Planar Chromatography - Analytical Chemistry (ACS Publications)

---

Anal. Chem. 2000, 72, 9R-25R

Planar Chromatography Joseph Sherma

Department of Chemistry, Lafayette College, Easton, Pennsylvania 18042 Review Contents GENERAL CONSIDERATIONS History, Books, Reviews, and Student Experiments Theory and Fundamental Studies Chromatographic Systems (Stationary and Mobile Phases) Apparatus and Techniques Detection and Identification of Separated Zones Quantitative Analysis Preparative Layer Chromatography and Radio-Thin-Layer Chromatography APPLICATIONS Acids and Phenols Amino Acids, Peptides, and Proteins Antibiotics Bases and Amines Carbohydrates Dyes and Pigments Hydrocarbons Lipids Pesticides Pharmaceuticals, Drugs, and Alkaloids Purines, Pyrimidines, and Nucleic Acids Steroids Surfactants and Detergents Toxins Vitamins Miscellaneous Organic Compounds Inorganics and Metal Organics LITERATURE CITED

10R 10R 10R 11R 12R 12R 13R 13R 14R 14R 14R 15R 15R 15R 15R 16R 16R 17R 17R 19R 19R 19R 19R 19R 20R 20R 20R

This is a selective review of the literature of thin-layer chromatography (TLC) and high-performance thin-layer chromatography (HPTLC) in Chemical Abstracts from November 1, 1997 to November 1, 1999. The literature search was augmented by consulting Analytical Abstracts, Chemical Titles, and Current Contents, and the following important journals publishing papers on TLC were searched directly: Journal of Chromatography (parts A and B and the bibliography issues), Journal of Chromatographic Science, Chromatographia, Analytical Chemistry, Journal of Liquid Chromatography & Related Technologies, Journal of AOAC International, Journal of Planar Chromatography-Modern TLC, and Acta Chromatographica. Publications in the past two years on the theory, techniques, and applications of TLC continued at a high level. Only a very small number of papers reported new research in paper chromatography, the other main classification of planar chromatography, but none of these was considered to be important enough to be included in this review. A computer-based search of Chemical Abstracts found that approximately 1800 publications on TLC were abstracted in the review period. The attempt was made to cite 10.1021/a1000001z CCC: $19.00 Published on Web 03/25/2000

© 2000 American Chemical Society

only the most important publications describing significant advances in theoretical studies, methodology, instrumentation, and applications in this review. The review is mostly limited to journals easily accessible to U.S. scientists. This eliminates coverage of many papers in foreign-language journals, most notably papers written in Chinese. Abstract citations are given for cited references not published in English. Most TLC papers originated from laboratories outside of the United States, especially Europe and Asia, but were published in English. Several important articles were published in 1998 and 1999 that reviewed the state of the art and future prospects in TLC. Bariska et al. (1) pointed out that TLC is relatively simple and inexpensive and has advantages that include separation of multiple samples in parallel, the possibility of two-dimensional (2-D) separation, the use of specific and sensitive detection reagents enabling detection and identification of zones, reliable quantitative evaluation by use of instrumental scanners, and continuous observation of the separation process because of the open format of the layer. In his overview of TLC, Poole (2) emphasized the complimentary features of TLC and column liquid chromatography (HPLC) and suggested attributes that provide reasons for using TLC, including simultaneous, parallel separation of samples for low-cost, high-throughput screening; the disposable stationary phase that allows analysis of crude samples; the use of static and sequential detection methods for identification and confirmation without time constraints; the storage of chromatograms on the layer for archiving, evaluation in different locations or at different times, and convenient fraction collection for multimodal column/ layer chromatography; and integrity of the total sample within the chromatogram in the absence of elution as in a column. Poole discussed future prospects for improved separation performance in TLC using zone refocusing, forced-flow, and electroosmotic flow methods, as well as increasing zone capacity by 2-D development and coupling to column chromatographic methods; advances in coupling TLC with spectroscopic methods for structural elucidation; and predictions for how TLC will be practiced in the future. Rabel (3) described the increasing use of TLC for mass screening of samples, e.g., in laboratories performing combinatorial chemistry for the purpose of drug discovery, and the use of multiple chemical and biological visualization and instrumental spectrometric procedures for compound detection and identification, including transfer from a silica gel layer onto a polymer membrane for further characterization. Staples (4) described TLC products currently available from manufacturers for various applications, including normal-phase silica gel 60 plates with and without fluorescent indicator, hydrocarbon-impregnated plates for reversedphase (RP) TLC, spherical gel plates for direct Raman resonance spectrometry and scouting solvents for HPLC separations, plates with preadsorbent or preconcentrating zone for convenient and Analytical Chemistry, Vol. 72, No. 12, June 15, 2000 9R

rapid sample application, speciality plates such as DEAE ion exchange cellulose and ion exchange/silica gel, and semiautomated and fully automated sample applicators. The 10th International Symposium on Instrumental Planar Chromatography was held on May 16-19, 1998, in Visegrad, Hungary, organized by Professor Dr. Sz. Nyiredy, editor-in-chief of the Journal of Planar Chromatography-Modern TLC. The 10th anniversary of this journal as well as the 60th anniversary of thinlayer chromatography were celebrated at this meeting. The lectures presented at the meeting, which was reported on by Studer (5), included the following topics: a structure-driven retention model for method development in TLC, new specialized phases for coupling of TLC and Fourier transform infrared spectrometry (FT-IR) in pharmaceutical analysis, and interfaces to electrospray (ES) ionization for coupling TLC with mass spectrometry (MS). The next regular International Planar Chromatography Symposium is scheduled to be held in Interlaken, Switzerland, in June, 2000, having as a theme combined or coupled techniques involving TLC. The First International Meeting on Imaging Techniques in Planar Chromatography was held in Jezersko, Slovenia, May 14-16, 1999. The lectures presented at this symposium, which like those listed above illustrate many of the important current research areas in techniques and applications in TLC, were reviewed by Vovk (6). The topics of these lectures included the following: recent advances in hardware and software of digital camcorders for use in TLC; current usage and future trends in TLC-video imaging in the pharmaceutical industry; cameras, illumination systems, image quality, and resultant quantitative data; use of conventional color photography and color video technology for documentation of TLC and overpressured layer chromatography (OPLC) spots; photothermal radiometric measurement of TLC plates; applications of TLC for investigating compounds from herbal plants; formation of a collection of digitized images that can be transmitted through computer networks to each user for appropriate processing in qualitative and semiquantitative determination; multifield image analysis for automation of lane and spot separation and microdot analysis; and programs for quantification of the images captured by different data acquisition systems, e.g., different digital or CCD cameras. Method development courses and TLC instrumentation workshops were offered periodically in Wilmington, NC, by Camag. A bibliography service (CBS) is offered by Camag to keep subscribers informed about publications involving TLC. This service is available free of charge by mail from Camag. A cumulative compilation of abstracts from volumes 51-82 (May, 1983 through March, 1999) can be purchased from Camag on a CD-ROM that is searchable by key word (author name, substance, technique, reagent, etc.). In addition to a review of the literature, issues of the Camag CBS contain a section on applications, e.g., separation of carotenoids by HPTLC with automated multiple development (HPTLC-AMD), HPTLC determination of histamine in fish and fish products, and quantification of levamisol in tissue and organs of pigs in issue 83, September 1999. A large number of applications notes are listed and can be requested on the Camag website . Diverse information on TLC methods and products is available on-line by entering the phrase “thin layer chromatography” on a website search engine such as or . 10R

Analytical Chemistry, Vol. 72, No. 12, June 15, 2000

An honorary issue of the Journal of Planar ChromatographyModern TLC was dedicated to Professor Karl-Artur Kovar, a pioneer in HPTLC-FT-IR, on the occasion of his 60th birthday, and his distinguished career was reviewed in the issue by Nyiredy (7). GENERAL CONSIDERATIONS History, Books, Reviews, and Student Experiments. The fourth edition of Thin Layer Chromatography: Techniques and Applications by Fried and Sherma was published (A1). Review articles were published on the following topics: the history of planar chromatography (A2), structure-retention relationships and physicochemical characterization of solutes (A3), hyphenated techniques in TLC (A4), enantiomer separations by TLC (A5), and chromatographic analysis of explosives, including TLC (A6). Other pertinent reviews are cited in the sections below. The nomenclature and means of classifying the methods of planar chromatography were discussed (A7). TLC laboratory experiments for high school and college students were devised to illustrate the teaching of the basics of pharmaceutical analysis (A8), identification of polycyclic aromatic hydrocarbons (PAHs) in candle black (A9), detection of volatile compounds in building materials (A10), and a microscale experiment for monitoring a reaction for the preparation of a fluorescent dye (A11). Theory and Fundamental Studies. A new approach to the separation number concept in TLC based on the effective diffusion of zones was introduced and verified for test solutes on silica gel and alumina (B1) and later applied to polyamide and cellulose (B2). Topological indexes and RM values on silica gel and Mn salt-impregnated silica gel were used to predict the pKa values of isomeric methylanilines and chloroanilines (B3). Multistep gradient elution RP-TLC was used to model the HPLC separation of colored pigments in red wines (B4). Photoacoustic spectroscopy was applied to the depth profiling of zone concentration distribution in adsorbent layers and to obtain values of thermal diffusivity of the plates (B5). Research on quantitative structure-retention relationships in TLC was reviewed (B6). The following theoretical and experimental studies of solute retention and separation mechanisms in TLC were published: mixed tris-p-diketonato metal complexes on polyacrylonitrile sorbent (B7); 33 metal ions on silica gel (B8); thermodynamic study of selected macrocycles using RP-HPTLC with methanol-water mobile phases (B9); phenolic acids on silica, alumina, and polyamide (B10); phenols, aniline derivatives, and quinoline bases on silica, alumina, and Florisil layers with binary isoelutropic mobile phases (B11); nitrophenones and their reduced derivatives on these same layers with mixtures of ethyl acetate or 2-propanol and heptane as mobile phases (B12); 10 dihydroxythiobenzanilides on C-8 and C-18 chemically bonded silica gel RP layers (B13); several coumarins and flavonoids on silica gel and Florisil layers (B14); 31 samples of variously substituted 2,5-anhydroaldohexose ethylene acetal derivatives on silica gel and C-18 layers (B15); aldopentose and aldohexose derivatives on silica gel (B16); tosylated xylitol derivatives on silica gel, C-8, and C-18 layers (B17); 18 steroid drugs on alumina (B18); application of multivariate mathematical-statistical methods to

compare the RP-TLC and -HPLC of 16 mono- and ditetrazolium salts (B19); 10 antioxidants on RP-TLC plates (B20); 60 2,4dihydroxythiobenzanilides in an RP-TLC system using methanol as an organic modifier (B21); influence of the modifier and molecular structure of 10 dihydroxythiobenzalides on RP-HPTLC retention (B22); factors controlling retention of amidines on silica gel layers (B23); closely related N-phenylamide derivatives of phenoxyacetic acid on silica gel and diol-bonded HPTLC plates (B24); influence of ligand nature on retention of heterocyclic azonaphthols and their chelates on silica gel (B25); and the use of naphthylbismuthol as a reagent for the TLC of metal chelates on silica gel (B26). Theoretical TLC studies were reported for the selectivity of mixture separation with multicomponent mobile phases (B27); physicochemical modeling of solute retention in adsorption systems with B-N and B1-B2 mobile phases (B28); development of an empirical second- or third-order polynomial equation for accurate prediction of retention (B29); a new adsorption-partition model of solute retention on chemically bonded layers (B30); a structure-driven retention model for method development in RP-TLC on C-18 layers (B31); and prediction of retention using polar and nonpolar surface areas of single-solvent mobile phases (B32). The following studies of mobile-phase optimization were reported: designing the optimum TLC system for identification of 12 phenozines and tri- and tetracyclic antidepressants (B33); retardation behavior of cyanobacterial hepatotoxins in the irregular part of the PRISMA optimization model (B34); application of principal component analysis to the choice of the optimum solvent system for separation of PAHs (B35); optimization in normal-phase (NP)-TLC with chemically bonded 3-cyanopropyl stationary phase (B36); a simple approach to optimize multicomponent quaternary mobile phases in adsorption TLC (B37); use of a four-parameter equation based on an adsorption-partition mechanism for optimizing separations in RP-TLC (B38); a structure-driven retention model for solvent selection and optimization in RP-TLC (B39); a computer-assisted method for optimizing ternary mobile phases in HPTLC (B40); and a new mathematical model for optimization of the separation of 1,4-dibenzodiazepines and comparison with other models (B41). Various strategies for mobile-phase optimization applied to TLC were reviewed (B42). Lipophilicity is the molecular parameter most frequently used in quantitative structure-activity relationship studies. It has important implications in terms of pharmaceutical and pesticide activity because it governs the penetration of bioactive compounds through hydrophobic cell membranes and uptake by target organs or organisms. TLC methods for estimation of lipophilicity (hydrophobicity) by RP-TLC on impregnated or chemically bonded silica gel layers were applied to N-alkyl derivatives of 1,2,3,4tetrahydroquinoline (B43), new isoxazolylnaphthoquinones (B44), nonsteroidal antiinflammatory drugs (B45, B46), 56 surfactants (B47), fused-ring nitrogen heterocycles (B48), N-hydroxyethylamides of aryloxyalkylene and pyridinecarboxylic acids (B49), aliphatic isomers using a modification of Rekker’s equation (B50), ethanolamine amido esters (B51), photosystem II inhibitors (B52), and nine aromatic compounds (B53). Studies were reported for the evaluation of hydrophobicity by principal component analysis (B54) and the effect of salts in the mobile phase on the

hydrophobicity parameters determined for 11 sulfosuccinic acid esters (B55). Chromatographic Systems (Stationary and Mobile Phases). The great majority of TLC analyses are carried out using NP silica gel layers. Because most HPLC analyses are performed using RP sorbents, the methods are complementary for achieving separations and confirming qualitative and quantitative results. Applications of different sorbent layers are cited in the individual sections below. Plates are usually precleaned prior to sample application, especially for analysis at trace (ppb) levels. A two-step procedure involving ascending development with methanol followed by washing by immersion in the same solvent yielded layers that were essentially free from surface contamination (C1). Photoacoustic spectrometry was used to investigate secondary chromatographic effects (vertical concentration distribution) (C2) and the effect of drying conditions on the depth distribution (C3) of compound zones on TLC plates. Photothermal beam deflection spectrometry provided accurate values for surface and depth distribution of TLC zones (C4). Raman spectrometry with a highpower Nd:WVO4 laser was used to characterize C-18 stationary phases, and results were compared with aminopropyl, cyanopropyl, diol, C-8, and C-1 phases (C5). Triangular and trapezoidal HPTLC layers and conical HPLC columns were prepared and evaluated, and sorbent geometry was found to have a large influence on separation speed, efficiency, resolution, zone shape, and depth distribution of compounds inside the sorbent (C6). Advantages of new quasi-column capillary- and forced-flow TLC methods using narrow planar sorbent layers were demonstrated using synthetic dye mixtures (C7). The properties and applications of Florisil in TLC and HPLC were reviewed and compared with other adsorbents (C8). Analyses of food components on unconventional starch and talc layers having specific characteristics compared to silica gel were described (C9). The following new TLC sorbents were prepared, characterized, and applied to TLC separations of various compounds: chemically modified mineral perlite (C10), two dicationic zeolites (metal cation-organic cation) (C11), volcanic tuff reacted with (γaminopropyl)trimethoxysilane (C12), alumina oxide plates impregnated with copper phthalocyanine dye (C13), and polar silica gel R chemically modified with (mercaptopropyl)trimethoxysilane (C14). Chitin modified with ferric chloride was used as a stationary phase for immobilized affinity chromatography applied to protein purification (C15). Metal ions were separated and determined on silica gelsupported Sn(IV) arsenosilicate layers with methanol-containing mobile phases (C16), tributyl phosphate-impregnated layers prepared from silica gel and stannic arsenate gel mixture with 1 M aqueous potassium thiocyanate mobile phase (C17), and trin-butyl phosphate-modified silica gel G layers in acetone-nitric acid-water solvent systems (C18). TLC is being used to an increasing degree for separation of enantiomeric compounds, especially compounds of pharmaceutical interest, on chiral layers, layers impregnated with a chiral selector, or layers developed with chiral mobile phases. The resolution of enantiomers by TLC on impregnated layers was reviewed (C19), and the following papers are examples of enantiomeric separations: optical isomers of amino acids on modified chitin and Analytical Chemistry, Vol. 72, No. 12, June 15, 2000

11R

chitosan layers (C20); 2-arylpropionic acids using 2-D TLC on silica gel impregnated with (-)-brucine chiral selector (C21); adrenergic drugs on molecularly imprinted chiral stationary phases prepared with R-agonists (C22); quinine and cinchonine diastereomers on synthetic polymers imprinted with quinine as the chiral stationary phase (C23); and 2-amino-1-butanol enantiomers (C24) and racemates of fluoxetine, norfluoxetine, and promethazine (C25) by RP-TLC using cyclodextrins as mobilephase additives. The retention of metallocyanide complexes was studied on DEAE cellulose anion-exchange layers developed with sodium perchlorate solutions (C26), and micellar mobile phases were evaluated for the separation of metal 1,3-diketonates (C27). Nonionic surfactant mobile-phase additives enhanced fluorescence intensity and signal-to-noise ratio in the quantitative densitometric determination of biogenic amines in plant tissue and beer samples using a fiber-optic scanner (C28). Apparatus and Techniques. Strategies for method development in HPTLC were reviewed (D1). Sample preparation methods for TLC are generally similar to those for gas chromatography (GC) and HPLC except that cruder samples with minimal purification can often be analyzed because of the single use of disposable plates. Sample cleanup and separation are often accomplished during the same run, and irreversibly sorbed impurities remaining at the origin cause no problems as they would in HPLC. Solid-phase extraction (SPE) is among the most widely used modern sample preparation methods prior to TLC, as noted in the applications sections below. A method was described for preconcentrating nonvolatile and lowvolatility impurities prior to their analysis by volatilizing the main component (matrix) (D2). Saponins in plant material were analyzed by TLC combined with TAS, in which samples packed in a glass cartridge are heated in a TAS oven and the escaping volatile substances transferred to the plate origin (D3). There are a variety of manual and automatic devices available for sample application; automated application instruments are required for maximum accuracy and precision in quantitative analysis. A simple streaking device was described for sample application in preparative layer chromatography (D4). The effects of pH, ionic strength, and buffer concentration of the mobile phase on the Rf values of acidic compounds were studied in ion pair TLC (D5). Displacement TLC with mobile phases containing organic bases was applied to the separation of toad poison bufadienolides in traditional Chinese drugs (D6). Separations on microthin layers (2-5 µm particle size, 50-70 µm thickness) with capillary and pressurized mobile-phase flow were compared to conventional HPTLC (D7). A novel, automated device was described for collecting fractions from TLC plates for biological and spectrometric analysis (D8). On-site field TLC analysis of contaminated soils was described using a digital camera and notebook personal computer for cost-effective visualization (D9). Capillary ascending flow in a normal glass tank continues to be the main technique reported for TLC and HPTLC development, with use of other techniques such as forced-flow, multiple, and 2-D development to improve resolution of complex mixtures in certain analyses. The influences of various sample application and development techniques on the efficiency of TLC systems were 12R

Analytical Chemistry, Vol. 72, No. 12, June 15, 2000

compared (D10). Two simple, low-cost, and rapid techniques for circular TLC using a capillary feeder and pressurized flow system were described (D11). A new method for 2-D TLC made use of paper fixed on polymer film as the stationary phase (D12). Electroosmotic flow has been proposed as a transport mechanism for TLC (D13), but early results have not led to practical, highefficiency systems. TLC with forced flow of mobile phase in microchannels packed with a sorbent was proposed as a simple version of analytical OPLC and a micropilot technique for HPLC (D14). OPLC was combined with digital autoradiography (DAR) and off-line MS to detect and identify 3H- and 14C-labeled deramciclane metabolites in different biological matrixes with a sensitivity of ∼2 µg (D15). The amount of the cationic ion-pairing reagent cetyltrimethylammonium bromide adsorbed on silica gel and C-18 layers was related to separation performance by use of continuous OPLC (D16). Two commercially available AMD instruments were compared for alkaloid separations, and separate optimization of conditions was found to be required for each (D17). Chromatographic and spectrometric investigation of irreversible adsorption was made in the course of multiple development in conventional TLC, HPTLC, and OPLC (D18). Silica gel HPTLC-AMD proved to be a powerful method for separation of hydroalcoholic plant extracts containing complex mixtures of compounds of widely different polarity, and the method was proposed for fingerprinting herbal products (D19). A systematic procedure for optimization of parameters in AMD was demonstrated for the separation of six phenolic compounds (D20). TLC was combined with atomic absorption spectrometry (AAS) for the analysis of mixed-ligand zinc complexes (D21) and with square-wave anodic stripping voltammetry (SWASV) for determination of lead in lake water containing humic acid (D22). Two systems were described for on-line coupling of TLC with HPLC to provide complementary 2-D separations combining NP and RP mechanisms (D23, D24). TLC on sorbent-coated quartz rods with flame ionization detection (Iatroscan TLC-FID) is used most widely for the analysis of hydrocarbons and lipids. Examples of some recent applications include the following: characterization of petroluemcontaminated soils (D25); group-type analysis of petroleum heavy fractions (D26); quantitative characterization of different fossil fuel types using calibration based on internal normalization (D27); analysis of heptane insolubles and the paraffin content of bitumin (D28); analysis of mixtures of fatty acid esters (D29); simultaneous quantification of ceramides and 1,2-diacylglycerol in tissues (D30); and determination of free bile acids in pharmaceuticals (D31). Detection and Identification of Separated Zones. Detection in TLC is usually based on natural color, fluorescence, or ultraviolet (UV) absorption (fluorescence quenching on phosphorimpregnated layers) or on the use of various universal or selective chemical or biological detection reagents. An advantage of TLC lies in the ability to use a number of nondestructive detection methods and reagents in sequence for a single layer to increase the amount of information obtained. Identification is tentatively based on the correspondence of Rf values and detection characteristics between sample and standard zones but must be confirmed by other evidence, such as off- or on-line coupling of TLC with spectrometric methods.

The following new TLC chromogenic detection reagents were reported: amido black 10B and other water-soluble stains for lipids (E1); Coomassie brilliant blue for sphingolipids and sphingolipid synthesis inhibitors (E2); the π-acceptors p-chloranil and chloranilic acid for R-methyldopa (E3) and quinine (E4); a mixture of mercuric chloride and potassium ferricyanide for heroin (E5); iodide-azide for tri- and pentavalent organophosphorus compounds (E6); bromine-o-tolidine for phenytoin (E7); and water for surfactants (E8). Antimicrobial compounds were detected in plant extracts by direct bioautography (E9), and TLC and bioluminescence were coupled for detection of biologically active substances at low-nanogram levels (E10). Spraying with 20% triethanolamine in 2-propanol followed by 30% paraffin oil in hexane enhanced and stabilized the fluorescence of dansyl derivatives of pesticides and allowed detection at 1-6 pmol/zone (E11) Research on HPTLC-FT-IR coupling over the past 10 years was reviewed (E12). An optimized layer combining silica gel 60 plus 50% magnesium tungstate as a reflection enhancer was prepared for on-line TLC-diffuse reflectance FT-IR spectrometry (DRIFT) (E13). Interaction of acids and bases with the binder in precoated HPTLC plates greatly altered their spectra in on-line coupled FT-IR analyses (E14). Coupled HPTLC-DRIFT was used to identify impurities in chlordiazepoxide bulk drug powder and tablets (E15). Low-volatility hydrocarbon classes were separated by TLC-FID and detected by CO2-selective nondispersive IR spectrometry (E16). Combined TLC-FT-Raman spectrometric methods for separation and identification of organic compounds at low-microgram levels were described (E17). The effects of preparation conditions when the layer was treated with silver or gold were studied for identification of zones by surface-enhanced Raman spectrometry (SERS) (E18). The combination of TLC with X-ray fluorescence spectrometry (XRFS) for direct, nondestructive imaging of elements in inorganic compounds and organic compounds was achieved by wrapping the dried silica gel or cellulose plate in polyethylene film prior to placing it onto the stage of the spectrometer (E19). Methods and applications of direct on-plate SWASV for detection of metals on layers at low-nanogram levels were reviewed (E20). Publications on the combination of TLC with MS were as follow: a critical appraisal of the state of the art in TLC/MS, including matrix-assisted laser desorption/ionization (MALDI), surface-assisted laser desorption (SALDI), and TLC/ES-MS (E21); a review of practical applications of TLC/MS-MS (E22); interface designs for TLC/ES-MS (E23); TLC combined with off-line electron impact MS (EIMS) for screening of biological samples for drugs and metabolites (E24); activated carbon SALDI TLC/ MS using a nitrogen laser to desorb analyte ions from the gel surface in a time-of-flight (TOF) mass spectrometer (E25); detection and quantification of organic compounds at picogram levels on silica gel and RP plates by TLC/MALDI-MS (E26); TLC/ MALDI-TOF-MS for pharmaceutical impurity testing (E27); identification of organic reaction products by TLC/MALDI-TOFMS (E28) and SALDI-TOF-MS (E29); confirmation of residues of thyreostatic drugs in thyroid glands by multiple MS after TLC screening (E30); detection and identification of morphine in urine extracts using TLC and tandem MS (E31); and analysis of cationic pesticides by TLC/MALDI-MS (E32).

Quantitative Analysis. Although the majority of planar chromatography analyses are carried out on a qualitative or semiquantitative (visual comparison) basis, modern computer-controlled slit-scanning densitometers that mechanically scan sample and standard chromatograms in tracks on the layer allow selective, sensitive, accurate, and precise quantitative analyses to be carried out by HPTLC. Reports of the use of electronic scanning or image analysis with a video densitometer (CCD camera) based on total illumination of the plate with the light source again increased during the review period, but sensitivity, resolution, selectivity, and applicability to absorption and fluorescence measurements do not match slit-scanning instruments at this time. Validation of results obtained by TLC is increasingly addressed in the literature because of the quality demands imposed on analyses, such as those performed under pharmaceutical compendial, regulatory, and GLP standards. TLC is considered by some analysts to be the most reliable and informative separation method because samples, standards, and control samples are developed at the same time on a single layer with the same mobile phase; samples are not eluted as in column procedures; and the chromatograms that are produced simultaneously but independently are stored in separate tracks on the layer and can be evaluated and validated in a variety of ways after the chromatographic procedure. Setup of a simple, low-cost image analysis system with a software package for qualitative and quantitative analysis was described (F1). The influence of measurement parameters for a dual-wavelength, flying-spot UV densitometer and a CCD video camera was studied for detection and documentation of plant phenolic separations at 254 and 366 nm, and the two instruments were found to be equivalent for this application (F2). Quantification by a densitometer versus a video camera was discussed and reviewed; the densitometer was found to be generally superior but the video camera useful in some applications depending on the necessary quality of the measurements and the cost efficiency (F3). For quantification of impurities and synthetic byproducts in theophylline, comparison of a slit scanner with video densitometry found the limit of detection and precision to be similar but that superior sensitivity and precision was achieved by slit scanning (F4). Validation of slit scanning and video densitometric determinations of pesticides showed that the former was more sensitive and precise (although the precision of video densitometry, 3.5-5.3%, was acceptable according to validation requirements) and that linearities were very good and almost identical for both techniques; the main advantages of video technology were speed (several seconds vs 20 min) and excellent archiving facility (F5). A quantitative OPLC purity test for phthaloylamplodipine was developed and validated in terms of specificity (retention factor, resolution, and asymmetry factors), detection and quantification limits, and precision (F6). Complete validation of an HPTLC method for caffeine in Coca Cola was performed according to International Conference on Harmonization (ICH) guidelines for slit scanning and video densitometry; the validation parameters were selectivity, stability of the analyte, linearity, precision (repeatability and intermediate precision), and robustness (F7). Preparative Layer Chromatography and Radio-Thin-Layer Chromatography. Although commercial instruments are available for forced-flow preparative layer chromatography (PLC), most notably rotation planar chromatography (RPC), the majority of Analytical Chemistry, Vol. 72, No. 12, June 15, 2000

13R

applications have involved classical ascending-development TLC techniques with thicker layers. PLC is used widely to isolate and recover larger amounts (e.g., 10-100 mg) of compounds for analysis by other methods, such as spectrometry, for example in pesticide metabolism studies. Such applications will not be covered in this review. The use of multiple development and incremental multiple development coupled with stepwise gradient elution was illustrated for improved micropreparative separations of closely related compounds (G1). The principal methods for detecting and quantifying radioactive zones in one-dimensional (1-D) and 2-D chromatograms are autoradiography, zonal analysis (scraping followed by scintillation counting), and direct measurement using radioimaging detectors. Techniques and applications of radio-TLC (also termed thin-layer radiochromatography) were reviewed (G2). The use of radio-TLC was reported in the following studies: investigation of methionine metabolism in human peripheral blood mononuclear cells using scintillation counting (G3); direct quantitative DAR coupled with TLC for analysis of neutral 14C-lipids neosynthesized by the human sebacious gland (G4); the use of TLC-DAR for studying drug metabolism (G5); rapid assay of nitric oxide synthase using TLC with radiometric scanning or scintillation counting (G6); phosphopeptide mapping by 2-D electrophoresis-TLC on a cellulose plate followed by film autoradiography or phosphor-imaging screen detection of 32P-containing peptide zones (G7); measuring the hepatic first-pass effect and metabolic rate of L-3,4-dihydroxyphenylalanine (DOPA), diazepam, and inulin in rat liver using TLC-autoradioluminography (G8); combination of HPLC and TLC for separation of five 32P-labeled adducted nucleotides isolated from rat liver (G9); detection and quantification of phosphatidylinositol-4-phosphate 5-kinase activity using TLC with autoradiography and scintillation counting (G10); OPLC-DAR for fast separation and detection of metabolites in biological samples (G11); study of bile acid kinetics in man by radio-TLC coupled with densitometry (G12); the use of the InstantImager electronic autoradiography system for direct nuclear quantification of TLC and gel mobility shift assays (G13); and analysis of human CYP3Acatalyzed testosterone 6-β-hydroxylation (G14). APPLICATIONS Unless otherwise stated, the references below involved TLC on precoated plates or sheets. TLC plates are rarely hand-coated today, unless a special layer is required that is not available commercially or the cost of precoated plates is prohibitive compared to in-house preparation. Acids and Phenols. Carboxylic acids separated on silica gel or RP plates were derivatized by reaction with 4-(bromomethyl)7-methoxycoumarin or panacyl bromide in the presence of a crown ether catalyst prior to detection in the low-picomole range by fluorodensitometry at 360 nm (H1). The main hydroxamic acids (DIMOBA and DIBOA) in wheat and rye extracts were quantified by TLC-densitometry in up to five samples simultaneously (H2). Phenols were analyzed on diol-bonded silica with mobile phases composed of heptane plus 1% acetic acid and ethyl acetate, 2-propanol, dioxane, or tetrahydrofuran (THF) as the polar modifier (H3) and on calcium sulfate layers with six organic solvents (H4). Ten new spray reagents, including alkacymetric indicators, were reported for detecting phenols at levels as low 14R

Analytical Chemistry, Vol. 72, No. 12, June 15, 2000

as 20 ng on silica gel, silica gel-Kieselguhr, and polyamide layers (H5). Free and bound phenolic acids from Ginko biloba L. leaves were identified by 2-D TLC in a horizontal DS chamber and quantified by HPLC (H6). The best solvent systems for silica gel phenol separations among six tested were chloroform-ethyl acetate-acetic acid (50:50:1) and benzene-dioxane-acetic acid (85:15:1) (H7). Complex mixtures of phenolic acids from Lycopus europaeus were separated using multiple gradient development on silica, propylamine, and diol layers and isocratic development with a C-18 layer (H8). Butylated hydroxytoluene in gum base was quantified at a level of 25 ng by densitometric scanning of silica gel plates after 2-fold development with hexane followed by petroleum ether (H9). Amino Acids, Peptides, and Proteins. The effects of structure on the chromatographic retention of selected protein amino acids in the form of thiohydantoin derivatives and their separation from free amino acids were studied (I1). Selected amino acid enantiomers were separated on a chiral layer, and a new valence optical topological index and valence optical Gutman index were proposed to enable distinction between isomers of L and D configuration (I2). TLC separation of 19 amino acids was studied on silica gel layers impregnated with different anions and developed with two new solvent systems, 1-butanol-methanolacetic acid (8:1:3) and 1-butanol-carbon tetrachloride-acetic acid (8:3:1); amino acids were kept in cationic form below their pI values, and the advantage of ion pair formation was taken into account (I3). The automated personal OPLC instrument was used for the separation of the main protein amino acids using a special double-layer cassette that provides the possibility of a long development distance and increased spot capacity and resolution (I4). Enantiomers of unusual aromatic amino acids were analyzed on Macherey-Nagel Chiralplates with acetonitrile-methanolwater (4:1:1 or 4:1:2) or acetonitrile-methanol-water-diisopropylethylamine (4:1:2:0.1) as mobile phase and ninhydrin detection reagent (I5). Amino acids were extracted from medicinal plants with 1% aqueous HCl and separated by 2-D development on cellulose plates (I6). Amino acids were identified in water conditioned by four strains of snails using cellulose HPTLC plates developed with 1-propanol-water (7:3) and detection with ninhydrin spray reagent; valine content was quantified by densitometric scanning at 610 nm (I7). TLC and HPLC methods for the separation of enantiomers and epimers of β-alkyl amino acids and peptides containing them were reviewed (I8). Rapid screening of synthetic peptides was carried out by use of the personal OPLC system with silica gel layers and 1-butanol-pyridine-acetic acid-water (12:4:1:4) mobile phase; it was shown that OPLC can be used orthogonally with HPLC or capillary electrophoresis (CE) for multimodal separations of closely related peptides (I9). Protein separations by TLC have been limited in the past to size exclusion chromatography on homemade, swollen gel layers, but the following two studies report adsorption TLC separations. A process was described for separating and analyzing hydrophobic proteins using TLC on a modified silica matrix and immunostaining for detection (I10). Thin-layer ion-exchange chromatography was used to separate four model proteins using silica gel layers,

0.01 M bicine, pH 8.5, and a three-step elution process with 0.01, 0.025, and 0.10 M NaCl (I11). Antibiotics. The retention behavior of five cephalosporins was investigated by gradient elution RP-ion pair TLC with variation of pH, type and concentration of ion pairing counterion, and concentration of the organic solvent in the aqueous mobile phase (J1). Eight cephalosporins were separated on silica gel and silanized C-18 layers by horizontal development using selected mobile phases; detection was based on UV absorption and use of different spray reagents (J2). HPLC with a C-18 column, precolumn derivatization with phthalaldehyde, and UV detection was compared to HPTLC using a silica gel GF layer, chloroformn-methanol-ammonia (1:1:1) mobile phase, detection with ninhydrin reagent, and densitometric scanning at 580 nm for the quantification of gentamicin in bulk drugs, and TLC was judged to be simpler, more accurate and economical, and faster (J3). A large number of antibiotics, including ceftriaxone and 12 cephalosporins, were quantified successfully by TLC-densitometry on hydrocarbon-impregnated silica gel HPTLC plates (J4-J6). Ofloxacin and enoxacin were determined simultaneously in body fluids at low-nanogram levels with 95-105% recovery by TLC with fluorescence densitometry on ethylenediaminetetraacetic acid (EDTA)-impregnated silica gel plates developed with chloroform-methanol-ethyl acetateTHF-aqueous ammonia (6:4.6:1.5:0.8:1.5) (J7). A stability-indicating method was reported for determining intact ceftazidime, cefuroxime sodium, and cefotaxime sodium in the presence of their degradation products in pharmaceutical formulations by densitometric scanning of fluorescence quenched zones on silica gel GF plates in the range of 2-16 µg with >99% accuracy (J8). Bases and Amines. Methodology for separation, detection, and quantitative determination of catecholamines, 5-hydroxytryptamine, and their acidic metabolites in biological tissue and fluids by TLC were reviewed, and selected procedures for analysis by fluorescence and visible absorption densitometry were described (K1). Procedures were outlined for separation and analysis of polyamines by TLC (K2). Effective separations of aromatic amines, including isomers, were accomplished on alumina thin layers developed with sodium dodecyl sulfate (SDS)-waterheptane-1-pentanol microemulsion (K3). Layers composed of zirconium molybdophosphate mixed with silica gel G with aqueous mobile phases containing sodium nitrate and HCl were used to separate 16 primary aromatic amine hydrochlorides (K4). Separations of multicomponent mixtures of aromatic amines that are reduction products of azo dyes were studied on silica gel (adsorption TLC) and bonded amino, cyano, and diol (normalphase partition) plates with 2-propanol-n-hexane mobile phases, and on C-2, C-8, and C-18 (reversed-phase partition) plates with methanol-water mobile phases (K5). (-)-Deprenyl and some structurally related amine analogues were characterized by elution and displacement TLC on silica gel and C-18 layers, and lipophilicity parameters were determined on paraffin-coated silica gel plates (K6). A simple and rapid method was developed and validated for determining biogenic amines in foods using 5% trichloroacetic acid extraction, cleanup by washing with ethyl ether, preparation of dansylated amines, multiple development silica gel TLC, and densitometry of the separated compounds at 254 nm (K7).

Carbohydrates. The retention behavior of some neutral sugars on silica gel layers impregnated with metal sulfates, chlorides, and nitrates was studied in terms of the effects of pH and of the concentration of impregnating solution, and attempts were made to explain the results from the standpoint of specific adsorption theory (L1). Applications of gradient elution AMD-TLC and OPLC were described for analysis of complex samples such as beet or cane molasses in the sugar industry; advantages cited include a low degree of sample cleanup, large spot capacities, and short development times (L2). Neutral sugar and uronic acid products of gum hydrolysis were simultaneously separated and determined by TLC on silica gel G impregnated with phosphate buffer solutions and developed with mobile phases containing polar aprotic solvents, and the results were used to identify the gums (L3). Seven aldoses and their corresponding alditols were separated on silica gel by two ascending developments with acetonitrile-ethyl acetate-1-propanol-water (85:20:20:25) and detected at 0.5-1 µg levels by dipping into alkaline silver nitratesodium thiosulfate reagent (L4). Systems devised for rapid (1520 min) analysis of 15 monosaccharides and disaccharides involved silica gel 60 plates and acetone-based mobile phases; detection was carried out with methanolic aniline-diphenylamine-phosphoric acid reagent (L5). Glucose in human blood was quantified using silica gel preadsorbent plates, acetonitrilewater (85:15) mobile phase, methanolic sulfuric acid detection reagent, and scanning densitometry; results were validated and compared with those given by a commercial home-monitoring kit intended for use by diabetics (L6). Maltose and glucose were identified as the primary sugars in the planorbid snail Biomphalaria glabrata by migration behavior on four layer types with different selectivities, use of specific detection reagents, and off-line GC/ MS of separated zones; the effect of starvation on the amounts of these carbohydrates in the snail digestive gland-gonad complex (DGG) was determined by scanning densitometry on silica gel after detection with 1-naphthol reagent (L7). Dyes and Pigments. Ten fluorescein dyes were separated by TLC with aqueous micellar solutions as mobile phases, and effects of mobile-phase acidity and the nature and concentration of the surfactant and dye were investigated; the cationic and anionic surfactants used resulted in conversion of the silica gel layer from a normal phase to a reversed phase (M1). TLC discriminated among 98 light and dark red automotive paint pigments in forensic applications (M2). Twenty red, blue, and green printing inks were analyzed by TLC using two different solvent systems for development; all brands could be differentiated on the basis of color, number, and Rf values of the separated components (M3). Two preliminary official TLC methods for identification of betanin, cochineal, red sandal wood, and angkak (red rice) pigments underwent a satisfactory collaborative study for detection of the four natural colorants in raw sausages (M4). Qualitative and quantitative determination of synthetic food colors in wine, wine-containing and nonalcoholic beverages, and spirits were performed with detection limits of