Improvement of capillary zone electrophoresis sensitivity with artificial seawater as the background electrolyte utilizing transient isotachophoresis for the determination of nitrate and nitrate ions in seawater

We describe capillary zone electrophoresis (CZE) with transient isotachophoresis (ITP) for the determination of low concentrations of nitrite and nitrate ions in seawater. Bromide-free artificial seawater was adopted as background electrolyte (BGE) to eliminate the interference of high concentrations of salts in seawater. To reverse the electroosmotic flow (EOF), 3 mM cetyltrimethylammonium chloride (CTAC) was added to the BGE. High concentrations of chlorate were added to sample solutions as the terminating ion to generate the ITP process before the CZE separation. In general, the stacking effect increased with increasing amounts of chlorate injected into the capillary. The limits of detection (LODs) for nitrite and nitrate were 0.063 and 0.033 mg/L when the chlorate concentration was 600 and 200 mM, respectively; these were half of those obtained by CZE without the transient ITP. The LODs were obtained at a signal to noise ratio (S/N) of 3. The relative standard deviations (RSD, n = 10) of the peak areas for these ions were 3.2 and 2.9%. The RSDs of peak heights for these ions were 1.6 and 2.1%. The RSDs of migration times for these ions were 0.67 and 0.46%.     

Determination of amino acids in tobacco samples by capillary electrophoresis/indirect absorbance detection with isolation of the electrolysis compartment and p-Aminobenzoic acid as a background electrolyte.

A fast, convenient and sensitive method of capillary zone electrophoresis (CZE) and indirect UV detection was proposed for the determination of 16 amino acids. p-Aminobenzoic acid (PAB) was selected as a background electrolyte (BGE). An isolated cell included a BGE buffer part and an electrode buffer one, which were jointed with a glass frit. The isolated cell can prevent PAB from the electrode reaction and improve the stability of the detection baseline. The separation conditions of amino acids were investigated, such as different BGEs, BGE concentration, buffer pH and electroosmotic flow (EOF) modifiers. Under the selected separation conditions, 14 amino acid peaks could be separated in 12 min. The detection limits of the amino acids were in the range of 1.7 - 4.5 micromol/L. The isolated cell is suitable for reagents reacting on the electrodes in capillary electrophoresis. The proposed method has been successfully applied to the determination of the amino acids in tobacco samples.     

Speciation of inorganic and methylated arsenic compounds by capillary zone electrophoresis with indirect UV detection. Application to the analysis of alkali extracts of As2S2 (realgar) and As2S3 (orpiment)

A capillary zone electrophoresis (CZE) method with indirect UV detection was developed to simultaneously separate inorganic and organic arsenic compounds including arsenite (iAsIII), arsenate (iAsV), monomethylarsonate and dimethylarsenic acid (DMAV). 2,6-Pyridinedicarboxylic acid (PDC) and n-hexadecyltrimethylammonium hydroxide (CTAOH) were selected to compose a background electrolyte (BGE), where PDC was used as chromophore and CTAOH functioned as electroosmotic flow (EOF) modifier to reduce/eliminate EOF. The choice of detection wavelength, the optimization of BGE pH, and effects of applied electric field strength and temperature on separation were further investigated. The limits of detection for the targeted analytes were between 0.19 and 0.23 ppm as molecule. Good linearity of more than three orders of magnitude was obtained. Repeatability of migration times and peaks areas were 0.8-1.7 and 3.4-6.9% R.S.D.; whereas reproducibility were 1.2-2.2 and 3.6-7.1% R.S.D., respectively. The established CZE method was then applied to analyze the alkali extracts of realgar (As2S2) and orpiment (As2S3). The main components in both alkali extracts were identified to be iAsIII and iAsV.     

Multiple effect of surfactants used as additives in background electrolytes in capillary zone electrophoresis: cetyltrimethylammonium bromide as example of model surfactant

Surfactants are frequently used in the preparation of background electrolytes (BGEs) in capillary zone elcetrophoresis (CZE) in order to affect and to optimize both the electroosmotic flow (EOF) and the separation process. Their effects are, however, always multiple, the resulting situation may be very complex and the separation process may even be destroyed. We use the surfactant cetyltrimethylammonium bromide (CTAB) as a model example and bring experimental results and related discussion which elucidate the multiple effect of surfactants in an integrated way. It is shown that even at concentration levels lower than 10(-4) M CTAB strongly reduces the cathodic EOF in bare fused-silica capillaries and converts it into anodic EOF. The magnitude and polarity of the EOF depends not only on the concentration of CTAB but also on the composition of BGEs used. The interactions of CTA cations with the bare capillary wall reduce sorption of cationic analytes and enables their analysis. CTA cations at levels below their critical micelles concentration (CMC) already interact with anionic analytes and reduce their mobilities. This association is strong with highly charged anions and by this, the reversal of the EOF, applying BGEs with highly charged anions is less effective. These interactions are competitive and also depend on the composition of the BGE used. At levels above its CMC, CTAB forms micelles and enables the application of the micellar electrokinetic capillary chromatography (MEKC) mode and the analysis of, e.g., neutral components. Simultaneously, it is shown that the presence of CTAB may increase the number of potentially formed system zones.     

The preparation of background electrolytes in capillary zone electrophoresis: golden rules and pitfalls

In this article the methodology of the design of suitable background electrolytes (BGEs) in capillary zone electrophoresis (CZE) is described. The principal aspects of the role of a BGE in CZE are discussed with respect to an appropiate migration behavior of analytes, including the transport of the electric current, the buffering of pH, the Joule heat, the electro-endosmotic flow (EOF) and the principal migration and detection modes. The impact of the composition of the BGE upon migration and detection is discussed. It is shown that the total concentration of the BGE is a principal factor and the adjustment of migrating analyte zones according to the Kohlrausch regulating function (KRF) is the principal effect in most of the sample stacking techniques. The number of co-ions and their properties are of key importance for peak shapes of the analyte peaks and for the existence of system zones. The detection of UV-transparent analytes may advanteously be done in the indirect UV mode, by using UV-absorbing co-ions, however, both peaks and dips may be expected in the UV trace in case of multiple co-ionic BGEs. Properties of BGEs can be predicted applying mathematical models and it is shown that with SystCharts, predictions can be given concerning the existence of system zones, detection modes and the peak shapes of analytes for a given BGE. Practical examples of methodological considerations are given in the design of suitable BGEs for four principal combinations of migration and detection modes. The properties of the BGEs selected are exemplified with experimental results. Golden rules are summarized for the preparation of suitable BGEs in CZE.     

Co‐electroosmotic capillary electrophoresis of basic proteins with 1‐alkyl‐3‐methylimidazolium tetrafluoroborate ionic liquids as non‐covalent coating agents of the fused‐silica capillary and additives of the electrolyte solution

The paper reports the results of a study carried out to evaluate the use of three 1‐alkyl‐3‐methylimidazolium‐based ionic liquids as non‐covalent coating agents for bare fused‐silica capillaries and additives of the electrolyte solutions (BGE) for CE of basic proteins in the co‐EOF separation mode. The three ionic liquids are differentiated from each other by the length of the alkyl group on the imidazolium cation, consisting of either an ethyl, butyl or octyl substituent, whereas tetrafluoroborate is the common anionic component of the ionic liquids. Coating the capillary with the ionic liquid resulted in improved peak shape and protein separation, while the EOF was maintained cathodic. This indicates that each ionic liquid is effective at masking the protein interaction sites on the inner surface of the capillary, also when its adsorption onto the capillary wall has not completely neutralized all the negative charges arising from the ionization of the silanol groups and the ionic liquid is not incorporated into the BGE employed for separation. Using the coated capillaries with BGE containing the ionic liquid employed for the coating, at concentration low enough to maintaining the EOF cathodic, both peak shape and protein separation varied to different extents, based on the particular ionic liquid used and its concentration. Fast and efficient separation of the model basic protein mixture in co‐electroosmotic CE is obtained with the 1‐butyl‐3‐methylimidazolium tetrafluoroborate coated capillary and 100 mM acetate buffer (pH 4.0) containing 4.4 mM 1‐butyl‐3‐methylimidazolium tetrafluoroborate as the BGE.     

DNA fragment separations by on‐line combination of capillary isotachophoresis–capillary zone electrophoresis with UV detection

A high‐speed DNA fragment separation system based on an on‐line combination of capillary ITP with CZE (CITP‐CZE) and using UV detection at 260 nm was developed. A novel CITP‐CZE buffer system of pH 6.1 was designed for the separation of ten DNA fragments with sizes ranging from 100 to 1000 bp. An effect of underivatized α‐, β‐ and γ‐cyclodextrins on the resolution of DNA fragments in the CZE step of the CITP‐CZE combination was systematically investigated. Methylhydroxyethylcellulose present in the BGE was used to eliminate the EOF. DNA ladder fragments were separated within 10 min with LODs in the range of 1–5 ng/μL (S/N = 3). The RSDs of the migration time and peak area of individual DNA fragments were in the range of 1–3 and 3–9%, respectively. The developed CITP‐CZE system was further applied to the analysis of digest plasmid DNA samples.     

Capillary zone electrophoresis of glycopeptides under controlled electroosmotic flow conditions coupled to electrospray and matrix-assisted laser desorption/ionization mass spectrometry

Defined conditions of EOF along with different pH values of the BGE were compared for the purpose of analyzing glycopetides by CZE coupled to MS (CZE-MS). Hyphenation to MS involved ESI as well as MALDI, and single-stage and multistage MS were applied. Variation of the EOF was accomplished by selecting appropriate coatings for the capillary, namely hexadimethrine bromide (HDMB) and HDMB/dextran sulfate. A high and reproducible anodic and cathodic EOF, respectively, was obtained in both approaches, allowing the detection of analytes with net positive as well as negative charge in one single run. Thus, a fast and sensitive determination of the glycopeptides in a tryptic digest of antithrombin, chosen as a test sample, was achieved. Ionization suppression effects, a phenomenon typically observed with glycopeptides in MS analysis, were minimized thanks to separation from other peptides present. The high stability of the coatings permitted the generation of mass spectra without interfering peaks originating from the coating polymers. The high EOF generated by the coatings facilitated the maintenance of a stable spray when coupling to ESI-MS, and a stable CZE current when working with a sheath flow-assisted analyte deposition onto MALDI targets, respectively. In conclusion, CZE-MS could be demonstrated as a robust complementary method to capillary RP-HPLC-MS in combination with both soft-ionization techniques, ESI and MALDI, generally, and particularly in the context of glycopeptide analysis.     

Capillary zone electrophoresis with electroosmotic flow controlled by external radial electric field.

A new way of regulation of electroosmotic flow (EOF) in capillary zone electrophoresis (CZE) by external electric field has been developed. A set of three high-voltage power supplies is used to form a radial electric field across the capillary wall. One power supply is applied in the usual way as a driving force of CZE and EOF to the ends of the inner capillary compartment dipped into the electrode vessels and filled with background electrolyte. Two power supplies are connected to the ends of the outer low-conductivity coating of the capillary which is formed by the dispersion of copolymer of aniline and p-phenylenediamine in polystyrene matrix. The difference between electric potentials on the outer capillary surface and inside the capillary determines the voltage of radial electric field across the capillary wall and affects the electrokinetic potential at the solid-liquid interface inside the capillary. The effect of magnitude and polarity of external radial electric field on the flow rate of EOF, on the migration times of charged analytes and on the separation efficiency and resolution of CZE separations of synthetic oligopeptides, diglycine, triglycine and octapeptide fragments of human insulin was evaluated. Through the EOF control by external electric field the dynamic effective length of the capillary was obtained and the speed of analysis and resolution of CZE separations of peptide analytes could be optimized.     

Determination of adenosine deaminase activity in human erythrocytes by on-column capillary isotachophoresis-capillary zone electrophoresis in the presence of electroosmotic flow

Transient capillary isotachophoresis (CITP)-capillary zone electrophoresis (CZE) in presence of electroosmotic flow (EOF) was utilized for the measurement of adenosine deaminase activity in human erythrocytes. Phosphates, dominant anions of the sample matrix, were used as leading ions for transient isotachophoresis, and borates (0.3 M, pH 10) were used as terminating ions and background electrolyte for CZE. Final experimental conditions made it possible to inject 70% of the total capillary volume (1.45 microL) with the sample. Enzymatic conversion products (inosine and hypoxanthine), present in the sample in the low-micromolar range, were determined using optimized conditions. The limit of detection was 28 nM using UV detection at 202 nm. The presented data shows that CITP-CZE can be performed in uncoated capillaries in the presence of strong EOF.     

Applications of a sulfonated-polymer wall-modified open-tubular fused-silica capillary in capillary zone electrophoretic separations

A fused-silica capillary that is wall-modified via chemically bonding a sulfonated polymer to the capillary wall has a uniform negative charge density on its surface and produces an electroosmotic flow (EOF) greater than 4 x 10(-4) cm2 V(-1) s(-1) The EOF is nearly independent of buffer pH over the pH range of 2 to 10 and is lower than the EOF obtained for the bare fused-silica capillary at the more basic pH but is higher at the more acidic buffer pH. Optimization of buffer pH can be based on analyte pKa values to improve the overall quality of the capillary zone electrophoresis (CZE) separation of complex mixtures of weak acid and base analytes. Because of the high EOF in an acidic buffer, the capillary is useful for the separation of weak organic bases which are in their cation forms in the acidic buffer. EOF for the sulfonic acid bonded phase capillary can be adjusted via buffer additives such as organic solvent, tetraalkylammonium salts, multivalent cations and alkylsulfonic acids. The advantages of utilizing buffer pH and the EOF buffer modifiers to enhance migration time, selectivity, and resolution in CZE separations with this capillary are illustrated using a series of test analyte mixtures of inorganic anions, carboxylic acids, alkylsulfonic acids, benzenesulfonic acids, sulfas, pyridines, anilines or small-chain peptides.     

System zones in capillary zone electrophoresis.

This paper brings an overview of system zones (SZs) in capillary zone electrophoresis (CZE) and their effects upon the migration of zones of analytes. It is shown that the formation and migration of SZs is an inherent feature of CZE, and that it depends predominantly on the composition of an actual background electrolyte (BGE). One can distinguish between stationary SZs and migrating SZs. Stationary SZs, which move due to the electroosmotic flow only, are induced in any BGE by sample injection. Migrating SZs may be induced by a sample injection in BGEs which show at least one of the following features: (i) BGE contains two or more co-ions, (ii) BGE has low or high pH whereby H+ or OH- act as the second co-ion, and (iii) BGE contains multivalent weak acids or bases. SZs do not contain any analyte and show always BGE-like composition. They contain components of the BGE only and the concentrations of these components are different from their values in the original BGE. Providing that some of the ionic components of the BGE are visible by the detector, the migrating SZs can be detected and they are present as system peaks/dips in the electropherogram. It is shown that a migrating SZ may be characterized by its mobility, and examples are given how this mobility can depend on the composition of the BGE. Further, the effects of the migrating SZs (either visible or not visible by the detector) upon the zones of analytes are presented and the typical disturbances of the peaks (extra broadening, zig-zag form, schizophrenic behavior) are exemplified and discussed. Finally, some conclusions are presented how to cope with the SZs in practice. The proposed procedure is based on the theoretical predictions and/or measurements of the mobilities of SZs and on the so-called unsafe region. Then, such operational conditions should be selected where the unsafe region is outside of the required analytical window.     

Simultaneous determination of deoxycytidine diphosphate and deoxycytidine triphosphate by capillary electrophoresis with transient isotachophoretic stacking: a sensitive monitoring method for ribonucleotide reductase activity

A simple and rapid capillary electrophoretic method was developed for simultaneous determination of sub‐micromolar 2′‐deoxycytidine 5′‐diphosphate (dCDP) and 2′‐deoxycytidine 5′‐triphosphate (dCTP) levels in enzyme assays without using radioactively labeled substrates. The separation was performed at 25°C using MES in the BGE as the terminating ion, the chloride ions in the sample buffer as the leading ion, and PEG 4000 in the BGE as the EOF suppressor for sample stacking by transient isotachophoresis (tITP). Several parameters affecting the separation were investigated, including the pH of the BGE, the concentration of sodium chloride in the sample buffer, and the concentrations of MES and PEG 4000 in the running buffer. Good separation with high separation efficiency was achieved within 6 min under optimal conditions. In comparison with the simple CZE method, the present tITP‐CZE method enabled a 150‐fold increase in the injection time without any decrease in resolution and the sensitivity was enhanced up to two orders of magnitude with the new method. The linear range of the method was 0.1–10 μM for dCDP and dCTP. The limits of detection of dCDP and dCTP were 85 and 73 nM, respectively. The proposed method was successfully applied for the activity assay of ribonucleotide reductase from Hep G2 and Sf9 cells.     

Electroosmotic pump-assisted capillary electrophoresis of proteins

A new method for protein analysis, that is, electroosmotic pump-assisted capillary electrophoresis (EOPACE), is developed and demonstrated to possess several advantages over other CE-based techniques. The column employed in EOPACE consists of two linked sections, poly(vinyl alcohol) (PVA)-coated and uncoated capillaries. The PVA-coated capillary column is the section for protein electrophoresis in EOPACE. Electroosmotic flow (EOF) is almost completely suppressed in this hydrophilic polymer coated section, so protein electrophoresis in the PVA-modified capillary is free of irreversible protein adsorption to the capillary inner wall. The uncoated capillary section serves as an electroosmotic pump, since EOF towards cathode occurs at neutral pH in the naked silica capillary. By the separation of a protein mixture containing cytochrome c (Cyt-c), myoglobin and trypsin inhibitor, we have demonstrated the advantages of EOPACE method over other relevant ones such as pressure assisted CE, capillary zone electrophoresis (CZE) with naked capillary and CZE with PVA-coated capillary. A significant feature of EOPACE is that simultaneous separation of cationic, anionic and uncharged proteins at neutral pH can be readily accomplished by a single run, which is impossible or difficult to realize by the other CE-based methods. The high column efficiency and good reproducibility in protein analysis by EOPACE are verified and discussed. In addition, separation of tryptic digests of Cyt-c with the EOPACE system is demonstrated.     

逍遥丸的毛细管电泳指纹图谱的建立

采用毛细管区带电泳法建立了逍遥丸(Xiaoyao Pill, XYP)的毛细管电泳指纹图谱(CEFP)。运用正方形优化法,以色谱指纹图谱分离量指数(RF)为优化的目标函数,对建立指纹图谱的实验条件进行了优化,确定了最佳背景电解质(BGE)溶液50 mmol/L硼砂-50 mmol/L磷酸氢二钠-150 mmol/L磷酸二氢钠-50 mmol/L碳酸氢钠(1:1:1:5, v/v/v/v; pH 7.40)、紫外检测波长228 nm、运行电压12 kV、重力进样25 s (高度14 cm)的分离检测条件。采用未涂层石英毛细管(70 cm×75 μm,有效分离长度57 cm)分离,以咖啡酸色谱峰为参照,确定13批逍遥丸样品的21个共有指纹峰。通过聚类分析确定用其中10批样品生成对照CEFP,以此为标准用系统指纹定量法鉴别13批逍遥丸的质量,结果显示: S3号样品的化学成分数量和分布比例不合格,S10和S12号样品含量明显偏高,其余批次质量均合格。所建立的正方形优化法操作简便,适用于中药的毛细管区带电泳BGE的选择;所建立的逍遥丸CEFP具有较好的精密度和重现性,可以为逍遥丸的质量控制提供新的参考。     

Ampholytes as background electrolytes in capillary zone electrophoresis: sense or nonsense? Histidine as a model ampholyte

A lot of phenomena, occuring in capillary zone electrophoresis (CZE), are linked with the ionic concentration of the background electrolyte (BGE). If weak bases and acids are used as BGEs in CZE, at a pH where they are scarcely ionized, the ionic concentration of the BGE is very low and this brings a strong peak broadening, limited sample stacking and low sample load. Because the electromigration dispersion increases extremely, moreover, the existence of low-conductivity BGEs in CZE is a contradiction in terms. The behavior of ampholytes as BGE in CZE is examined, by means of histidine as a model ampholyte. For BGEs consisting of histidine, important parameters, including the ionic concentrations, buffer capacity, transfer ratio, and the indicator for electromigration dispersion E(1)m(1)/E(2)m(2), are calculated at various pH. Although the transfer ratio is fairly constant over the whole pH traject, the ionic concentration and buffer capacity decrease whereas the electromigration dispersion strongly increases near the pI of histidine. I.e., that ampholytes can be applied as BGEs in CZE, however, just not at pH near their pI value, except as the difference between the pK values of the basic and acidic group, the deltapK value, is very small. For ampholytes with a low deltapK value or at high concentrations, all the before-mentioned effects are less fatal, but in that case we can not speak of a real low-conductivity BGE. If ampholytes are used at pH near their pK values, the use of ampholytes as BGE is not advantageously compared with simple weak bases and acids. This has been confirmed by calculations and experiments.     

Position and intensity of system (eigen) peaks in capillary zone electrophoresis

The intensity of system (or eigen) peaks encountered in capillary zone electrophoresis (CZE) can be predicted by considering mass balances for each of the analyte constituents and each of the constituents in the background electrolyte (BGE). As a result of coherence, in each zone the proportions in which the constituent concentrations vary are fixed; they are determined by the composition of the BGE and the nature of the analyte constituent (if present) and described as eigenvectors of a transport matrix. Considering the effect of an injection, the mass balances for all constituents can be satisfied only via the intensity of each zone. This leads to an n-equations, n-unknowns problem, with the intensities as the unknowns and the mass balances as equations.The latter can be easily solved to obtain the intensities of the zones, of analytes as well as of system peaks. In this work the approach has been applied to CZE systems with two co-ions in the BGE, and experimental results have been compared to the predictions obtained from the model. Agreement was seen to be reasonable, but the quantitative comparison often failed, probably due to experimental difficulties.     

Success and failure with phthalate buffers in capillary zone electrophoresis

Phthalate buffers are currently used in capillary electrophoresis as robust electrolyte systems for indirect detection. This contribution demonstrates that these buffers show regularly not only successful regions of mobilities of analytes (sample window) but also regions of failure where the migration of analytes is strongly deteriorated due to the presence of a system zone. System zones in phthalate buffers may be easily detected by UV detection and manifest themselves as peaks or dips. Peak shape diagrams are advantageously used for the prediction of the migration behavior of system zones in phthalate background electrolyte (BGE) systems at various pH. It is shown that the mobility of the system zone varies strongly with pH, is practically zero at pH values below 4 and above 7, and shows a maximum at pH 5. Thus, the system peak may coincide either with the peaks of various analytes or with the electroosmotic flow (EOF) peak. Experiments are given showing the effects of such coincidences as, e.g., zigzag detection patterns, double EOF peaks, and/or unusually broad peaks/dips. The message of this contribution is to show how to understand the electrophoretic properties of phthalate BGEs that, regardless of possible failure regions, may be successfully used in the analytical practice of capillary zone electrophoresis (CZE).     

Theory of system zones in capillary zone electrophoresis

In the last years, it has been shown that the formation and migration of system zones is an inherent feature of capillary zone electrophoresis (CZE) and that it depends predominantly on the composition of an actual background electrolyte (BGE). In most of the currently used BGEs, the SZs are invisible by the UV absorbance detection system, however, the comigration of SZs with the zones of analytes deteriorates the analytical performance of CZE and may be fatal for its utilization. Therefore, the theoretical predictions of the existence and migration of SZs is of key importance for the expediency of CZE. This is a review of the theoretical treatments of SZs which reveals the origin and the properties of SZs and shows how to cope with them. Also, a table of some typical BGEs is presented where the existence and mobilities of SZs are given.     

Modification of the electroosmotic flow and separation selectivity of anions in electrochromatography with pseudo-stationary phases of C14-alkyldimethylammoniopropane sulfonate zwitterionic surfactants by addition of salts to the background electrolyte

The effects of salts (NaCl, NaClO4, MgCl2, CeCl3) added to background electrolyte (BGE) solutions (10 mmol L(-1) sodium phosphate, pH 7.2) on electroosmotic flow (EOF) and the separation selectivity of anions (chloride, bromide, iodide, nitrite, nitrate, chlorate, thiocyanate, iodate, chromate, and molybdate ion) by capillary electrochromatography using the zwitterionic surfactant 3-(N,N-dimethylmyristylammonio)propane sulfonate (C14N3S) as a pseudo-stationary phase were investigated. There are two mechanisms affecting the separations: 1. the cations and anions of the added salts interact with the zwitterionic surfactant to varying degrees, thus changing the overall retention of the analytes; and 2. they change the EOF and the resulting apparent mobilities. It was shown that a BGE containing perchlorate and a low concentration of zwitterionic surfactant (2 mmol L(-1)) gave a stable and reproducible EOF and the concentration of perchlorate could be used to manipulate the separation selectivity for polarizable anions, such as iodide and thiocyanate. These effects are discussed in terms of measured association constants describing the interaction of anions and cations with the zwitterion.