to thepI, the ampholyte will provide buffering [68, 69]. Unfortu-nately, few ampholytes that are suitable as BGE buffersfulfil this requirement. Doble et al. [33] have used lysineand glutamic acid as buffers for indirect detection ofanions. Bromocresol green was used as a probe and theBGE was buffered with 10 mM lysine at pH 9.7 for the in-direct detection of C2–C8 alkanesulfonic acids. The mod-erately high EOF meant that separations were performedin a counter-EOF mode and detection of high mobilityanions could not be performed with this approach. Suchanions were determined using potassium indigotetra-sulfonate (ITS) as the probe with the BGE buffered with10 mM glutamic acid at pH 3.22. EOF was sufficiently lowto allow the counter-EOF mode to be used with a negativevoltage polarity. However, this system was restricted tothe detection of relatively strong acid anions due to thelow pH of the BGE. Histidine has been recently used as abuffer for the indirect detection of a wide range of organicand inorganic anions. The highly absorbing anionic dyestartrazine and naphthol yellow S [70] and Orange G [71]were used as probes in electrolytes buffered with histidineat its pI. Reproducibility measurements over a series ofruns without BGE replenishment or replacement showedthat histidine at its pI (7.7) provided effective buffering.

Isoelectric ampholytic buffers will have very low conduc-tivity at their pI and can be added to the BGE in sufficientlyhigh concentrations to provide adequate buffering with-out increasing Joule heating effects.

2.5 Detector linearity

Indirect photometric detection must be performed at abackground absorbance which is within the linear rangeof the detector used in order to provide a linear responsewithout a loss of detection sensitivity. Probes are typicallyused at concentrations which result in a background ab-sorbance of around 0.05–0.15 AU. Some common exam-ples are the use of chromate at concentrations rangingfrom 5 to 10 mM for detection of anions and imidazolefrom 5 to 10 mM for cation detection. As a result, morehighly absorbing probes are often used at concentrationsof roughly an order of magnitude less so as to keep thebackground absorbance below about 0.15 AU. In orderto minimize electromigrational dispersion and to utilizethe benefits of stacking effects, the ionic strength of theBGE should be as high as possible in comparison withthe ionic strength of the sample. This is achieved withdirect detection by increasing the buffer concentrationuntil excessive separation current and Joule heatingresults. Practically, a simple plot of buffer concentrationversus separation current (at constant voltage) can beused to determine the maximum acceptable buffer con-centration. This is the concentration at which the aboveplot deviates substantially from linearity. The advantagesof buffer concentration maximization include narrowerpeaks, increased separation efficiency and resolution,increased peak heights, and the ability to handle samplesof higher ionic strength. Similar benefits can be expectedto apply to the use of indirect detection.

By comparison, indirect detection has often been usedwith probe concentrations set at arbitrary values withoutrigorous investigation. Improvements in peak shapes andsymmetries have been noted by increasing the concen-tration of pyromellitic acid [66]. Further increases in probeconcentrations (and hence higher background absorb-ances) are generally avoided for fear of exceeding thelinear range of the detector. Surprisingly, little work hasbeen performed on the linearity of detectors for use inCE [72–75]. The most common approach has been tomeasure the absorbance of a series of probe solutionsand to plot absorbance versus probe concentration andto estimate the absorbance at which deviation from line-arity occurs [76]. Macka et al. [72] reported the use of anapproach in which sensitivity (absorbance/concentration)was plotted versus absorbance and the linearity limitwas defined as the concentration at which sensitivity


decreased by a certain amount from its optimal value. Fit-ting curves to experimentally measured sensitivity datawas used to derive the effective pathlength of the detec-tor. This approach was applied successfully to the evalu-ation of linearity of various instruments and probes. Upperdetector linearity limits of a series of instruments anddetector configurations [77] were evaluated using a mod-ification of this technique. Importantly, this work sug-gested that most detectors maintained a linear responseat absorbances far in excess of values previously ex-pected. Hence, it is feasible to use highly absorbingprobes at similar concentrations to less absorbing probeswithout a loss of detection sensitivity. An estimate ofeffective pathlengths could also be made which alloweda qualitative comparison of the quality of detector optics.Also, the technique allowed that irregularly shaped detec-tion cells could be evaluated using a similar procedure.The upper detector linearity limit and effective pathlengthof an Agilent Technologies Extended Light Path Capillaryand Agilent Technologies High Sensitivity Detection Cellwere determined using this approach [78].

2.6 Highly absorbing probes

The most effective approach to improve detection sensi-tivity is to use more highly absorbing probes. The use ofhighly absorbing dyes as probes is an attractive optionbut to date has found limited application. Examples ofthis approach are reviewed below.

2.6.1 Use of anionic dyes as probes

Most work involving the use of dyes as probes hasfocused on the detection of anions. Xue and Yeung [28]determined pyruvate using the anionic dye bromocresolgreen. A 0.5 mM bromocresol green, pH 8.0, unbufferedBGE was used, but unstable baselines were observed.Mala et al. [25] detected anions with the dyes indigo car-mine and chlorphenol red. Unfortunately, both BGEs usedin this work were buffered with a combination of Tris/acetic acid, and the acetate present would act as com-peting co-ions. This would account for the worse thanexpected sensitivity and LODs reported in this study.Siren et al. [79] used nitrosonaphthol dyes for detectionof organic acids and inorganic anions. Attempts to bufferthe BGE at pH 8 with phosphate were unsuccessful assensitivity was compromised. Accordingly, the dyes werefurther used in unbuffered BGEs. Doble et al. [33] usedbromocresol green for the detection of C2�C8 alkane-sulfonic acids and BGEs buffered with the counter-iondiethanolamine (DEA) and the isoelectric ampholytic buf-fer lysine were compared. Detection limits with lysine as abuffer ranged from 1 to 2 �M and were approximately an

order of magnitude greater than those obtained with theuse of DEA as a buffer. The slopes of calibration curveswith lysine were about 70% of those with DEA. Theauthors explained both observations by noting that lysinecontained approximately 2% carbonate, which acted asa co-ion and hence caused a decrease in sensitivity. Thehigher mobility dye potassium indigotetrasulfonate wasused for detection of more mobile analyte anions (Fig. 2)[33], and excellent sensitivity (with detection limits rang-ing from 0.1 to 2 �M) was observed when the BGE was

Figure 2. Electropherogram of 14 anions obtained from(a) a high-concentration standard and (b) near their detec-tion limits with ITS as probe. Conditions: capillary, 75 �mID, 70 cm�50 cm; electrolyte, 0.5 mM ITS, 2.67 mM Bis/Tris, pH 6.8; separation voltage, �30 kV; 0.6 s pressureinjection at 5” Hg; detection wavelength, 314 nm; tem-perature, 30�C. Peak identification: 1, bromide (a) 200 �M,(b) 10 �M; 2, chloride (a) 200 �M, (b) 10 �M; 3, sulfate(a) 80 �M, (b) 2 �M; 4, thiocyanate (a) 80 �M, (b) 2 �M; 5,chlorate (a) 80 �M, (b) 2 �M; 6, malonate (a) 40 �M, (b) 1 �M;7, tartrate (a) 40 �M, (b) 0.5 �M; 8, bromate (a) 40 �M,(b) 1 �M; 9, formate (a) 40 �M, (b) 1 �M; 10, citrate (a) 30 �M,(b) –; 11, succinate (a) 30 �M, (b) 0.8 �M; 12, phthalate(a) 30 �M, (b) 0.8 �M; 13, iodate (a) 60 �M, (b) 1.5 �M;14, phosphate 60 �M, (b) 1.5 �M. Reprinted from [33],with permission.


buffered with Bis/Tris. Use of the ampholytic buffer glu-tamic acid in the BGE was applicable only for detectionof relatively strong acid anions due to the low BGE pH of3.22. However, sensitivity was very good, with detectionlimits between 0.7 and 1 �M. Tartrazine and naphthol yel-low S were used at low concentrations in BGEs bufferedwith histidine at its pI [70].

Awidemobility rangeofanions was detected in the counter-EOF mode, with EOF significantly suppressed by the addi-tion of the neutral polymer hydroxypropylmethylcellulose(HPMC). Limits of detection were generally in the sub-micromolar range. Tartrazine was also used at increasedprobe concentrations for the detection of alkanesulfonicacids [80]. EOF could not be suppressed by the addition ofHPMC and separations were performed in the counter-EOFmode with a positive polarity voltage and high EOF. Theaddition of a zwitterionic surfactant minimized baseline dis-turbances and provided a suitably high EOF to allow anionsof moderate mobility to be transported past the detectionwindow. The use of the probe at increased concentrationsmarkedly improved peak shapes (at higher analyte concen-trations), separation efficiency and resolution without anyloss of detection sensitivity. The highly absorbing dyeOrange G was used as a probe at high concentrations forthe detection of inorganic and organic anions [71]. Limits ofdetection were decreased by increasing the probe concen-tration. EOF was suppressed by a combination of a poly-ethylenimine-coated capillary and addition of HPMC. Anoptimized BGE was used for the analysisofair filter samplesand the results obtained were comparable with those fromion chromatography.

2.6.2 Use of cationic dyes as probes

Cationic dyes have been used as probes even less fre-quently than anionic dyes. Xue and Yeung [28] detectedpotassium with a 0.5 mM malachite green, pH 3, BGE.Mala et al. [25] used methyl green for the detection ofcaesium, potassium, calcium, magnesium, sodium andlithium ions. The BGE was buffered with 1 mM Tris titratedwith acetic acid to pH 6.5. Tris would act as a competingco-ion at this pH and would therefore reduce detectionsensitivity. It was noted that no separation was observedwhen the competing co-ion (Tris) was used at a concen-tration of 10 mM. This approach was adapted by Butler etal. [81] who substituted pyronine G for methyl green inorder to better match the emission wavelength of an LEDlight source. Potassium, calcium, sodium, and lithiumions were separated in a 0.15 mM pyronine G, 1 mM Tris,pH 4.0, BGE. Tetrazolium violet was used by Shamsi andDanielson [82] for separation of C12�C18 dialkyldimethylquaternary ammonium compounds and cis/trans-ada-

mantane isomers. Detection limits ranged from 0.05 to1 mg/L. Buffering was provided at pH 6.0 by 100 mM

borate. Significant amounts of methanol (50–85% v/v)were present in the BGEs used.

In a similar approach to that used for the analysis ofanions with Orange G, the highly absorbing dye chrysoi-dine was used as a cationic probe [83]. The probe con-centration was increased as high as possible whilst main-taining optimum detection sensitivity. In a significantadvance over previous uses of cationic dyes, separationselectivity was introduced by the addition of weak acidauxiliary complexing agents. This allowed a wide rangeof metal ions to be separated. Additionally, the pH ofthe BGE was close to the pKa of the weak acid in orderto provide effective buffering.

2.7 Separation and indirect photometricdetection of cations

Many metal ions have very similar electrophoretic mobili-ties and cannot be easily separated by CE. The introduc-tion of auxiliary ligands which can form complexes withmetal ions can be used to achieve changes in selectivity.Two approaches are possible. A strongly complexing aux-iliary ligand, such as cyanide [84–87], EDTA and trans-1,2-diaminocyclohexane-N,N,N’,N’-tetraacetic acid (CDTA)[88–102], and metallochromic ligands [103–107], can beused to form highly stable (usually anionic) complexes.Complexation is normally performed prior to injection,and this approach is referred to as precapillary complexa-tion. Detection of the metal complexes is performed in thedirect mode. As the majority of the metal ion is com-plexed, it is difficult to change the separation selectivity.An alternative is the use of weakly complexing ligandswhere the metal is only partially complexed, so that selec-tivity changes can be accomplished by varying thedegree of complexation using changes in pH or ligandconcentration. Complexation is normally performed afterthe injection and occurs when the sample contacts theBGE containing the auxiliary ligand. This is referred to ason-capillary complexation. Detection of the complexes isnormally performed indirectly. This is the most commontechnique for the separation of metal ions, typically usingweak acid ligands such as 2-hydroxyisobutyric acid(HIBA), citric acid, lactic acid, acetic acid, glycolic acid,succinic acid, and acetic acid.

The first demonstration of this approach was by Foretet al. [108] who separated all 14 lanthanides by theaddition of HIBA. Jandik and others [109–111] separated15 alkali, alkaline earth and transition metals and 13 lanth-anides using 4-methylbenzylamine (UV Cat 1) as a probewith HIBA and citric acids as complexing agents. Chen


Figure 3. Electrophoretic separation of 16 metal ions.Capillary: 75 �m ID, 100 cm�108 cm. Carrier electrolyte,10 mM 4-aminopyridine-6.5 mM HIBA (pH 4.5 adjustedwith sulfuric acid); hydrostatic injection, 20 s from 10 cmheight; voltage, 25 kV. Peaks according to the sequence:1, K (0.4 ppm); 2, Na (0.8 ppm); 3, Ba (1.2 ppm); 4, Sr(0.6 ppm); 5, Ca (0.6 ppm); 6, Mg (0.6 ppm); 7, Mn(0.6 ppm); 8, Li (0.4 ppm); 9, Fe (0.6 ppm); 10, Co(0.6 ppm); 11, Cd (0.6 ppm); 12, Ni (0.6 ppm); 13, Zn(0.6 ppm); 14, Pb (0.4 ppm); 15, Cr (0.8 ppm); 16, Cu(1.0 ppm). Reprinted from [143], with permission.

and Cassidy [112] achieved baseline separation of 26metals with N,N’-dimethylbenzylamine as probe andHIBA as complexing agent. Shi and Fritz [113] separated27 metals using 4-methylbenzylamine and lactic acid. Linet al. [114] investigated the use of 10 carboxylic acids(acetic, glycolic, lactic, HIBA, oxalic, malonic, malic, tar-taric, succinic, and citric) as complexing agents. It wasfound that oxalic and citric acid caused the largest selec-tivity changes. The optimum pH in terms of separationselectivity was found to be near the pKa1 of the complex-ing agent. An example of the separation of metal ionsusing HIBA is shown in Fig. 3. Optimization of separationsinvolving large numbers of metals has been thoroughlyinvestigated [113, 115].

2.8 Simultaneous separation and indirectphotometric detection of anions and cations

The simultaneous separation and detection of anions andcations poses a number of difficulties. Firstly, anions andcations must be transported past the detection window insuch a way as to allow separation and secondly the BGEmust contain suitable probes to allow the detection ofthe ions. A small number of approaches has been used

to address this task. It is possible to employ an EOF whichhas sufficient velocity to transport ions which are migrat-ing in the opposite direction to the EOF past the detectionpoint. The simplest way to increase the magnitude of theEOF is to increase the BGE pH. However, this may not besuitable for the detection of cations as alkaline earths andother metal ions may form hydroxide complexes or preci-pitates. Additionally, cationic probes may not be proto-nated at higher pH values. Alternatively, external pressurecan be applied to provide a flow in addition to the EOF, asdemonstrated by Haumann et al. [116]. The most suc-cessful means of simultaneous detection involves theuse of sample injection at both ends of the capillary.Cations and anions can then be allowed to migrate pastthe detector due to their electrophoretic mobilities. TheBGE should contain both an anionic and cationic probein order to provide detection. Clearly, from indirect detec-tion principles the presence of competing co-ions shouldbe minimized. This suggests that probes should be pre-sent as free acids and bases. The BGE must be at a pHat which both probes will be charged.

Padarauskas et al. [117] used two different BGEs forsimultaneous detection. Imidazole (cationic probe) andnitrate (anionic probe) were used at pH 4 to separate anddetect potassium, chloride, sulfate, calcium, sodium, andmagnesium ions. Buffering was provided by fumaric acid.An electrolyte containing copper(II)-ethylenediamine andnitrate at pH 8.5 allowed detection of chloride, sulfate,potassium, ammonium, hydrogencarbonate, calcium, so-dium, and magnesium ions. Fumaric acid was utilized asa complexing agent to improve separation selectivity be-tween calcium and sodium. Triethanolamine was used asa buffer, which resulted in a system peak occurring in amobility region away from that of the analytes. The sam-ple was injected electrokinetically at both ends of thecapillary and a detection window near the middle of thecapillary was used. This approach was further developedby Padarauskas et al. [118] using a BGE containing cop-per(II)-ethylenediamine as the cationic probe and chro-mate (in the form of chromic trioxide) as the anionicprobe. Potassium, calcium, magnesium, sodium, ammo-nium, chloride, hydrogencarbonate, sulfate, and nitrateions were separated. Triethanolamine provided bufferingat pH 8.0. This method was used for the analysis of tap,river and mineral waters.

Kuban and Karlberg [119] used a BGE containing 4-ami-nopyridine and chromic acid at pH 8.0 for detection ofcations and anions, respectively. The sample was in-jected hydrodynamically at both ends of the capillaryand again a detection window in the middle of the capil-lary was used. Separation of 22 small inorganic andorganic anions and alkali and alkaline earth metal ions


was achieved in less than 5 min. Transition metal cationsand barium could not be analyzed as hydroxide forma-tion and/or precipitation with chromate occurred.

Xiong and Li [120] investigated the use of benzylamineand imidazole (cationic probes) and benzenesulfonicacid, sulfosalicylic acid and pyromellitic acid (anionicprobes) in BGEs for simultaneous detection of anionsand cations. Sample was injected hydrostatically at theanodic end of the capillary and ions were transportedpast the detection window by a high-velocity EOF. ABGE containing imidazole and sulfosalicyclic acid wasused to detect potassium, sodium, lithium, hydrogenphosphate, fluoride, hydrogen carbonate, chlorate, andperchlorate ions. A benzylamine-pyromellitic acid BGEwas used to detect potassium, sodium, lithium, acetate,hydrogen phosphate, fluoride, chlorate, perchlorate,nitrate, nitrite, chloride, and sulfate ions.

Xiong and Li [121] further developed this approach bystudying the use of imidazole, 1,2-dimethylimidazole,benzylamine (cationic probes) and sulfosalicylic acid, tri-mellitic acid and pyromellitic acid (anionic probes) (Fig. 4).They found that a BGE consisting of 1,2-dimethylimida-zole/trimellitic acid was the best combination. Imidazolewas not used further as it was desired to perform separa-tions at a pH higher than the pKa of imidazole, meaningit would not function as a probe. Benzylamine was dis-carded due to its lower absorptivity (between 200 and230 nm) compared with 1,2-dimethylimidazole. Trimelliticacid was selected as the anionic probe based on superiordetection sensitivity and peak shapes compared with theother anionic probes. This electrolyte was then optimized,allowing the separation and detection of 4 alkali metalions and 12 organic acids in 6 min. Detection limits rang-ed from 0.08 to 5 �g/mL. The method was used suc-cessfully for the analysis of apple, grape and orangejuices.

Raguenes et al. [122] used a BGE containing Tris/1,5-naphthalenedisulfonic acid to separate and detect potas-sium, ammonium, sodium, lithium, ascorbate, sorbate,benzoate, lactate, acetate, hydrogencarbonate, phos-phate, formate, and fluoride ions in less than 9 min withdetection limits between 0.1 and 1 ppm. The analysis ofa real sample yielded results comparable with inductivelycoupled plasma-mass spectroscopy and ion chromatog-raphy. Haumann et al. [116] used imidazole-thiocyanateand dimethyldiphenylphosphonium iodide-trimesic acidBGEs in conjunction with hydrodynamic sample injectionat both ends of the capillary for detection of inorganicanions and cations, small organic acids and aliphaticamines. The analysis of drinking water, mineral water andan alcoholic beverage demonstrated the utility of thissystem.

Figure 4. Separation of 4 cations and 12 anions inthree binary BGEs at pH 7.0. (A) 1,2-Dimethylimidazole(4.0 mM)-sulfosalicylic acid (1.6 mM), � = 217 nm; (B) 1,2-dimethylimidazole (4.0 mM)-trimellitic acid (1.0 mM), � =210 nm; (C) 1,2-dimethylimidazole (5.0 mM)-pyromelliticacid (0.8 mM), � = 210 nm. 18-Crown-6, 2.0 mM in all threebinary BGEs; Injection, 15 s; capillary, 50 �m ID, 358 �M

OD, length 38.0�44.0 cm; voltage, �30 kV. Peak identifi-cation (�g/mL): 1, NH�

4 (12.7); 2, K� (35.9); 3, Na� (173);4, Li� (8.7); 5, H2O; 6, ascorbate (13.3); 7, sorbate (15.5);8, benzoate (27.5); 9, lactate (10.1); 10, acetate (43.9);11, succinate (8.4); 12, malate (8.2); 13, tartrate (20.4);14, maleate (8.9); 15, malonate (8.5); 16, perchlorate(8.7); 17, oxalate (6.7). Reprinted form [121], with permis-sion.

3 Tabular summary of the literature

Tables 1–5 provide a comprehensive overview of pub-lished indirect detection methods for anions and cations.The major emphasis of this search of the literature hasfocused on high sensitivity detection. Less absorbing,commonly used probes such as chromate and imidazolehave already been thoroughly reviewed and are thereforeomitted from the tables. Probes which have reportedabsorbances of less than 10 000 L�mol�1cm�1 have also


Table 1. Benzene carboxylate probes for indirect detection of anions

Probe BGE range(mM)

Detection wave-length (nm)

Buffering method LODa) Analytes Ref.

p-Aminobenzoate 7.5–20 250, 264 None (probe only) 10–40 nM (Large-volume injection) chloride,sulfate, nitrite, nitrate, formate,acetate, propionate, methane-sulfonate, oxalate, malonate,maleate, succinate, glutarate, adi-pate, pimelate, suberate, azelate,sebacate

[123]

Benzoate 100 254 Counterion (Tris) 50 �M �,�,-Cyclodextrins [56]

20 228 Counterion (Tris) 0.3–1.1 fmol Bromide, chloride, nitrate, sulfate [30]

40 254, 270, 300 None (probe only) 1–2.5 �M C2–C14 fatty acids [57]

o-Benzylbenzoic acid 20 228 None (probe only) 0.3–1.1 fmol Bromide, chloride, nitrate, sulfate [30]

3,4-Dimethoxycinnamicacid

12 256, 267, 310 None (probe only) 0.01–0.04 mM Fructose, glucose, maltose,sucrose

[124]

p-Hydroxybenzoate 5 254 None (probe only) N/A Bromide, butanesulfonate,carbonate, chloride, citrate,ethanesulfonate, hexanesulfonate,molybdate, nitrate, nitrite, pen-tanesulfonate, phthalate, propane-sulfonate, sulfate, tungstate

[125]

Phthalate 3 214, 254 Counterion (Tris) 2–4 �M Sulfate, disulfide, tetrathionate,thiosulfate

[55]

5 254 Co-ion (borate) 12 �M Acetate, carbonate [53]

5 254, 270 Co-ion (borate) 200 ng/mL Bisinositol phosphate, hexa-kisinositol phosphate, mono-inositol phosphate, trisinositolphopshate

[42]

5 230 None (probe only) 0.1–0.4 �g/mL Bromoacetate, chloroacetate,formate, monofluorioacetate

[126]

5 254, 280 Co-ion (boric acid) 100–200 ng/mL Acetate, bromide, butyrate,chloride, formate, nitrate, nitrite,oxalate, propionate, sulfate

[63]

1.5 230 Co-ion (carbonate) 60–360 pg Adipate, �-ketoglutarate, citrate,ethylmalonate, glutarate, lactate,malonate, methylmalonate, meth-ylsuccinate, oxalate, pyruvate,succinate, tartrate

[127]

5 185 None (probe only) 0.7–1.1 mg/L Acetate, citraconate, citrate,crotonate, hydroxyisobutyrate,itaconate, mesaconate, meth-acrylate, pyruvate

[128]

Pyromellitate 1.5 214, 254 Counterion (Tris) 2–4 �M Sulfate, sulfide, tetrathionate,thiosulfate

[55]

2.25 254 Counterion (tri-ethanolamine)

1–3 mg/L Bromide, chloride, fluoride, nitrate,nitrite, phosphate, sulfate,thiosulfate

[64]

3 220 Counterion (diethyl-enetriamine)

0.01–1.0 mg/L Acetate, chloride, citrate, lactate,malate, nitrate, oxalate, phos-phate, succinate, sulfate, tartrate

[129]

a) Units quoted are those used in the original references.


Table 1. Continued

Probe BGE range(mM)


Buffering method LOD Analytes Ref.

2.25 254, 280 Counterion(triethanolamine)

100–200 ng/mL Acetate, bromide, butyrate,chloride, formate, nitrate, nitrite,oxalate, propionate, sulfate

[63]

2.25 254 Counterion(triethanolamine)

0.04–0.15 mg/L Bromide, chloride, nitrate, nitrite,oxalate, sulfate

[130]

2.25 254 Counterion(triethanolamine)

0.01–0.04 mg/L Bromide, chloride, sulfate, nitrite,nitrate, fluoride, phosphate

[131]

2-Sulfobenzoic acid 20 228 None (probe only) 0.3–1.1 fmol Bromide, chloride, nitrate, sulfate [30]

5-Sulfosalicylic acid 5 214 None (probe only) 5 fmol Heparin fragments [132]

0.1 214 None (probe only) 100–900 ppb Phosphate, fluoride, carbonate,chlorate, perchlorate

[120]

N/A, not available

Table 2. Naphthalene carboxylate probes for indirect detection of anions

Probe BGE range(mM)


Buffering method LODs Analytes Ref.

2,6-Naphthalenedicar-boxylate

1 240 None (probe only) 0.025 mg/L Acetate, butyrate, propionate [133]

2 280 None (probe only) 20 �g/mL Acetate, adipate, azelate, benzoate,butyrate, carbonate, chloro-acetate, dichloroacetate, formate,fumarate, glutarate, methane-sulfonate, phthalate, pimelate,propionate, sebacate, suberate

[134]

2 254, 280 None (probe only) 100–200 ng/mL Acetate, bromide, butyrate, chloride,formate, nitrate, nitrite, oxalate,propionate, sulfate

[63]

4 214 Counterion (Bis/Tris) 47–180 mg/L Oxalate, malonate, formate, malate,succinate, phthalate, methane-sulfonate, glutarate, glycolate, pyru-vate, suberate, acetate, glyoxylate,lactate, propionate, benzoate

[135]

1-Naphthylacetic acid 0.1–2 222 None (probe only) 0.1 mM Carbohydrates [136]

Bis/Tris, 2,2-bis(hydroxymethyl)-2,2’,2’’-nitrilotriethanol

not been included. Tables are used to group variousprobe types, with benzene carboxylate probes beinglisted in Table 1, naphthalene carboxylate probes listedin Table 2, aromatic sulfonate probes listed in Table 3,and miscellaneous anionic probes listed in Table 4. Cat-ionic probes are listed in Table 5. Each entry containsinformation on the probe concentration which has been

used, detection wavelength(s), ions analyzed and LODs(where given in the original publication), and the methodof buffering. The complexing agent and mode of com-plexation (if any) is also given for all cationic probeentries. Absorptivities of anionic probes are given inTable 6 and cationic probe absorptivities are collated inTable 7.


Table 3. Aromatic sulfonate probes for indirect detection of anions

Probe BGE range(mM)



1,5-Naphthalene-disulfonate

3 214, 254 Counterion (Tris) 2–4 �M Sulfate, sulfide, tetrathionate,thiosulfate

[55]

4–8.3 214, 284, 288 Co-ion (borate) 10–275 �g/L(inorganic anions),30–60 �g/L(organic acids)

Inorganic anions, organicacids

[137]

2-Naphthalenesulfonate 4–10 254, 270, 300 Counterion (Tris) 1–2.5 �M C2–C14 fatty acids [57]

1,3,6-Naphthalene-trisulfonate

2 214, 254 Counterion (Tris) 2–4 �M Sulfate, sulfide, tetrathionate,thiosulfate

[55]

4–8.3 214, 284, 288 Co-ion (borate) 8–250 �g/L(inorganic anions),20–50 �g/L(organic acids)

Inorganic anions, organicacids

[137]

Table 4. Miscellaneous probes for indirect detection of anions

Probe BGE range(mM)



Bromocresol green 0.5 633 None (probe only) 2�10�16 mol Pyruvate [28]

0.5 618 Ampholytic (lysine),counterion(diethanolamine)

0.1–0.2 �M C2–C8 alkanesulfonic acids [33]

Chlorphenol red 0.5 578, 620, 635 Co-ion (acetic acid) 2�10�13�8�10�16 mol

Bromide, chloride, citrate,fumarate, glutamate, malonate,nitrate, sulfate

[25]

1,3-Dihydroxynaph-thalene

0.1 222 None (probe only) 0.1 mM Carbohydrates [136]

Indigo carmine 0.5 578, 620, 635 Co-ion (acetic acid) 2�10�13�8�10�16 mol

Bromide, chloride, citrate,fumarate, glutamate, malonate,nitrate, sulfate

[25]

Potassium indigo-tetrasulfonate

0.5 314 Counterion(Bis/Tris)

0.1–2 �M Bromide, chloride, sulfate,thiocyanate, chlorate, malonate,tartrate, bromate, formate,succinate, phthalate, iodate,phosphate

[33]

0.5 314 Ampholytic(glutamic acid)

0.7–1 �M Sulfate, nitrate, perchlorate,chlorate, bromate

[33]

2,6-Pyridinedi-carboxylate

5 214 None (probe only) 1–2.5 mg/L Acetate, bromide, butyrate, citrate,formate, heptanoate, hexanoate,iodide, lactate, malate, nitrate,octanoate, oxalate, pyruvate,succinate, tartrate, valerate

[138]

20 230 None (probe only 6–12 mg/L Inorganic and organic anions,amino acids

[44]

20 230 None (probe only) 23–37 mg/L Carbohydrates

Riboflavin 12 256, 267, 310 None (probe only) 0.01–0.04 mM Fructose, glucose, maltose,sucrose

[124]


Table 4. Continued

Probe BGE Range(mM)



Sorbate 6 256 None (probe only) 0.23–0.29 mM D-Fructose, L-fucose, D-galactose,D-glucose, D-glucuronic acid,galacturonate, N-acetyl-galactos-amine, N-acetyl-glucosamine,N-acetyl-neuraminic acid

[139]

6 256 None (probe only) 2 pmol Carbohydrates [140]

6 254 None (probe only) 2000 fmol 2-Deoxy-D-galactose, 2-deoxy-D-ribose, cellobiose, D-arabonicacid, D-fructose, D-fucose,D-galactonic acid, D-galactose,D-galacturonic acid, D-gluconicacid, D-glucose, D-glucuronic acid,D-lyxose, D-mannonic acid,D-mannose, D-mannuronic acid,D-ribonic acid, D-ribose, D-xylose,lactose, lactulose, I-arabinose,I-rhamnose, I-sorbose, maltose,maltotriose, melibiose, raffinose,saccharose

[141]

12 256, 267, 310 None (probe only) 0.01–0.04 mM Fructose, glucose, maltose,sucrose

[124]

Tartrazine 0.5 258 Ampholytic (histidine) 0.1–0.8 �M Organic and inorganic anions [70]

0.5–5 258 Ampholytic (histidine) 0.63–0.89 �M C1–C8 alkanesulfonic acids [80]

Naphthol yellow S 0.5 223, 476 Ampholytic (histidine) 0.3–1.3 �M Organic and inorganic anions [70]

Orange G 1–7 248, 476 Ampholytic (histidine) 0.22–0.91 �M Organic and inorganic anions [71]

Bis/Tris, 2,2-bis(hydroxymethyl)-2,2’,2’’-nitrilotriethanol

Table 5. Miscellaneous probes for indirect detection of cations

Probe Complexingagent

Complexformation

BGErange(mM)

Detectionwavelength(nm)

Bufferingmethod

LODs Analytes Ref.

4-Aminopyridine 18C6, HIBA ON 5–15 254 Counterion(HIBA,acetate)

� low ppm Lithium, sodium, po-tassium, ammonium,manganese, ironII,cobaltII, cadmium,nickel, zinc

[142]

HIBA ON 10 214 Counterion(HIBA)

92–454 ppb Potassium, sodium,barium, strontium,calcium magnesium,manganese, lithium,iron, cobalt, cadmiumnickel, zinc, lead,chromium, copper

[143]

Cry22 ON 5 254 Counterion(acetate)

N/A Sodium, potassium,ammonium

[144]


Table 5. Continued

Probe Complexingagent

Complexformation

BGErange(mM)

Detectionwavelength(nm)

Bufferingmethod

LODs Analytes Ref.

Tetrazoliumviolet

None None 5–10 300 Counterion(borate)

0.05–0.50 mg/L Chloroethyltrimethyl-ammonium, tetra-butylammonium,tetrahexylammonium,didodecyldimethyl-ammonium

[82]

Methyl green None None 0.1 630 Co-ion(Tris)

0.8 fmol Lithium, sodium,potassium, caesium,magnesium, calcium

[25]

1,1’-Di-n-heptyl-4,4’-bipyridiniumhydroxide(DHBPOH)

Glycine ON 5 280 Counterion(glycine)

9–60 ppb Lithium, sodium,potassium, rubidium,caesium, magnesium,calcium, strontium,barium, manganese,cadmium, ammonium

[145]

Chrysoidine HIBA, lactate ON 5 230 Counterion(HIBA,lactate)

0.12–1.43 �M Potassium, barium,strontium, calcium,sodium, magnesium,manganese, chromium,iron, cobalt, lithium,lanthanum, cerium,praseodymium, neo-dymium, samarium,europium, gadolinium

[83]

18C6, 18-crown-6; Cry22, polyether cryptand-22

Table 6. Absorptivity data for probes for detection of anions

Probe Absorptivity (L�mol�1cm�1)

p-Aminobenzoate (1) 13 600, 264 nm, pH 9.6 [123]

Benzoate (1) 44 480, 194 nm, pH 6.5 [138](2) 11 900, 228 nm, pH 6.5 [30](3) 809, 254 nm, pH 8.0 [12]

o-Benzylbenzoic acid (1) 19 000, 228 nm, pH 6.5 [30]

3,4-Dimethoxycinnamic acid (1) 27 000, 310 nm, N/A [124]

p-Hydroxybenzoate (1) 10 299, 254 nm, pH 8.0 [12]

Phthalate (1) 9 950, 214 nm, pH 8.1 [55](2) 37 160, 196 nm, pH 6.5 [138](3) 1 357, 254 nm, pH 8.0 [12]

Pyromellitate (1) 26 200, 214 nm, pH 8.0 [55](2) 23 900, 214 nm, pH 6.5 [138](3) 7 062, 254 nm, pH 8.0 [12](4) 24 000, 210 nm, N/A [146]

2-Sulfobenzoic acid (1) 40 000, 228 nm, pH 6.5 [30]

5-Sulfosalicylic acid (1) 44 000, 210 nm, N/A [146](2) 22 000, 217 nm, N/A [146](3) 9 200, 226 nm, N/A [146]


2,6-Naphthalenedicarboxylate (1) 7 667, 254 nm, N/A [134](2) 10 020, 280 nm, N/A [134

1-Naphthylacetic acid (1) 81 100, 222 nm, pH 12.1 [136]

1,5-Naphthalenedisulfonate (1) 31 000, 214 nm, pH 8.1 [55](2) 16 420, 288 nm, N/A [137](3) 89 000, 224 nm, N/A [146]

2-Naphthalenesulfonate (1) 11 520, 206 nm, pH 6 [147]

1,3,6-Naphthalenetrisulfonate (1) 31 600, 214 nm, pH 8 [55](2) 7 750, 284 nm, N/A [137]

Bromocresol green (1) 45 000, 618 nm, pH 9.2 [33]

Chlorphenol red (1) 28 000, 578 nm, pH 6.5 [25](2) 33 000, 578 nm, pH 7.3 [25]

1,3-Dihydroxynaphthalene (1) 14 680, 256 nm, pH 12.1 [136]

Indigo carmine (1) 12 000, 620 nm, pH 4.2–7.3 [25]

Potassium indigotetrasulfonate (1) 32 100, 314 nm, pH 6.8 [33]

2,6-Pyridinedicarboxylate (1) 43 680, 192 nm, pH 6.5 [138]


Table 6. Continued


Riboflavin (1) 30 000, 267 nm, N/A [124]

Sorbate (1) 24 120, 254 nm, pH 12.1 [136](2) 28 800, 254 nm, pH 12.1 [148]

Orange G (1) 24 345, 248 nm, pH 7.7 [71](2) 19 511, 476 nm, pH 7.7 [71]

Naphthol yellow S (1) 23 125, 223 nm, pH 8 [70](2) 17 000, 428 nm, pH 8 [70]

Tartrazine (1) 21 350, 258 nm, pH 8 [70](2) 21 600, 426 nm, pH 8 [70]

Table 7. Absorptivity data for probes for detection ofcations


1,1’-Di-n-heptyl-4,4’-bipyridiniumhydroxide (DHBPOH)

(1) 24 000, 252 nm, N/A [149]

Methyl green (1) 15 000, 635 nm, pH 4.2 [25]

4-Aminopyridine (1) 18 500, 261 nm, N/A [150](2) 16 000, 210 nm, N/A [146]

Tetrazolium violet (1) 5 800, 300 nm, N/A [82](2) 18 000, 254 nm, N/A [82](3) 9 700, 280 nm, N/A [82]

Chrysoidine (1) 26 733, 453 nm, pH 4 [83]

4 Conclusions

Some general comments can be made regarding pub-lished approaches to indirect detection. (i) The majorityof work has been performed using low or moderatelyabsorbing probes. In the rare cases when highly absorb-ing probes have been used, they have been utilized at lowBGE concentrations. (ii) Most BGEs have been unbuf-fered or incorrectly buffered with the addition of co-ions.Very few BGEs have been buffered using an isoelectricampholytic buffer. (iii) Most detection has been performedat wavelengths in the low-UV region. There has been littleuse of detection of probes at longer wavelengths, suchas in the visible region. (iv) The small amount of work onthe detection of cations using highly absorbing probeshas been severely restricted as no complexing agentshave been used to allow separation of a range of metalions. (v) Concentration detection limits are generally atthe micromolar level. Potential areas which may meritfurther work include the use of alternative light sources,particularly LEDs, in conjunction with probes which ab-

sorb in the visible region, the use of highly absorbingprobes at increased concentrations and the developmentand synthesis of isoelectric ampholytic buffers.

Received January 23, 2003

5 References

[1] Pacakova, V., Stulik, K., J. Chromatogr. A 1997, 789, 169–180.

[2] Fritz, J. S., J. Chromatogr. A 2000, 884, 261–275.[3] Macka, M., Haddad, P. R., Electrophoresis 1997, 18, 2482–

2501.[4] Swinney, K., Bornhop, D. J., Electrophoresis 2000, 21,

1239–1250.[5] Doble, P., Haddad, P. R., J. Chromatogr. A 1999, 834, 189–

212.[6] Ackermans, M. T., Everaerts, F. M., Beckers, J. L., J. Chro-

matogr. 1991, 549, 345–355.[7] Kuhn, R., Hoffstetter-Kuhn, S., Capillary Electrophoresis:

Principles and Practice, Springer Laboratory 1993.[8] Rabiloud, T., Electrophoresis 1994, 15, 278–282.[9] Bruin, G. J. M., Van Asten, A. C., Xu, X., Poppe, H., J. Chro-

matogr. 1992, 608, 97–107.[10] Poppe, H., Anal. Chem. 1992, 64, 1908–1919.[11] Nielen, M. W. F., J. Chromatogr. 1991, 588, 321–326.[12] Cousins, S. M., Haddad P. R., Buchberger, W., J. Chroma-

togr. A 1994, 671, 397–402.[13] Buchberger, W., Cousins, S. M., Hadded, P. R., Trends Anal.

Chem. 1994, 13, 313–319.[14] Doble, P., Andersson, P., Haddad, P. R., J. Chromatogr. A

1997, 770, 291–300.[15] Collet J., Gareil, P., J. Chromatogr. A 1995, 716, 115–122.[16] Beckers, J. L., J. Chromatogr. A 1994, 679, 153–165.[17] Xu, X., Kok, W. T., Poppe, H., J. Chromatogr. A 1996, 742,

211–227.[18] Beckers, J. L., J. Chromatogr. A 1995, 693, 347–357.[19] Beckers, J. L., J. Chromatogr. A 1996, 741, 265–277.[20] Beckers, J. L., J. Chromatogr. A 1997, 764, 111–126.[21] Mikkers, F. E. P., Everaerts, P. M., Verheggen, T. P. E. M.,

J. Chromatogr. 1979, 169, 1–10.[22] Mikkers, F. E. P., Everaerts, F. M., Verheggen, T. P. E. M.,

J. Chromatogr. 1979, 169, 11–20.[23] Foret, F., Krivankova, L., Bocek, P., Capillary Zone Electro-

phoresis, VCH Publishers, Weinheim 1993.[24] Jandik, P., Bonn, G., Capillary Electrophoresis of Small

Molecules and Ions, VCH Publishers, Weinheim 1993.[25] Mala, Z., Vespalec, R., Bocek, P., Electrophoresis 1994, 15,

1526–1530.[26] Bazzanella, A., Lochmann, H., Mainka, A., Bachmann, K.,

Chromatographia 1997, 45, 59–62.[27] Takeuchi, T., Yeung, E. S., J. Chromatogr. 1986, 370, 83–92.[28] Xue, Y. J., Yeung, E. S., Anal. Chem. 1993, 65, 2923–2927.[29] Macka, M., Paull, B., Andersson, P., Haddad, P. R., J. Chro-

matogr. A 1997, 767, 303–310.[30] Ma, Y. F., Zhang, R. L., J. Chromatogr. 1992, 625, 341–348.[31] Steiner, F., Beck, W., Engelhardt, H., J. Chromatogr. A 1996,

738, 11–23.[32] Weston, A., Brown, P. R., Jandik, P., Heckenberg, A. L.,

Jones, W. R., J. Chromatogr. 1992, 608, 395–402.


[33] Doble, P., Macka, M., Haddad, P. R., J. Chromatogr. A 1998,804, 327–336.

[34] Foret, F., Fanali, S., Ossicini, L., Bocek, P., J. Chromatogr.1989, 470, 299–308.

[35] Beck, W., Engelhardt, H., Chromatographia 1992, 33, 313–316.

[36] Kenndler, E., Friedl, W., J. Chromatogr. 1992, 608, 161–170.[37] Zhu, T., Sun, Y. L., Zhang, C. X., Ling, D. K., Sun, Z. P., J. High

Resolut. Chromatogr. 1994, 17, 563–564.[38] Strege, M. A., Lagu, A., J. Liq. Chromatogr. 1991, 16, 51–68.[39] Macka, M., Andersson, P., Haddad, P. R., Anal. Chem. 1998,

70, 743–749.[40] Corstjens, H., Billiet, H. A. H., Frank, J., Luyben, K. C. A. M.,

Electrophoresis 1996, 17, 137–143.[41] Bello, M. S., J. Chromatogr. A 1996, 744, 81–91.[42] Henshall, A., Harrold, M. P., Tso, J. M. Y., J. Chromatogr.

1992, 608, 413–419.[43] Thompson, C. O., Trenerry, V. C., Kemmery, B., J. Chroma-

togr. A 1995, 704, 203–210.[44] Soga, T., Imaizumi, M., Electrophoresis 2001, 22, 3418–

3425.[45] Neubert, R., Raith, K., Schiewe, J., Pharmazie 1997, 52,

212–215.[46] Oehrle, S. A., Bossle, P. C., J. Chromatogr. A 1995, 692,

247–252.[47] Buscher, B. A. P., Tjaden, U. R., Irth, H., Andersson, E. M.,

van der Greef, J., J. Chromatogr. A 1995, 718, 413–419.[48] Santoyo, E., Garcia, R., Martinez-Frias, J., Lopez-Vera, F.,

Verma, S., J. Chromatogr. A 2002, 956, 279–286.[49] Goebel, L. K., McNair, H. M., Rasmussen, H. T., McPherson,

B. P., J. Microcol. Sep. 1993, 5, 47–50.[50] Shamsi, S. A., Danielson, N. D., Anal. Chem. 1995, 67,

1845–1852.[51] Shamsi, S. A., Danielson, N. D., Anal. Chem. 1995, 67,

4210–4216.[52] Pianetti, G. A., Taverna, M., Baillet, A., Mahuzier, G., Baylocq

Ferrier, D., J. Chromatogr. 1993, 630, 371–377.[53] Hepler, R. W., Yu Ip, C. C., J. Chromatogr. A 1994, 680, 201–

208.[54] Wang, T., Hartwick, R. A., J. Chromatogr. 1992, 589, 307–

313.[55] Motellier, S., Gurdale, K., Pitsch, H., J. Chromatogr. A 1997,

770, 311–319.[56] Larsen, K. L., Mathiesen, F., Zimmermann, W., Carbohydr.

Res. 1997, 298, 59–63.[57] Roldan Assad, R., Gareil, P., J. Chromatogr. A 1995, 708,

339–350.[58] Nardi, A., Fanali, S., Foret, F., Electrophoresis 1990, 11, 774–

776.[59] Dose, E. V., Guiochon, G. A., Anal. Chem. 1991, 63, 1063–

1072.[60] Vogt, C., Vogt, J., Wittrisch, H., J. Chromatogr. A 1996, 727,

301–310.[61] Vogt, C., Werner, G., J. Chromatogr. A 1994, 686, 325–332.[62] Wang, T., Hartwick, R. A., J. Chromatogr. 1992, 607, 119–

125.[63] Dabek-Zlotorzynska, E., Dlouhy, J. F., J. Chromatogr. A

1994, 671, 389–395.[64] Rhemrev Boom, M. M., J. Chromatogr. A 1994, 680, 675–

684.[65] Ehmann, T., Bachmann, K., Fabry, L., Rufer, H., Pahlke, S.,

Kotz, L., Chromatographia 1997, 45, 301–311.[66] Harrold, M. P., Wojtusik, M. J., Riviello, J., Henson, P.,

J. Chromatogr. 1993, 640, 463–471.

[67] Wojtusik, M. J., Harrold, M. P., J. Chromatogr. A 1994, 671,411–417.

[68] Rilbe, H., Electrophoresis 1992, 13, 811–816.

[69] Svensson, H., Acta Chem. Scand. 1962, 16, 456–466.

[70] Johns, C., Macka, M., Haddad, P. R., Electrophoresis 2000,21, 1312–1319.

[71] Johns, C., Shaw, M. J., Macka, M., Haddad, P. R., Electro-phoresis 2003, 24, 557–566.

[72] Macka, M., Andersson, P., Haddad, P. R., Electrophoresis1996, 17, 1898–1905.

[73] Walbroehl, Y., Jorgenson, J. W., J. Chromatogr. 1984, 315,135–143.

[74] Bruno, A. E., Gassmann, E., Pericle, N., Anton, K., Anal.Chem. 1989, 61, 876–883.

[75] Bruin, G. J. M., Stegeman, G., Van Asten, A. C., Xu, X.,Kraak, J. C., Poppe, H., J. Chromatogr. 1991, 559, 163–181.

[76] Cassidy, R., Janoski, M., LC-GC 1992, 10, 692–694, 696.

[77] Johns, C., Macka, M., Haddad, P. R., King, M., Paull, B.,J. Chromatogr. A 2001, 927, 237–241.

[78] Johns, C., Macka, M., Haddad, P. R., LC-GC 2003, 16,290–295.

[79] Siren, H., Maattanen, A., Riekkola, M. L., J. Chromatogr. A1997, 767, 293–301.

[80] Johns, C., Macka, M., Haddad, P. R., Electrophoresis 2002,23, 43–48.

[81] Butler, P. A. G., Mills, B., Hauser, P. C., Analyst 1997, 122,949–953.

[82] Shamsi, S. A., Danielson, N. D., J. Chromatogr. A 1996, 739,405–412.

[83] Johns, C., Macka, M., Haddad, P. R., J. Chromatogr. A, inpress.

[84] Aguilar, M., Huang, X., Zare, R. N., J. Chromatogr. 1989, 20,427–431.

[85] Aguilar, M., Farran, A., Martinez, M., J. Chromatogr. 1993,635, 127–131.

[86] Buchberger, W., Semenova, O. P., Timerbaev, A. R., J. HighResolut. Chromatogr. 1993, 16, 153–156.

[87] Buchberger, W., Haddad, P. R., J. Chromatogr. A 1994, 687,343–349.

[88] Timerbaev, A. R., J. Capil. Electrophor. 1995, 2, 165–174.

[89] Motomizu, S., Oshima, M., Matsuda, S., Obata, Y., Tanaka,H., Anal. Sci. 1992, 8, 619–625.

[90] Kajiwara, H., Sato, A., Kaneko, S., Biosci. Biotechnol. Bio-chem. 1993, 57, 1010–1011.

[91] Desbene, P. L., Morin, C., Spectra 2000 [Deux Mille] 1994,181, 35–41.

[92] Timerbaev, A. R., Semenova, O. P., Bonn, G. K., Analyst1994, 119, 2795–2799.

[93] Desbene, P. L., Morin, C. J., Desbene Monvernay, A. M.,Groult, R. S., J. Chromatogr. A 1995, 689, 135–148.

[94] Timerbaev, A. R., Semenova, O. P., J. Chromatogr. A 1995,690, 141–148.

[95] Wang, T. L., Li, S. F. Y., J. Chromatogr. A 1995, 707, 343–353.

[96] Colburn, B. A., Starnes, S. D., Sepaniak, M. J., Hinton, R.,Sep. Sc. Technol. 1995, 30, 1511–1527.

[97] Haumann, I., Baechmann, K., J. Chromatogr. A 1995, 717,385–391.

[98] Timerbaev, A. R., Semenova, O. P., Buchberger, W., Bonn,G. K., Fresenius’ J. Anal. Chem. 1996, 354, 414–419.

[99] Semenova, O. P., Timerbaev, A. R., Gaegstadter, R., Bonn,G. K., J. High Resolut. Chromatogr. 1996, 19, 177–179.


[100] Norden, M., Dabek Zlotorzynska, E., J. Chromatogr. A1996, 739, 421–429.

[101] Schaffer, S., Gareil, P., Dezael, C., Richard, D., J. Chroma-togr. A 1996, 740, 151–157.

[102] Conradi, S., Vogt, C., Wittrisch, H., Knobloch, G., Werner,G., J. Chromatogr. A 1996, 745, 103–109.

[103] Patterson, D. H., Harmon, B. J., Regnier, F. E., J. Chroma-togr. A 1994, 662, 389–395.

[104] Timerbaev, A. R., Semenova, O. P., Jandik, P., Bonn, G. K.,J. Chromatogr. A 1994, 671, 419–427.

[105] Motomizu, S., Kuwabara, M., Oshima, M., Bunseki Kagaku1994, 43, 621–628.

[106] Liu, Y., Lopez Avila, V., Zhu, J. J., Wiederin, D. R., Beckert,W. F., Anal. Chem. 1995, 67, 2020–2025.

[107] Xu, J., Ma, Y. F., J. Microcol. Sep. 1996, 8, 137–143.[108] Foret, F., Fanali, S., Nardi, A., Bocek, P., Electrophoresis

1990, 11, 780–783.[109] Bondoux, G., Jandik, P., Jones, W. R., Spectra 2000 [Deux-

Mille] 1991, 158, 47–51.[110] Jandik, P., Jones, W. R., Weston, A., Brown, P. R., LC-GC

1991, 9, 634–636, 638, 640, 642, 644.[111] Weston, A., Brown, P. R., Jandik, P., Jones, W. R., Hecken-

berg, A. L., J. Chromatogr. 1992, 593, 289–295.[112] Chen, M., Cassidy, R. M., J. Chromatogr. 1993, 640, 425–

431.[113] Shi, Y., Fritz, J. S., J. Chromatogr. 1993, 640, 473–479.[114] Lin, T. I., Lee, Y. H., Chen, Y. C., J. Chromatogr. A 1993,

654, 167–176.[115] Corr, J. J., Anacleto, J. F., Anal. Chem. 1996, 68, 2155–

2163.[116] Haumann, I., Boden, J., Mainka, A., Jegle, U., J. Chroma-

togr. A 2000, 895, 269–277.[117] Padarauskas, A., Olsauskaite, V., Schedt, G., J. Chroma-

togr. A 1998, 800, 369–375.[118] Padarauskas, A., Olsauskaite, V., Paliulionyte, V., J. Chro-

matogr. A 1998, 829, 359–365.[119] Kuban, P., Karlberg, B., Anal. Chem. 1998, 70, 360–365.[120] Xiong, X., Li, S. F. Y., Electrophoresis 1998, 19, 2243–2251.[121] Xiong, X., Li, S. F. Y., J. Chromatogr. A 1998, 822, 125–136.[122] Raguenes, C., Xiong, X., Lee, H. K., Li, S. F. Y., J. Liq. Chro-

matogr. Relat. Technol. 1999, 22, 2353–2365.[123] Roeder, A., Bachmann, K., J. Chromatogr. A 1995, 689,

305–311.[124] Xu, X., Kok, W. T., Poppe, H., J. Chromatogr. A 1995, 716,

231–240.[125] Jones, W. R., J. Chromatogr. 1993, 640, 387–395.

[126] Guan, F. Y., Wu, H. F., Luo, Y., J. Chromatogr. A 1996, 719,421–426.

[127] Chen, H., Xu, Y., Van Lente, F., Ip, M. P. C., J. Chromatogr. B.1996, 679, 49–59.

[128] Volgger, D., Zemann, A. J., Bonn, G. K., Antal, M. J.,J. Chromatogr. A 1997, 758, 263–276.

[129] Arellano, M., Andrianary, J., Dedieu, F., Couderc, F., Puig,P., J. Chromatogr. A 1997, 765, 321–328.

[130] Dabek-Zlotorzynska, E., Dlouhy, J. F., Houle, N., Pie-chowski, M., Ritchie, S., J. Chromatogr. A 1995, 706, 469–478.

[131] Hiissa, T., Siren, H., Kotiaho, T., Snellman, M., Hautojarvi,A., J. Chromatogr. A 1999, 853, 403–411.

[132] Damm, J. B. L., Overklift, G. T., J. Chromatogr. A 1994, 678,151–165.

[133] Tindall, G. W., Wilder, D. R., Perry, R. L., J. Chromatogr.1993, 641, 163–167.

[134] Dabek-Zlotorzynska, E., Dlouhy, J. F., J. Chromatogr. A1994, 685, 145–153.

[135] Dabek-Zlotorzynska, E., Piechowski, M., McGrath, M., Lai,E. P. C., J. Chromatogr. A 2001, 910, 331–345.

[136] Lee, Y. H., Lin, T. I., J. Chromatogr. B 1996, 681, 87–97.[137] Shamsi, S. A., Danielson, N. D., Anal. Chem. 1994, 66,

3757–3764.[138] Soga, T., Ross, G. A., J. Chromatogr. A 1997, 767, 223–

230.[139] Klockow, A., Paulus, A., Figueiredo, V., Amado, R., Widmer,

H. M., J. Chromatogr. A 1994, 680, 187–200.[140] Vorndran, A. B., Oefner, P. J., Scherz, H., Bonn, G. K.,

Chromatographia 1992, 33, 163–168.[141] Oefner, P. J., Vorndran, A. E., Grill, B., Huber, C., Bonn,

G. K., Chromatographia 1992, 34, 308–316.[142] Beck, W., Engelhardt, H., Fresenius’ J. Anal. Chem. 1993,

346, 618–621.[143] Shakulashvili, N., Faller, T., Engelhardt, H., J. Chromatogr.

A 2000, 895, 205–212.[144] Chiou, C. S., Shih, J. S., Analyst 1996, 121, 1107–1110.[145] Dabek-Zlotorzynska, B., Dlouhy, J. F., J. Chromatogr. A

1995, 706, 527–534.[146] Xiong, X., Li, S. F. Y., J. Chromatogr. A 1999, 835, 169–185.[147] Shamsi, S. A., Danielson, N. D., Anal. Chem. 1995, 67,

4210–4216.[148] Lu, B., Westerlund, D., Electrophoresis 1996, 17, 325–332.[149] Sato, H., J. Chromatogr. 1989, 469, 339–349.[150] Whang, K. S., Whang, C. W., Electrophoresis 1997, 18,

241–246.

enhancement of detection sensitivity for indirect photometric detection of anions and cations in...

Documents