Encyclopedia of chromatography by jack cazes 2

Organic Acids, Analysis by Thin Layer ChromatographyNatasˇa Brajenovic´ Rud¯er Bosˇkovic´ Institute, Zagreb, Croatia INTRODUCTION A very large number of papers report new research works

Optical activity detectors are capable of specifically

detecting chiral compounds, taking advantage of their

unique interactions with polarized light Much of the

work on the development of prisms and other devices

for the production of polarized light was done in the

early part of the nineteenth century However, the

measurement of optical activity is often used for

enan-tiomeric purity determination of chiral compounds,

which by definition have either a center or plane of

asymmetry Enantiomers rotate the plane of polarized

light in opposite directions, although in equal amounts

The isomer that rotates the plane to the left

(counter-clockwise) is called the levo isomer and is designated

(), whereas the one that rotates the plane to the right

(clockwise) is called the dextro isomer and is

desig-nated () Questions of optical activity are of extreme

importance in the field of asymmetric chemical

synthe-sis and in the pharmaceutical industry

Detection Principle

Figure 1 shows the basic optimal system of the optical

rotation detector, which is based on the nonmodulated

polarized beam-splitting method The light radiated

from the light source is straightened by the plane

po-larizer, then to the lens for beam formation and

con-centration, and then to the flow cell

The plane-polarized light which goes through the

flow cell is rotated by optically active substances

(chi-ral compounds) according to their specific optical

rota-tions and concentrarota-tions The light then enters the

po-larized beam splitter and is divided into two beams

according to the polarized beam directions These

beams are detected by two photodiodes as shown

The angle of the plane polarizer is adjusted so that

the two photodiodes may receive the same beam

in-tensity when no optically active substance is present in

the flow cell When optically active substances are

present in the flow cell, the difference between the

beam intensities received by the two photodiodes is

not zero Therefore, the difference has a linear relation

Copyright © 2002 by Marcel Dekker, Inc All rights reserved

with specific optical rotation and concentration of theoptically active substance and can be expressed by

where is the difference of beam intensities received

by the two photodiodes (i.e., output of signal level), K

is a constant determined by cell structure and light tensity of the light source, [a] is the specific optical ro-

in-tation of the chiral compound, and C is the

concentra-tion of the chiral compound

Polarimetry Theory

Most forms of optical spectroscopy are usually cerned with the measurement of the absorption oremission of electromagnetic radiation Ordinary,natural, unreflected light behaves as though it con-sists of a large number of electromagnetic waves vi-brating in all possible orientations around the direc-tion of propagation If, by some means, we sort outfrom the natural conglomeration only those rays vi-brating in one particular plane, we say that we haveplane-polarized light Of course, because a lightwave consists of an electric and a magnetic compo-nent vibrating at right angles to each other, the term

con-“plane” may not be quite descriptive, but the ray can

be considered planar if we restrict ourselves to ing the direction of the electrical component Circu-

not-V0

V0 K 3a4C

Fig 1 Optical rotation detector.

measurable quantities, which are obtained in arbitraryunits and related to the circular polarization condition

of the luminescence It is appropriate to considerCPLS spectroscopy as a technique that combines theselectivity of CD with the sensitivity of luminescence.The major limitation associated with CPLS spectros-copy is that it is confined to emissive molecules only

Vibrational Optical Activity

The optical activity of vibrational transitions has beenconducted The infrared (IR) bands of a small mole-cule can easily be assigned with the performance of anormal coordinate analysis, and these can usually bewell resolved One of the problems associated with vi-brational optical activity is the weakness of the effect.Instrumental limitations of infrared sources and detec-tors create additional experimental constraints on thesignal-to-noise ratios

Two methods suitable for the study of vibrationaloptical activity have been developed:

Vibrational Circular Dichroism: Vibrational lar dichroism (VCD) could be measured at good signal-to-noise levels Vibrational optical activity

circu-is observed in the classic method of Grosjean andLegrand

Raman Optical Activity: The Raman optical ity (ROA) effect is the differential scattering ofleft- or right-circularly polarized light by a chiralsubstrate where chirality is studied through Ra-man spectroscopy

activ-Fluorescence-Detected Circular Dichroism

Fluorescence-detected circular dichroism (FDCD) is achiroptical technique in which the spectrum is ob-tained by measuring the difference in total lumines-cence obtained after the sample is excited by left- andright-circularly polarized light For the FDCD spec-trum of a given molecular species to match its CD spec-trum, the luminescence excitation spectrum must beidentical to the absorption spectrum

Factors Affecting the Measurement

of Optical Rotation

The rotation exhibited by an optically active substancedepends on the thickness of the layer traversed by thelight, the wavelength of the light used for the measure-ment, and the temperature of the system In addition,

if the substance being measured is a solution, then the

lar polarized light represents a wave in which the

electrical component (and, therefore, the magnetic

component also) spirals around the direction of

propagation of the ray, either clockwise

(“right-handed” or dextrorotatory) or counterclockwise

(“left-handed” or levorotatory) If, following the

passage of the plane-polarized ray through some

material, one of the circularly polarized

compo-nents, say the left circularly polarized ray, has been

slowed down, then the resultant would be a

plane-polarized ray rotated somewhat to the right from its

original position In addition, lasers have been

in-corporated into two optical rotation methods to

date: polarimetry and circular dichroism

Optical Rotation and Optical Rotatory Dispersion

A polarimeter measures the direction of rotation of

plane-polarized light caused by an optically active

sub-stance The specific optical activity of an asymmetrical

molecule varies with the wavelength of the light used

for its determination This variation is called optical

rotatory dispersion (ORD) In ORD, rotations are

measured over a range of wavelengths rather than at a

single wavelength, usually covering the ultraviolet

(UV) as well as the visible region

Circular Dichroism

In this technique, the molecular extinction coefficients

of a compound are measured with both left and right

circularly polarized light, and the difference between

these values is plotted against the wavelength of the

light used The phase angle between the projections of

the two circularly polarized components is altered by

passage through the chiral medium, but their

ampli-tudes will be modified by the degree of absorption

ex-perienced by each component This differential

ab-sorption of left- and right-circularly polarized light is

termed circular dichroism (CD) So, circular dichroism

measurements provide both absorbance and optical

rotation information simultaneously

Circularly Polarized Luminescence Spectroscopy

Circularly polarized luminescence spectroscopy

(CPLS) is a measure of the chirality of a luminescent

excited state The excitation source can be either a

laser or an arc lamp, but it is important that the source

of excitation be unpolarized to avoid possible

photos-election artifacts The CPLS experiment produces two

concentration of the optically active material is also

in-volved and the nature of the solvent may also be

im-portant There are certain substances that change their

rotation with time Some are substances that change

from one structure to another with a different rotatory

power and are said to show mutarotation

Mutarota-tion is common among the sugars Other substances,

owing to enolization within the molecules, may rotate

so as to become symmetrical and, thus, lose their

rota-tory power These substances are said to show

racem-ization Mutarotation and racemization are influenced

not only by time, but also by pH, temperature, and

other factors Of course, rotations that determined for

the same compound under the same conditions are

identical Therefore, in expressing the results of any

polarimetric measurement, it is, therefore, very

impor-tant to include all experimental conditions

Temperature

Temperature changes have several effects on the

rota-tion of a solurota-tion or liquid An increase in temperature

increases the length of the tube; it also decreases the

density, thus reducing the number of molecules

in-volved in the measurement It causes changes in the

ro-tatory power of the molecules themselves, due to

asso-ciation or dissoasso-ciation and increased mobility of the

atoms, and affects other properties In addition,

tem-perature changes cause expansion and contraction of

the liquid and a consequent change in the number of

active molecules in the path of the light

The unique ability of the optical rotation detector

to respond to the sign of rotation allows precise

enantiomeric purity determination even if the

enan-tiomers are only partially resolved The sign of

rota-tion is also useful in establishing enantiomer elurota-tion

order

Because the optical rotation detectors only respond

to optically active compounds, enantiomeric purity

de-termination to precisions of better than 0.5% can be

achieved and is possible in even the complex mixtures

The detection can also be used as part of a flow

injec-tion analysis system to determine amount and

enan-tiomeric purity of a drug in dosage form

The applications using optical rotation detectors

in-clude the following:

1 Qualitative analysis of chiral compounds, cluding drugs, pesticides, carbohydrates, aminoacids, liquid crystals, and other biochemicals

in-2 Determination of enantiomeric purity of chiralcompounds

3 Monitoring an enzymatic reaction

4 Qualitative analysis of proteins

5 Use as a conventional polarimeterHowever, the disadvantages of optical rotation detec-tors may be limited by shot or flicker noise, which are de-pendent on the optical and mechanical properties of thesystem or by noise in the detector electronics Generally,the usefulness of this technique has been limited by thelack of sensitivity of commercially available instruments

Suggested Further Reading

Allenmark, S., Techniques used for studies of optically

active compounds, in Chromatographic aration: Methods and Application, 2nd ed., Ellis

Enantiosep-Horwood Ltd., London, 1991

Beesley, T E and R P W Scott, An introduction to

chiral chromatography, in Chiral Chromatography,

John Wiley & Sons, Inc., New York, 1998, pp 1–11.Dodziuk, H., Physical methods as a source of informa-tion on the spatial structure of organic molecules, in

Modern Conformational Analysis, Elucidating Novel Exciting Molecular Structures, VCH, New

York, 1995, pp 48 –54

Edkins, T J and D C Shelly, Measurement conceptsand laser-based detection in high-performance mi-

cro separation, in HPLC Detection: Newer Methods

(G Patonay, ed.), VCH, New York, 1992, pp 1–15.Goodall, D M and D K Lloyd, A note on an opticalrotation detector for high-performance liquid chro-

matography, in Chiral Separations (D Stevenson

and D Wilson, eds.), Plenum Press, New York,

1988, pp 131–133

Sheldon, R A., Introduction to optical isomersion, in

Chirotechnology: Industrial Synthesis of Optically Active Compounds, Marcel Dekker, Inc., New

York, 1993, pp 25 –27

Weston, A and P R Brown, HPLC and CE Principles and Practice, Academic Press, San Diego, CA, 1997 Yeung, E S., Polarimetric detectors, in Detectors for Liquid Chromatography (E S Yeung, ed.), John

Wiley & Sons, New York, 1986, pp 204 –228

Quantitative evaluation of thin-layer chromatograms

can be performed by direct, in situ visual, and indirect

elution techniques Visual evaluation involves

compar-ison of the sizes and intensities of color or fluorescence

between sample and standard zones spotted,

devel-oped, and detected on the same layer The series of

standards is chosen to have concentrations or weights

that bracket those of the sample zones After matching

a sample with its closest standard, accuracy and

preci-sion are improved by respotting a more restricted

se-ries of bracketing standards with a separate sample

spot between each of two standard zones Accuracy no

greater than 5 –10% is possible for trained personnel

using visual evaluation The determination of

myco-toxins in food samples is an example of a practical

ap-plication of visual comparison of fluorescent zones

The elution method involves scraping off the

sepa-rated zones of samples and standards and elution of

the substances from the layer material with a strong,

volatile solvent The eluates are concentrated and

an-alyzed by use of a sensitive spectrometric method, gas

or liquid column chromatography, or electroanalysis

Scraping and elution must be performed manually

be-cause the only commercial automatic

microprepara-tive elution instrument has been discontinued by its

manufacturer The elution method is tedious and

time-consuming and prone to errors caused by the incorrect

choice of the sizes of the areas to scrape, incomplete

collection of sorbent, and incomplete or inconsistent

elution recovery of the analyte from the sorbent

How-ever, the elution method is being rather widely used

(e.g., some assay methods for pharmaceuticals and

drugs in the USP Pharmacopoeia)

Introduction to Densitometry

In order to achieve the optimum accuracy, precision,

and sensitivity, most quantitative analyses are

per-formed by using high-performance thin-layer

chroma-tography (TLC) plates and direct quantification by

means of a modern optical densitometic scanner with

Copyright © 2002 by Marcel Dekker, Inc All rights reserved

a fixed sample light beam in the form of a rectangularslit that is variable in height (e.g., 0.4 –10 mm) andwidth (20 mm to 2 mm) Densitometers measure thedifference in absorbance or fluorescence signal be-tween a TLC zone and the empty plate backgroundand relate the measured signals from a series of stan-dards to those of unknown samples through a calibra-tion curve Modern computer-controlled densitome-ters can produce linear or polynomial calibrationcurves relating absorbance or fluorescence versusweight or concentration of the standards and deter-mine bracketed unknowns by automatic interpolationfrom the curve Samples and standards are best appliedusing an automated instrument such as the one shown

in Fig 1 Use of manual spotting and less efficient TLCplates results in greater errors and poorer repro-ducibility in quantitative results

Instrumental Design and Scanning Modes

A commercial densitometer and a schematic diagram

of the light-path arrangement used in scanning areshown in Fig 2 The plate is mounted on a moveablestage controlled by a stepping motor drive that allowseach chromatogram track to be scanned in or againstthe direction of development A tungsten or halogenlamp is used as the source for scanning colored zones

in the 400 – 800-nm range (visible absorption) and adeuterium lamp for scanning ultraviolet (UV)-absorb-ing zones directly or as quenched zones on phosphor-containing layers (F-layers) in the 190 – 450-nm range.The monochromator used with these continuous-wavelength sources can be a quartz prism or, more of-ten, a grating The detector is a photomultiplier orphotodiode placed above the layer to measurereflected radiation [Some scanners (e.g., Fig 2) makeuse of a reference photomultiplier in addition to themeasuring photomultiplier in the single-beam mode;the reference photomultiplier puts out a constant sig-nal that is compared to the signal from the measuringphotomultiplier to produce a difference signal that ismore accurate than a direct signal from a single mea-suring photomultiplier would be.]

tor, were designed earlier with two photomultiplier tectors simultaneously recording the two beams (doublebeam in space), but, today, such densiotmeters areequipped with a chopper and one detector (doublebeam in time) For dual-wavelength, single-beam scan-ning, which will correct for scattering of the absorbedlight by subtracting out the (presumably equal) scatter-ing at a nonabsorbed wavelength, a light beam is se-lected by a mirror and passes through two separatemonochromators to isolate the two different wave-lengths The two beams are alternated by a chopper andrecombined into a single beam representing a differencesignal at the detector Zigzag or meandering scanningwith a small point or spot of light is possible with densit-ometers having two independent stepping motors to

de-move the plate in the X and Y axes Computer

algo-rithms integrate the maximum absorbance ments from each swing to produce a distribution profile

measure-of zones having any shape The potential advantages measure-ofscanning with a moving light spot are offset by problemswith lower spatial resolution and errors in data process-ing, and the method is not as widely used as conven-tional scanning of chromatographic tracks with a fixedslit Some densitometers have the ability to rotate theplate while scanning for measurement of circular andanticircular chromatograms

Single-wavelength, single-beam, fixed-slit scanning

is most often used and can produce excellent results

For normal fluorescence scanning, a high-intensity

xenon continuum source or a mercury vapor line

source is used, and a cutoff filter is placed between the

plate and detector to block the exciting UV radiation

and transmit the visible emitted fluorescence For

fluorescence measurement in the reversed-beam

mode, a monochromatic filter is placed between the

source and plate and the monochromator between the

plate and detector In this mode, the monochromator

selects the emission wavelength, rather than the

excita-tion wavelength as in the normal mode

Simultaneous measurement of reflection and

trans-mission, or transmission alone, can be carried out by

means of a detector positioned on the opposite side of

the plate (Fig 2) Ratio-recording double-beam

densit-ometers, which can correct for background disturbances

and drift caused by fluctuations in the source and

detec-Fig 1 Automatic TLC sampler (ATS 3) used for

computer-controlled application of precisely computer-controlled volumes of

samples and standards between 10 nL and 50 mL from a rack

of vials as spots or bands to preselected origins on a plate.

(Courtesy of Camag Scientific Inc., Wilmington, NC.)

Fig 2 Photograph of the DESAGA Densitometer CD 60 with a superimposed schematic diagram of the light path including (right to left) the source lamp, two mirrors, grating monochromator, mirror, beam splitter, plate with chromato- grams to be scanned, reference and measuring detectors (reflection) above the plate and detector (transmission) be- low the plate (Courtesy of DESAGA GmbH, Weisloch, Germany.)

when high-quality plates and analytical techniques are

employed

Spectral Measurement

Many modern scanners have a computer-controlled

mo-tor-driven monochromator that allows automatic

recording of in situ absorption and fluorescence

excita-tion spectra These spectra can aid compound

identification by comparison with stored standard

spec-tra, test for identity by superimposition of spectra from

different zones on a plate, and check zone purity by

su-perimposition of spectra from different areas of a single

zone The spectral maximum determined from the in situ

spectrum is usually the optimal wavelength for scanning

standard and sample areas for quantitative analysis

Data Handling

The densitometer is connected to a recorder, integrator,

or computer A personal computer with software

de-signed specifically for TLC is most common for data

han-dling and automation of the scanning process in modern

instruments With a fully automated system, the

com-puter can carry out the following functions: data

acquisi-tion by scanning a complete plate following a preselected

geometric pattern with control of all scanning

parame-ters; automated peak searching and optimization of

scanning for each fraction located; multiple-wavelength

scanning to find, if possible, a common wavelength for all

substances to be quantified, to optically resolve fractions

incompletely separated by TLC, and to identify fractions

by comparison of spectra with standards

cochromato-graphed on the same plate or stored in a spectrum

li-brary through pattern recognition techniques; baseline

location and correction; computation of peak areas

and/or heights of samples and codeveloped standards

and processing of the analog raw data to quantitative

dig-ital results, including calculation of calibration curves by

linear or polynomial regression, interpolation of sample

concentrations, statistical analysis of reproducibility, and

presentation of a complete analysis report; and storage

of raw data on disk for later reintegration, calibration,

and evaluation with different parameters

Calibration Curves

Densitometric calibration curves relating absorption

signal and concentration or weight of standards on

the layer are usually nonlinear, especially for higher

amounts of standards, and do not pass through theorigin Fluorescence calibration curves are generallylinear and pass through the origin, and analysesbased on fluorescence are more specific and 10 –1000times more sensitive The advantages of fluorescencemeasurement may be realized for nonfluorescentcompounds by prechromatographic or postchro-matographic derivatization reactions with suitablefluorogenic reagents

Because the incident monochromatic light is sorbed, reflected, and scattered by the opaque layermaterial, the theoretical relationship between amount

ab-of absorption and amount ab-of substance does not followthe simple Beer –Lambert law that is valid for solu-tions The Kubelka –Munk equation is the most ac-cepted theoretical relationship for TLC, but its use isnot necessary because of the ability of densitometersoftware to handle empirical nonlinear regressionfunctions

Image Analysis (Videodensitometry)

Video camera systems are available from severalmanufacturers for documentation and densitomet-ric quantification of TLC plates As an example, theCamag VideoScan instrument consists of a lightingmodule with short- and long-wave UV and visiblesources upon which the layer is placed, a charge-coupled device (CCD) camera with zoom and long-time integration capability, and a PC under MS-Windows control with frame grabber, monitor, andprinter The available software for quantitative eval-uation allows the display of the tracks of the chro-matogram image acquired with the video camera asanalog curves and calculation of their peak proper-ties ( height, area, height percent, and area per-cent) For quantification, the computer creates astandard curve from the areas or heights of the stan-dards and interpolates unknown values from thecurve

Video scanners have potential advantages, ing rapid data collection, simple design with virtually

includ-no moving parts, and ability to quantify sional chromatograms, but they have not yet beenshown to have the required capabilities, such assufficient spectral discrimination or the ability to il-luminate the plate uniformly with monochromaticlight of selected wavelength, to replace slit-scanningdensitometers Current video scanners can measurespots in the visible range in transmittance, reflec-tance, or fluorescence modes, but they cannot per-form spectral analysis

two-dimen-R f,

6 Apply a detection reagent, if necessary, byspraying or dipping The reagent should pro-duce a stable colored, UV-absorbing, or fluores-cent zone having high contrast with the layerbackground.

7 Scan the natural or induced absorption or rescence of the standard and sample zones onthe plate using a densitometer with optimizedparameters

fluo-8 Generate a calibration curve by linear or nomial regression of the scan areas andweights of the standards and interpolate theweights in the sample zones from the curve

poly-9 Calculate the concentration of analyte in thesample from the original weight of the sample,the original total volume of the sample test so-lution, the aliquot volume of the test solutionthat is spotted, the interpolated analyte weight

in that spotted volume from the calibrationcurve, and any numerical factor required be-cause of dilution or concentration stepsneeded for the test solution to produce abracketed scan area for the analyte zone in thesample chromatogram

10 Validate the precision of the TLC analysis byreplicated determination of the sample and ac-curacy by comparison of the results to thoseobtained from analysis of the same sample by

an established independent method or tion of recovery from analysis of a spiked pre-analyzed sample or spiked blank sample.The following are some advantages of TLC densito-metry compared to HPLC:

calcula-1 The simultaneous analysis of multiple samples

on a single plate leads to higher samplethroughput (lower analysis time) and less costper sample Up to 36 tracks are available forsamples and standards on a 10-cm  20-cmhigh-performance TLC plate

2 The ability to generate a unique calibrationcurve using standards developed under thesame conditions as samples on each plate (in-system calibration) leads to statistical improve-ment in data handling and better analytical pre-cision and accuracy and eliminates the need for

an internal standard for most analyses

3 Detection is versatile and flexible because themobile phase is removed prior to detection Be-cause the detection process is static (the zonesare stored on the layer), multiple, complemen-tary detection methods can be used

Applications and Practical Aspects

of Densitometry

Densitometric quantification has been applied to

virtu-ally every type of analyte and sample For example, the

greatest number of applications is for the analysis of

drug and pharmaceutical compounds, most of which

have structures including chromophores that cause

strong UV absorption These compounds are readily

quantified in the fluorescence quenching mode on

F-layers or in the direct UV absorption mode on

unim-pregnated layers Lipids are compounds that are not

easily analyzed by gas chromatography (GC) or

high-performance liquid chromatography (HPLC) because

they lack volatility and the presence of a chromophore

leading to UV absorption The most successful way to

quantify lipids is by densitometry after separation and

detection on the layer with a chromogenic reagent, most

notably phosphomolybdic acid The quantification of

amino acids after detection with ninhydrin is another

example of densitometry in the visible absorption

mode Fluorescence densitometry has been applied to

the determination of naturally fluorescent compounds

(e.g., quinine in tonic water) or compounds derivatized

with a fluorogenic reagent pre-TLC or post-TLC (e.g.,

amino acids reacted with fluorescamine, or carbamate

pesticides with dansyl chloride after hydrolysis)

The steps in a typical densitometric quantitative

analysis, regardless of analyte type, are the following:

1 Prepare a standard reference solution

2 Prepare a sample solution in which the analyte

is completely dissolved and impurities have

been reduced to a level at which they do not

in-terfere with scanning of the analyte

3 Choose a layer and mobile-phase combination

that will separate the analyte as a compact

zone with an value in the range 0.2 – 0.8

4 Apply the standard and sample aliquots to the

layer using an instrument (Fig 1) or manually

with a micropipette, onto preadsorbent, laned

plates Generally, three or four standard zones

are applied in constant volumes from a series

of standard solutions with increasing

concen-trations, or in a series of increasing volumes

from a single standard solution The sample

volume applied must provide an amount of

an-alyte zone with a weight or concentration that

is bracketed by the standard amounts

5 Develop the plate in an appropriate chamber

and dry the mobile phase under in a fume

hood or oven

R f

4 Storage of the chromatogram also allows

scan-ning to be repeated with various parameters

without time constraints and assures that the

entire sample is available for detection and

scanning

5 Less sample cleanup is often required because

plates are not reused Every sample is analyzed

on a fresh layer without sample carryover or

cross-contamination

6 Solvent use is very low for TLC, both on an

absolute and per-sample basis, leading to

re-duced purchase and disposal costs and safety

concerns

Suggested Further Reading

Fried, B and J Sherma, Thin Layer

Chromatogra-phy — Techniques and Applications, 4th ed., Marcel

Dekker, Inc., New York, 1999, pp 197–222

Jaenchen, D E., Instrumental thin layer

chromatogra-phy, in Handbook of Thin Layer Chromatograchromatogra-phy,

2nd ed (J Sherma and B Fried, eds.), MarcelDekker, Inc., New York, 1996, pp 129 –148

Petrovic, M., M Kastelan-Macan, K Lazaric, and S.Babic, Validation of thin layer chromatographyquantitation with CCD camera and slit-scanning

densitometer, J AOAC Int 82: 25 –39 (1999).

Pollak, V A., Theoretical foundations of optical

quan-titation, in Handbook of Thin Layer phy (J Sherma and B Fried, eds.), Marcel Dekker,

Chromatogra-Inc., New York, 1991, pp 249 –281

Poole, C F and S K Poole, Chromatography Today,

Elsevier, New York, 1991, pp 649 –734

Prosek, M and M Pukl, Basic principles of optical

quan-titation in TLC, in Handbook of Thin Layer tography, 2nd ed (J Sherma and B Fried, eds.), Mar-

Chroma-cel Dekker, Inc., New York, 1996, pp 273 –306

Robards, K., P R Haddad, and P E Jackson, Principles and Practice of Modern Chromatographic Methods,

Academic Press, San Diego, CA, 1994, pp 180 –226

The principal task of chromatography is the separation

of mixtures of substances By “optimization” of the

chromatographic process, we mean enhancement of

the quality of the separation by changing one or more

parameters of the chromatographic system An ability

to foresee, correctly, the direction and scope of these

changes is the most important goal of each

optimiza-tion procedure

Use of chemometrics to devise procedures

suit-able for the most crucial stage of optimization,

opti-mization of selectivity, is generally performed in

three steps:

1 Selection of the experimental method which

best suits the analytical problem considered At

this stage, a chromatographic technique is

cho-sen that ensures that the best possible range of

retention parameters is obtained for each

indi-vidual component of the separated mixture

2 Establishing the experimental conditions that

enable quantification of the influence of the

op-timized parameters of a chromatographic

sys-tem on solute retention

3 Fixing the experimental conditions at values

that provide the optimum separation selectivity

Chemometric optimization of the chromatographic

system consists, in fact, in predicting local maxima in

multiparametric space and, then, in further deciding

which of these parameters is global with regard to the

overall efficiency of a given chromatographic system

Quality of Chromatographic Separations

Elementary Criteria

The simplest way of quantifying the separation of two

chromatographic bands, 1 and 2, is to calculate the

Copyright © 2002 by Marcel Dekker, Inc All rights reserved

ference between their respective retention parameters;that is, the difference between their values,

(1)

or between their values,

(2)where and are the capacity (retention) factors of thechromatographic bands and a is the separation factor.The terms most frequently used to characterize theseparation of two chromatographic bands are the sep-aration factor, a,

(3)where  and the resolution, [1],

(4)where and are the distances of the geometric centers

of two chromatographic bands, 1 and 2, from the origin,

l is the distance from the origin to the mobile phase front,

and and are the diameters of the two graphic bands, measured in the direction of eluent flow.Other elementary criteria include the separation

chromato-factor, S [2],

(5)

the peak-to-valley ratio of the bands, P [3],

(6)

(where f and g are, respectively, the average peak

height and valley depth, characteristic of a given pair ofneighboring solutes on a chromatogram), the frac-tional peak overlap, FO [4],

the product of the separation factors, [8], theproduct of the fractional peak overlap, [9], andthe product of the peak-to-valley ratio of the bands,[10].

There are also other, more complex criteria,

includ-ing the normalized resolution product, r [11],

of chromatograms in liquid chromatography [12].Other criteria are the total peak overlap, [13],

(12)the informing power, [14],

(13)and the chromatographic response function, CRF [10],

(14)(where is the peak-to-valley ratio for the ith pair of

chromatographic bands)

Performance of the Chromatographic System

One measure of the performance of a given graphic system is the number of the theoretical plates

chromato-per chromatographic band (N ) In its simplest form,

this can be defined as

(15)

where l is the distance from the origin to the eluent front and H is the height equivalent to one theoretical plate (H is sometimes also denoted HETP).

The average height equivalent to one theoreticalplate ( ) can be calculated from the relationship [15]H

(where is the surface area of the part of the band

originating from the pure single compound, is

the surface area of the fractional overlap of the nth and

(n 1)th bands, and is the surface area of the

fractional overlap of the nth and (n 1)th bands), and

the selectivity parameter, [5],

(8)where 

Criteria for the Quality of Chromatograms

One method which can be used to establish the

opti-mum conditions for the separation of a complex

mix-ture (i.e., not only a pair) of compounds consists in

searching for the maximum of a function denoted the

chromatogram quality criterion The evaluation of

separation selectivity can be conducted with the aid of

different criteria of chromatogram quality such as the

sum of resolution, [6], the sum of separation

fac-tors, [2], and other sums and products of

elemen-tary criteria, selected examples of which are the

Fig 1 Graphical interpretation of the selected elementary

cri-teria: (a) resolution, (b) the peak-to-valley ratio, P; (c) the

fractional peak overlap, FO.

R s;

(16)where is the standard deviation, which characterizes

the width of the chromatographic spot, or the band

width on the densitogram, is the spot width (or the

width of the peak base on the densitogram),  is

the distance from the origin to the eluent front, and

is the distance from the origin to the geometric center

of the chromatographic spot

The relationship between the height equivalent to

one theoretical plate (H ) and the velocity of the

mo-bile-phase flow is given by the simplified van Deemter

equation [16]

(17)

where u is the linear velocity of the mobile phase, A is

a constant characterizing eddy diffusion, B is a constant

characterizing molecular diffusion, and C is a constant

characterizing resistance to interphase mass transfer

This particular issue is of considerable significance in

planar chromatographic separations, during the course

of which the velocity of the mobile phase changes

The concept of separation number (SN) in planar

chromatography is a practical approach to the task of

quantification of chromatographic system

perfor-mance According to this concept, such performance

can simply be evaluated by calculating how many

com-ponents of the separated mixture can be comfortably

accommodated (i.e., without any overlap of adjacent

components) along the direction of migration of the

eluent A convenient relationship proposed in Ref 17

enables easy calculation of the numerical value of SN:

or, simplified,

(18)where is the width at half-height of a spot at the ori-

gin and is the width at half-height of a spot at 

1 (extrapolated) ( and are in units)

Semiempirical Optimization Strategies

Strategies used for optimization of selectivity can

bacally be divided into three separate groups: (a) the

si-multaneous strategy, (b) the sequential strategy, and

(c) the interpretative strategy

R F

b1 b0

experi-Sequential Strategy

In this strategy, the optimum experimental conditionsare approached in a series of consecutive steps Thechoice of any step results strictly from the outcome ofall those accomplished previously One example of arelevant algorithm is the simplex method [18]; thePRISMA [19] geometrical method is a suitable ex-ample of the overall optimization approach

Interpretative Strategy

This method enables prediction of the quality of a ration on the basis of a relatively limited number of theexperimental data, collected in previous experiments.According to this approach, the chromatographic re-sults are interpreted in terms of the retention functions,valid for each individual solute separately Some goodexamples of the interpretative strategy are the so-called

sepa-“window diagrams” approach [20] and the search forthe extremum of the multiparameter response functionwith the aid of the genetical algorithm [21]

References

1 T Kowalska, Theory and mechanism of thin-layer

chro-matography, in Handbook of Thin-Layer raphy (J Sherma and B Fried, eds.), Chromatographic

Chromatog-Science Series Vol 55, Marcel Dekker, Inc., New York,

8 P J Schoenmakers, A C J H Drouen, H A H Billiet,

and L de Galan, Chromatographia 15: 688 – 696 (1982).

16 J J Van Deemter, F J Zuiderweg, and A Klinkenberg,

Chem Eng Sci 5: 271 (1965).

17. F Geiss, Fundamentals of TLC (Planar phy), Hüthig, Heidelberg, 1987.

Chromatogra-18. W Spendley, G R Hext, and F R Hinsworth, metrics 4: 441 (1962).

Techno-19 Sz Nyiredy, B Meier, C A J Erdelmeier, and

O Sticher, J High Resolut Chromatogr Chromatogr Commun 8: 186 –188 (1985).

20. R J Laub and J H Purnell, J Chromatogr 112: 71–79

(1975).

21. J H Holland, Adaptation in Natural and Artificial tems, University of Michigan Press, Ann Arbor, 1975.

9. R Smits, C Vanroelen, and D L Massart, Fresenius

Zeitschr Anal Chem 273: 1–5 (1975).

10. S L Morgan and S N Deming, J Chromatogr 112:

267–285 (1975).

11 A C J H Drouen, P J Schoenmakers, H A H Billiet,

and L de Galan, Chromatographia 16: 48 –52 (1982).

12. S N Deming and M L H Turoff, Anal Chem 50:

546 –548 (1978).

13. J C Giddings, Anal Chem 32: 1707–1711 (1960).

14. D L Massart and R Smits, Anal Chem 46: 283 –286

(1974).

15. G Guiochon and A M Siouffi, J Chromatogr 245:

1–20 (1982).

Organic Acids, Analysis by Thin Layer Chromatography

Natasˇa Brajenovic´

Rud¯er Bosˇkovic´ Institute, Zagreb, Croatia

INTRODUCTION

A very large number of papers report new research works

on analyses of organic acids by thin layer

chromatogra-phy, paper chromatograchromatogra-phy, high-performance liquid

chromatography, and other analytical techniques One of

the most popular and widely used separation techniques

for qualitative and quantitative analyses in the laboratory

is thin layer chromatography

The reason for using thin layer chromatography is its

wide applicability to a great number of different types of

samples, high sensitivity, and speed of separation with

relatively low cost This technique is very fast, and many

separations can be accomplished in less than an hour

The development of different methods of analysis in

thin layer chromatography is a very important area of

organic chemistry and biochemistry Analysis of organic

acids by thin layer chromatography is widely applied in

different fields of environmental, pharmaceutical,

indus-try, industrial foods, organic chemisindus-try, cosmetics,

clinical, and biochemical assays

OVERVIEW

Here, we present scientific activity in the analysis of

organic acids by thin layer chromatography in a period

from 1993 to 2004 The review is based on a search of

Current Contents and Science Citation Index, using

different combinations of key words relevant for thin

layer chromatography, organic acids, and different kinds

of organic acids, such as amino acids, carboxylic acids,

humic acids, aromatic carboxylic acids, and fatty acids In

addition, the journals publishing papers covering specific

topics related to the analysis of organic acids by thin layer

chromatography were searched directly: Analytica

Chi-mica Acta, Analytical Chemistry, Journal of

Microbio-logical Methods, Journal of Liquid Chromatography and

Related Technologies, Journal of Chromatography (Parts

A and B), Chromatographia, and Journal of

Pharmaceu-tical and Biochemical Analysis

GENERAL CONSIDERATIONSStationary Phase

Typical thin layer separations are performed on flat glass orplastic plates that are coated with a thin and adherent layer

of particles, which constitute the stationary phase.[1]mercial plates come in two categories: conventional [thicklayers (200–250 mm) of particles having sizes of 20 mm orgreater] and high-performance plates (film thickness of

Com-100 mm and particles whose diameters are 5 mm or less).Silica gel is the most extensively used adsorbent in thinlayer chromatography because it leads to excellent, un-complicated separations It can be successfully employedfor both qualitative and quantitative thin layer chroma-tographic analyses It is usually used as a stationary phase

in separations and analysis of alkaloids, various organicacids (especially amino acids and their derivatives), ste-roids, lipids, vitamins, plant pigments, pesticides, drugs,carbohydrates, phenolic substances, etc

Besides silica gel, cellulose, and aluminum oxide,various other impregnated plates are also frequently used

as stationary phases.[1]

Mobile PhaseMany different solvents and mixtures of solvents are used

as mobile phases for the analysis of organic acids by thinlayer chromatography, such as chloroform, ethyl acetate,methanol, benzene, etc

IdentificationFor identification in qualitative thin layer chromatography,

a great number of visualization reagents are used.[1] Ifcompounds are not colored, a UV lamp may be used tovisualize the plates The quantitative determination of samplecomponents is performed according to the following:

1 Extracting the stained spot with solvent and analyzing

ORDER REPRINTS

APPLICATIONS

Pharmaceutical Industry, Medicine,

Biochemistry, and Biology

Thin layer chromatography is used for the analysis of free

amino acids from sanguine plasma in different

progres-sion states in maladies: diabetes, renal syndrome, and

hepatic cirrhosis.[2] Elution was performed on cellulose

plates and the densitometry was achieved with a

photo-densitometer (Shimadzu CS-9000) at 575 nm In the case

of hepatic cirrhosis, a better resolution was obtained A

mixture of n-butanol–acetone–acetic acid–water (35:35:

7:23 vol/vol/vol/vol) was used as the mobile phase

A simple and fast method for identification of

bifidobacteria using thin layer chromatographic analysis

of short chain fatty acids in a culture broth is proposed

(Table 1).[3] This approach has many advantages: the

total time required to analyze organic acids is

approxi-mately 50 min; and the identification protocol is simpler,

quicker, and more economical than conventional

identi-fication methods

Mycolic acids analysis by thin layer chromatography

has been employed by several laboratories worldwide as a

method for fast identification of Mycobacterium.[4]

Mycobacterium tuberculosis strains identified by classical

methods were confirmed by their mycolic acid content

Using aminopropyl-modified silica gel plates in a

normal phase system, the retention behavior of 12 acidic

drugs and biologically active aromatic acids was

inves-tigated by high-performance thin layer chromatography.[5]

The metabolism of aromatic amino acids

(phenylala-nine and tyrosine) can be studied following the excretion

of their characteristic phenolic acid metabolites in urine

using chromatographic methods These apply acids to the

investigations of amino acids themselves in diagnostics

Thin layers of cellulose or silica gel on aluminum foilwere used as stationary phases.[6]The retention factor (Rf)

of clinically important compounds in the three solventsystems is presented in Table 2

Thin layer chromatography is also used for directenantiomeric resolution ofD,L-arginine,D,L-histidine,D,L-lysine, D,L-valine, and D,L-leucine on silica gel platesimpregnated with optically pure (1R, 3R, 5R)-2-azabicy-clo[3,3,0]octan-3-carboxylic acid, which serves as a chiralselector in the pharmaceutical industry.[7]To successfullyresolve D,L-amino acids, various combinations of aceto-nitrile–methanol–water were proposed The spot wasdetected by ninhydrin (0.2% in acetone)

Thin layer chromatography is often applied as anindustrial control procedure in the synthesis of 2-hydroxy-3-naphthalenecarboxylic acids, on silica gel plates withchloroform–methanol–acetic acid (50:20:1) as developer.[8]2-Hydroxy-3-naphthalenecarboxylic acid is an importantintermediate in the synthesis of dyestuffs and drugs.Separation of amino acids and their identification indifferent mixtures are frequent tasks encountered inbiochemistry Thin layer chromatography is a fast, simple,and inexpensive approach to attain this goal Becausesome of the components are UV-inactive, other methods,such as vibrational spectroscopy, should be applied fordetection and identification Comparative study based onRaman spectroscopy of thin layer chromatography spots

of some weak Raman scatterers (essential amino acids)was carried out using four different visible and near-infrared laser radiation wavelengths: 532, 633,785, and

1064 nm.[9] The best results were obtained with simplesilica gel plates

Table 1 Retention time (R f ) and detection color of standard

organic acids

Methyl red + bromophenol blue

a BzAc = benzene–acetic acid (glacial)–water (70:29:1).

b IprBuAm = isopropanol–n-butanol–t-butanol–ammonia–water (40:20: 20:10:20).

ORDER REPRINTS

Environmental, Water, Plant,

and Soil Applications

The occurrence of chlorinated organic compounds in fish

from polluted waters is rather frequent.[10] Chlorinated

carboxylic acids of fatty acid character have also been

shown to account for up to 90% of the extractable

organically bound chloride (EOCl) in fish Purification by

thin layer chromatography of methyl esters of

dichlorote-tradecanoic, dichlorohexadecanoic, and

dichlorooctadeca-noic acids was used They were detected at 1200 ppm of

EOCl in fish

Organic acid complexes with metal ions significantly

affect the mobility of metal ions in plants and soils Toxic

metals also react with organic acids and have a harmfulinfluence on the environment

Many plants contain a variety of free acids such asacetic acid, citric acid, fumaric acid, malic acid, succinicacid, oxalic acid, glycolic acid, etc.[11–13] They are com-ponents of citric cycle, whereas the others are intermedi-ates in the pathway from carbohydrates to aromatic com-pounds.[11] Following extraction, organic acids can beseparated and detected with a variety of techniques Thinlayer chromatographic methods have been also employed

to separate certain organic acids,[11,14,15] as presented inTable 3

Thin layer chromatography of some plant phenolics,which play an important role in plant metabolism, was

Table 3 Thin layer chromatographic procedures for separation of organic acids

(7:2:1, vol/vol) Cellulose powder or silica gel 1-Propanol concentrated in ammonium

hydroxide (7:3 or 3:2) 2-Hydroxybenzoic acid (salicylic acid) Silica gel or Fe(III)-impregnated

silica gel or aluminum oxide

Tap water

silica gel or aluminum oxide

Tap water 3,4,5-Trimethoxybenzoic acid Silica gel or Fe(III)-impregnated

silica gel or aluminum oxide

Tap water 3,4,5-Trihydroxybenzoic acid (gallic acid) Silica gel or Fe(III)-impregnated

silica gel or aluminum oxide

Tap water 4-Hydroxy-3,5-dimethoxybenzoic acid Silica gel or Fe(III)-impregnated

silica gel or aluminum oxide

Tap water 1,2-Benzenedicarboxylic acid Silica gel or Fe(III)-impregnated

silica gel or aluminum oxide

Tap water

Source: Refs [11], [14], and [15].

Table 4 Chromatographic behavior of some plant phenolic acids

Compound

F = fluorescence; Q = quenching; DSA = diazotized sulfanilic acid; DNA = diazotized p-nitroaniline.

Developers: A = 2% formic acid; B = 20% potassium chloride; C = isopropyl alcohol–ammonium hydroxide–water (8:1:1); D = 10% acetic acid Source: Ref [12].

ORDER REPRINTS

carried out on cellulose.[12]The solvents were: 2% formic

acid, 20% potassium chloride, isopropyl

alcohol–ammo-nium hydroxide–water (8:1:1), and 10% acetic acid The

plates were examined under UV light after development

The results of the chromatographic analysis of some

phenolic acids are shown in Table 4

The composition of lignin (an important component of

plant cell walls) is very complex.[13] p-Coumaric acid,

cinnamic acid, ferulic acid, and others are among the

products of lignin biodegradation (Fig 1) Thin layer

chromatography is a very rapid method for their

separation and is usually completed in a very short time

Six different solvents were used as developer (Table 5),

with silica gel being used as a stationary phase

Humic acids are also products of lignin biodegradation

The characterization of humic acids by thin layer

chromatography on Fe(III)-impregnated silica gel with

tap water as developer has been presented.[14]During the

chromatographic process, complexes between Fe(III)

from the support group and the active functional group

from humic acid are formed, causing successive attaching

and detaching of Fe(III) from the support of Fe(III)

hydroxy/oxide These results could partially answer

how the process of metal migration in soils and

sedi-ments progresses

Some carboxy and benzene derivatives related to

humic materials were also examined by thin layer

chromatography.[15] Aluminum oxide, silica gel, andFe(III)-impregnated silica gel plates were used assupports, whereas the mobile phase was water The resultsare presented in Table 3 It was concluded that thehydroxy/oxide layer of iron and aluminum can affect themobility of simple organic molecules in soil On the otherside, organic molecules having carboxy and hydroxygroups can improve the dissolution of hydroxy/oxides ofiron or aluminum in soils and sediments

Fig 1 Chemical structure of simple phenolics and some components of lignin (From Refs [13] and [14].)

Table 5 Different solvent systems were used as developers for thin layer chromatographic separation for simple phenolic and related compounds as cinnamic acid, p-coumaric acid, ferulic acid, and tannic acid

Solvent systems were used as developers for simple phenolic compounds and some acids, which are products

of lignin biodegradation 1) Chloroform–methanol–acetic acid (90:10:1) 2) Petroleum ether (60–80°C)–ethyl acetate–formic acid (40:60:1)

3) Benzene–dioxane–acetic acid (85:15:1) 4) Chloroform–ethyl acetate–acetic acid (50:50:1) 5) Toluene–acetonitrile–formic acid (70:30:1) 6) Petroleum ether (60–80°C)–methanol–acetic acid (90:10:1) Source: Ref [13].

ORDER REPRINTS

Food Analysis, Agriculture, and Industry

There are several acids that are widely used in industries

For example, synthetic carboxylic acid esters are used in the

perfume industry Benzoic acids are used as sodium salts in

the food industry as inhibitors of microorganism growth.[16]

Benzoic acid derivatives often contain amino, hydroxy,

carboxy, and nitro groups Analysis of substituted benzoic

acids by thin layer chromatography was performed on silica

gel, polyamide, and cellulose containing UF254fluorescent

indicator.[17]For the mobile phase, different mixtures were

used: hexane–acetic acid; hexane–ethyl acetate–formic acid;

chloroform–methanol–phosphoric acid; cyclohexane–acetic

acid; benzene–ethanol; etc Because benzoic acid

deriva-tives have similar retention parameters, their separation

requires a thorough optimization of conditions (the nature of

the stationary phase, the composition of the mobile phase,

and the pH of the solutions)

For the separation of benzoic acids, planar

electro-chromatography was used.[18] In this approach, an

electroosmotic flow is used to drive the mobile phase

in thin layer chromatography Planar

electrochromatog-raphy has several advantages over classical thin layer

chromatography, especially substantially faster

separa-tion For example, separation by planar

electrochroma-tography can be 10 times faster than that using ordinary

thin layer chromatography

Thin layer chromatographic analysis of agricultural

products, foods, beverages, and plant constituents is

described by Sherma[19]in a review paper In laboratories

throughout the world, thin layer chromatography is widely

used for food analysis and qualitative control Numerous

applications of thin layer chromatography have been

reported in the area of food composition, involving

determinations of compounds such as lipids, sugars,

amines, vitamins, and organic acids such as amino acids

and fatty acids

VARIOUS ORGANIC COMPOUNDS

Thin layer chromatographic analysis is also highly

applicable to the determination of aromatic organic

acids.[20] In human organisms, aromatic acids are

syn-thesized as metabolites in intoxication by toluene, xylene,

and ethyl benzene.[16] These compounds are easily

ab-sorbed through the skin or respiratory system, and are

oxidized to aromatic acids The separation,

identifica-tion, and quantitative analyses of aromatic acids are also

necessary because they appear as semiproducts of the

biosynthesis of aromatic amino acids in plants (phenolic

acids), and metabolites of numerous toxic substances,

drugs, and catecholamines Polar adsorbents and

polar-bonded stationary phases are also widely used incarboxylic separation by thin layer chromatography, oftencoupled with densitometry

Dansyl chloride (DNS-Cl; lene-1-sulfonyl chloride) is used in analytical chemistry tofluorescently label substances This process of dansylationcreates fluorescent derivatives, which can be detectedwith great sensitivity The method for the dansylation ofhydroxyl (–OH) and carboxylic acid (–COOH) functionalgroups has been described.[21] Fluorescent labeling bydansyl chloride has applications in liquid chromatogra-phy, high-performance liquid chromatography, thin layerchromatography, and mass spectrometry Fast thin layerchromatography was accomplished using acetone asthe resolving solvent, and resulted in good differentiation

5-dimethylaminonaphtha-of analytes

Many thin layer chromatography systems for theseparation of amino acids have been described.[22–24]Copper sulphate and polyamide were tried as impreg-nants for improving the separation of 20 amino acids onsilica gel layers MeOH–BuOAc–AcOH–pyridine (20:20:10:5) was used as the solvent system.[22]

D-enantiomers of amino acids have been frequentlyreported in various tissues of diverse organisms A simpleand rapid method of separating optical isomers ofamino acids on a reverse-phase thin layer chromatogra-phy plate is described.[23]Amino acids, derivatized with1-fluoro-2,4-dinitrophenyl-5-L-alanine amide, were spot-ted onto a reverse-phase thin layer chromatography plate.Acetonitrile in triethylamine–phosphate buffer was used

as the developer

For the evaluation of protein structure, identification ofamino acids is extremely important Thin layer chro-matography is an appropriate method in this field Avariety of spray reagents are used, among which ninhy-drin is the most popular one due to its high sensitivity.[24]Ninhydrin produces a purple/violet color with most aminoacids A typical experimental setup includes chromato-graphic plates prepared from silica gel; n-propanol–water(70:30) mixture was used as a mobile phase For com-plex mixtures of substances, two-dimensional chroma-tography is preferred, using n-propanol–water (70:30)mixture and methanol–chloroform (3:1) mixture as thetwo developers

CONCLUSIONThin layer chromatography is a very widely used chroma-tographic technique in research activities of analyticalchemists in many laboratories in the world Many articlesdealing with the development of new analytical methodsfor the analysis and separation of different organic acids in

ORDER REPRINTS

various fields have been published There is a constant

need for qualitative and quantitative analyses of organic

acids in the pharmaceutical, cosmetic, and food industries;

in medicine, biology, organic chemistry, and biochemical

analysis; and in environmental studies The important

reasons for frequent applications of thin layer

chromatog-raphy in qualitative and quantitative analyses of organic

acids are its high sensitivity, fast separation of

compo-nents, and relatively low cost

REFERENCES

1 Skoog, D.A.; Holler, J.F.; Nieman, T.A High-Performance

Liquid Chromatography In Principles of Instrumental

Analysis, 5th Ed.; Vondeling, J., Sherman, M., Bortel, J.,

Messina, F., Pecuilis, V., Eds.; M Brooks/Cole, Thomson

Learning, Inc.: London, 1998; 725 – 766.

2 Gocan, S.; Ghizdavu, L.; Ghizdavu, L TLC of some free

amino acids from sanguine plasma J Pharm Biomed.

Anal 2001, 26, 681 – 685.

3 Ki-Yong, L.; Jae-Seong, S.; Tae-Ryeon, H Thin layer

chromatographic determination of organic acids for rapid

identification of bifidobacteria at genus level J Microbiol.

Methods 2001, 45, 1 – 6.

4 Leite, C.Q.F.; Desouza, C.W.O.; Leite, S.R.D

Identifica-tion of mycobacteria by thin layer chromatographic

analysis of mycolic acids and conventional biochemical

method—Four years of experience Mem Inst Oswaldo

Cruz 1998, 93 (6), 801 – 805.

5 Bieganovska, M.L Retention behaviour of some acids

drugs and biologically active compounds on silica and

aminopropyl silica layers Chem Anal 1995, 40 (6), 859 –

867.

6 Ersser, R.S.; Oakley, S.E.; Seakins, J.W.T Urinary

phenolic acids by thin-layer chromatography Clin Chim.

Acta 1970, 30, 243 – 249.

7 Bhushan, R.; Martens, J.; Thuku Thiongo, G Direct thin

layer chromatography enantioresolution of some basic

DL-amino acids using a pharmaceutical industry waste as

chiral impregnating reagent J Pharm Biomed Anal.

2000, 21, 1143 – 1147.

8 Revilla, A.L.; Havel, J.; Borovcova´, J.; Vrchlabsky, M.

Capillary zone electrophoresis of

hydroxynaphthalenecar-boxylic acids Purity monitoring of b-hydroxynaphthoic

acid in industry J Chromatogr., A 1997, 772, 397 – 402.

9 Istva´n, K.; Keresztury, G.; Sze´p, A Normal Raman and

surface enhanced Raman spectroscopic experiments with

thin layer chromatography spots of essential amino acids

using different laser excitation sources Spectrochim Acta,

Part A: Mol Biomol Spectrosc 2003, 59, 1709 – 1723.

10 Mu, H.; Wese´n, C.; Nova´k, T.; Sundin, P.; Skramstad, J.; Odham, G Enrichment of chlorinated fatty acids in fish lipids prior to analysis by capillary gas chromatography with electrolytic conductivity detection and mass spec- trometry J Chromatogr., A 1996, 731, 225 – 236.

11 Dashek, W.V; Micales, J.A Isolation, Separation, and Characterization of Organic Acids In Methods in Plant Biochemistry and Molecular Biology; Dashek, W.V., Ed.; CRC Press: Boca Raton, FL, 1997; 107 – 113.

12 Jangaard, N.O Thin-layer chromatography of some plant phenolics J Chromatogr 1970, 50, 146 – 148.

13 Sharma, O.P.; Bhat, T.K.; Singh, B Thin-layer tography of gallic acid, methyl gallate, pyrogallol, phloroglucinol, catechol, resorcinol, hydroquinone, cate- chin, epicatechin, cinnamic acid, p-coumaric acid, ferulic acid and tannic acid J Chromatogr., A 1998, 822, 167 – 171.

chroma-14 Iskric´, S.; Hadzˇija, O.; Kveder, S Behaviour of humic acids on Fe(III)-impregnated silica gel compared with model substances J Liq Chromatogr 1994, 17 (7), 1653 – 1657.

15 Kveder, S.; Iskrc´, S.; Zambeli, N.; Hadzˇija, O The behaviour of some benzene derivatives on thin layers of aluminum oxide—Comparison with plain and Fe(III) impregnated silica gel J Liq Chromatogr 1991, 14 (18), 3277 – 3282.

16 Waksmundzka-Hajnos, M Chromatographic separation of aromatic carboxylic acids J Chromatogr., B 1998, 717,

93 – 118.

17 Sumina, E.G.; Shtykov, S.N.; Dorofeeva, S.V Ion-pair reversed-phase thin-layer chromatography and high-per- formance liquid chromatography of benzoic acids J Anal Chem 2002, 57 (3), 210 – 214.

18 Nurok, D.; Koers, J.M.; Carmichael, A Role of buffer concentration and applied voltage in obtaining a good separation in planar electrochromatography J Chroma- togr., A 2003, 983, 247 – 253.

19 Sherma, J Thin-layer chromatography in food and tural analysis J Chromatogr., A 2000, 880, 129 – 147.

agricul-20 Sherma, J Planar chromatography Anal Chem 1992, 64, 134R – 147R.

21 Bartzatt, R Dansylation of hydroxyl and carboxylic acid functional groups J Biochem Biophys Methods 2001,

47, 189 – 195.

22 Srivastava, S.P.; Bhushan, R.; Chauhan, R.S TLC separations of amino acids on silica gel impregnated layers J Liq Chromatogr 1984, 7 (7), 1359 – 1365.

23 Nagata, Y.; Iida, T.; Sakai, M Enantiomeric resolution of amino acids by thin-layer chromatography J Mol Catal.,

B Enzym 2001, 12, 105 – 108.

24 Laskar, S.; Sinhababu, A.; Hazra, K.M A modified spray reagent for the detection of amino acids on thin layer chromatography plates Amino Acids 2001, 21, 201 – 204.

Organic Extractables from Packaging Materials:

Chromatographic Methods Used for Identification

and Quantification

Dennis Jenke

Baxter Healthcare Corporation, Round Lake, Illinois, U.S.A.

INTRODUCTION

Plastic materials are widely used in medical items, such

as solution containers, transfusion sets, transfer tubing,

devices, and packaging systems The physiochemical

na-ture of these materials provides medical products with

their necessary, desirable performance characteristics

While an important performance characteristic of

plas-tics used in medical application is chemical inertness,

interactions between a plastic material and a contacted

pharmaceutical product are well documented Such

in-teractions may include sorption, the uptake of product

components by the plastic material, or leaching, i.e., the

release of plastic material components into the product In

the case of leaching, both the identities of the leached

substances and their accumulation levels may impact the

ultimate utility of the product

A review, related to the chromatographic methods used

to assess the accumulation of leachable substances from

packaging materials used for pharmaceutical products, is

provided This considers methods used to identify and/or

quantify such leachables in actual products or

product-simulating solvent systems

DISCUSSION

The assessment of the impact of the accumulation of

leached substances in pharmaceutical products contacted

by a plastic material during their manufacture, storage,

and/or use is a multifaceted undertaking involving

dis-ciplines within the applied physical, chemical, and

bio-logical sciences While numerous strategies can be

en-visioned, and have been utilized to perform such an

assessment, considerations include the identification of

the leached substances and the measurement of the actual

or probable accumulation levels of the identified

sub-stances The identification process is an extensive

in-vestigation that utilizes sensitive and information-rich

scouting of analytical methods for the dual purposes of

first revealing the leachables and then providing vant information (e.g., formula and structure) that leads

rele-to their identification In the worst-case scenario, such

an analytical investigation is conducted blind; that is,the analytical team is faced with the unenviable chal-lenge of finding an unknown number of unknown com-pounds, many of which accumulate in the product atlevels much lower than its other constituents These cons-tituents may include both additives and nonmaterial-related contaminants such as ingredient impurities anddegradation products This search for material-derivedleachables in pharmaceutical products is greatly facili-tated if it is conducted with information-rich analyticalmethodologies that exhibit a comprehensive ability torespond to a large population of analytes in both a uni-versal, but very specific, manner The dual performancerequirements of universality and specificity are the pri-mary reasons why chromatographic methods are almostexclusively used in investigations specifically associatedwith organic leachables

Given the variety of packaging materials used inpharmaceutical applications, the population of potentialprimary and secondary organic leachables is large andcompositionally diverse While an analytical chemist has

a multitude of chromatographic tools with which to form a leachables assessment, some guidance in terms ofsuccessfully applied strategies and methods can greatlyfacilitate the assessment Thus this article contains a ge-neral compilation of published chromatographic methodsand strategies that have been successfully applied to theidentification and quantification of packaging materialleachables Examples are provided for each major sepa-ration strategy [e.g., high-performance liquid chromato-graphy (HPLC), gas chromatography (GC), thin-layerchromatography (TLC), and supercritical fluid chromato-graphy (SFC)] and for most commonly employed de-tection methods [e.g., ultraviolet (UV), mass spectrometry(MS), and flame ionization detection (FID)] While thecompilation in Tables 1 – 4 is by no means exhaustive, it issufficiently broad in scope to provide the investigatorwith a general overview of the ways in which chromato-

DOI: 10.1081/E-ECHR 120018659

Copyright D 2003 by Marcel Dekker, Inc All rights reserved.

Polyolefin packaging material

graphy has been applied to meet the objectives of a

leachables investigation

Tables 1 – 4 provide general method details, such as

column type, elution and detection conditions, and other

operating conditions The materials investigated, as well

as the specific leachables examined, are also indicated

General comments are provided in terms of sample

preparation Given the number of methods cited, it is not

possible here to provide detailed chromatographic

pro-files, which are readily available in the cited references

Preinjection sample preparation is not a

chromato-graphic issue per se Nevertheless, it is an important

consideration in the successful application of a complete

analytical process Nerin et al.[36]reviewed sample

treat-ment techniques applicable to polymer extract analysis,

including headspace methods, supercritical fluid

extrac-tion, and solid phase microextraction

REFERENCES

1 Aignasse, M.F.; Prognon, P.; Stachowicz, M.; Gheyouche,

R.; Pradeau, D A simple and rapid HPLC method for

determination of DEHP in PVC packaging and releasing

studies Int J Pharm 1995, 113, 241 – 246.

2 Yagoubi, N.; Baillet, A.; Pellerin, F.; Ferrier, D

Physico-chemical behavior of b-irradiated plastic materials

cur-rently used as packages and medical products Nucl.

Instrum Methods Phys Res 1995, 105, 340 – 344.

3 Begley, T.H.; Hollifield, H.C Migration of dibenzoate

plasticizers and polyethylene terephthalate cyclic

oligo-mers from microwave susceptor packaging into

food-simulating liquids and foods J Food Prot 1990, 53,

1062 – 1066.

4 Begley, T.H.; Hollifield, H.C Liquid chromatographic

determination of residual reactants and reaction

by-products in polyethylene terephthalate J.A.O.A.C 1989,

72, 468 – 470.

5 Rybak, K.E.; Sarzynski, W.; Dawidowicz, A.L Migration

of antioxidant additives from polypropylene investigated

by means of reversed phase high performance liquid

chromatography Chem Anal 1992, 37, 149 – 159.

6 Yagoubi, N.; Baillet, A.; Mur, C.; Baylocq-Ferrier, D.

Determination of phenolic antioxidants in pharmaceutical

formulations by liquid chromatography and migration

study on HDPE packagings Chromatographia 1993, 35,

455 – 458.

7 Sarbach, C.; Yagoubi, N.; Sauzieres, J.; Renaux, C.;

Ferrier, D.; Postaire, E Migration of impurities from a

multi-layer plastics container into a parenteral infusion

solution Int J Pharm 1996, 140, 169 – 174.

8 Reif, O.W.; Solkner, P.; Rupp, J Analysis and evaluation

of filter cartridge extractables for validation in

phar-maceutical downstream processing PDA J Pharm Sci.

Technol 1996, 50, 399 – 410.

9 Biles, J.E.; McNeal, T.P.; Begley, T.H.; Hollifield, H.C.

Determination of Bisphenol-A in reusable polycarbonate food-contact plastics and migration into food-simulating liquids J Agric Food Chem 1997, 45, 3541 – 3544.

10 Buiarelli, F.; Cartoni, G.; Coccioli, F HPLC and GC-MS determination of compounds released to mineral waters stored in plastic bottles of PET and PVC Annal Chim.

1993, 83, 93 – 104.

11 Ulsaker, G.A.; Teien, G Identification of caprolactam as a potential contaminant in parenteral solutions stored in overwrapped PVC bags J Pharm Biomed Anal 1992,

10, 77 – 80.

12 Arbin, A.; Jacobsson, S.; Hanninen, K.; Hagman, A.; Ostelius, J Studies on contamination of intravenous solutions from PVC-bags with dynamic headspace GC-

MS and LC-diode array techniques Int J Pharm 1986,

28, 211 – 221.

13 Snell, R.P Solid-phase extraction and liquid graphic determination of monophthalates and phthalide extracted from solution administration sets J.A.O.A.C.

chromato-1993, 76, 531 – 534.

14 Tiller, P.R.; El Fallah, Z.; Wilson, V.; Huysman, J.; Patel,

D Qualitiative assessment of leachables using dependent liquid chromatography/mass spectrometry and liquid chromatography/tandem mass spectrometry Rapid Commun Mass Spectrom 1997, 11, 1570 – 1574.

data-15 Yu, K.; Block, E.; Balogh, M LC-MS analysis of polymer additives by electron and atmospheric-pressure ionization: Identification and quantification LC GC 2000, 18, pp.

162, 164, 166, 168, 170, 172, 174, 176, 178.

16 Danielson, J.W Toxicity potential of compounds found in parenteral solutions with rubber stoppers J Parenter Sci Technol 1992, 46, 43 – 47.

17 Dobias, J.; Voldrich, M.; Proks, M Migration of polyethylene terephthalate oligomers from packaging into food simulant liquids Potav Vedy 1996, 14, 25 – 32.

18 Snell, R.P Gas chromatographic determination of hexanone leached from hemodialysis tubing J.A.O.A.C.

cyclo-1993, 76, 1127 – 1132.

19 Snell, R.P Capillary GC analysis of compounds leached into parenteral solutions packaged in plastic bag J Chromatogr Sci 1989, 27, 524 – 528.

20 Ulsaker, G.A.; Teien, G Determination of stearate ester in intravenous solutions stored in poly(vinyl chloride) bags, using gas chromatography – single-ion monitoring mass spectrometry Analyst 1984, 109, 967 – 971.

9,10-epoxy-21 Milano, C.J.; Bailey, L.C Evaluation of current compendial physicochemical test procedures for pharmaceutical elasto- meric closures and development of an improved HPLC procedure PDA J Pharm Sci Technol 1999, 53, 202 – 210.

22 Kim-Kang, H.; Gilbert, S.G Permeation characteristics of and extractables from gamma-irradiated and non-irradiated plastic laminates for a unit dosage injection device Packag Technol Sci 1991, 4, 35 – 48.

23 Kim-Kang, H.; Gilbert, S.G Isolation and identification of potential migrants in gamma-irradiated plastic laminates

by using GC/MS and GC/IR Appl Spectrosc 1991, 45,

572 – 580.

Organic Extractables from Packaging Materials: Chromatographic Methods Used for Identification and Quantification 11

24 Kim, H.; Gilbert, S.G.; Johnson, J.B Determination of

potential migrants from commercial amber polyethylene

terephthalate bottle wall Pharm Res 1990, 7, 176 – 179.

25 Veiga-Rial, M.; Sarria-Vidal, M.; de la

Montana-Migu-eslez, J.; Simal-Gandura, J Identification of residual

constituents in plastic packaging for dairy products Rec.

Res Devel Agric Food Chem 1999, 3, 305 – 311.

26 Demertzis, P.G.; Franz, R.; Welle, F The effects of

g-irradiation on compositional changes in plastic packaging

films Packag Technol Sci 1999, 12, 119 – 130.

27 Salmona, G.; Assaf, A.; Gayte-Sorbier, A.; Airaudo, C.B.

Mass spectral identification of benzothiazole derivatives

leached into injections by disposable syringes Biomed.

Mass Spectrom 1984, 11, 450 – 454.

28 Castle, L.; Offen, C.P.; Baxter, M.J.; Gilbert, J Migration

studies from paper and board food packaging materials I.

Compositional analysis Food Addit Contam 1997, 14,

35 – 44.

29 Eiceman, G.A.; Karasek, F.W Identification of residual

organic compounds in food packages J Chromatogr 1981,

210, 93 – 103.

30 Vitali, M.; Leoni, V.; Chiavarni, S.; Cremisini, C

Deter-mination of 2-ethyl-1-hexanol as contaminant in drinking

water J.A.O.A.C 1993, 76, 1133 – 1137.

31 Azuma, K.; Hirata, T.; Tsunoda, H.; Ishitani, T.; Tanaka,

Y Identification of volatiles from low density ene film irradiated with an electron beam Agric Biol Chem 1983, 47, 855 – 860.

polyethyl-32 Berg, B.E.; Hegna, D.R.; Orlien, N.; Greibrokk, T termination of low levels of polymer additives migrating from polypropylene to food simulated liquids by capillary SFC and solvent venting injection Chromatographia 1993,

De-37, 271 – 276.

33 Dilettato, D.; Arpino, P.J.; Nguyen, K.; Bruchet, A vestigation of low mass oligomers and polymer additives from plastics Part II: Application to polyolefin soxhlet extracts J High Res Chromatogr 1991, 14, 335 – 342.

In-34 Moulder, R.; Kithinji, J.P.; Raynor, M.W.; Bartle, K.D.; Clifford, A.A Analysis of chemical additives in poly- propylene films using capillary supercritical fluid chro- matography J High Res Chromatogr 1989, 12, 688 – 691.

35 Corti, P.; Murratzu, C.; Franchi, G.; Lencioni, E.; Pancini,

R Evaluation of chemical residues in food and drugs: Evaluation of an antioxidant given from high density polyethylene food containers Acta Toxicol Ther 1988, 9,

205 – 221.

36 Nerin, C.; Rubio, C.; Salafranca, J.; Batlle, R The simplest sample treatment techniques to assess the quality and safety of food packaging materials Rev Anal Chem.

2000, 19, 435 – 465.

12 Organic Extractables from Packaging Materials: Chromatographic Methods Used for Identification and Quantification

Conventional thin-layer chromatography (TLC) in our

experience, known under the name planar

chromatog-raphy, uses horizontal or vertical glass or Teflon

cham-bers for the development of chromatograms As

sta-tionary phases, commonly known adsorbents or

supports based on silica gel, aluminium oxide,

magne-sium silica, cellulose, and so forth are used; particle

sizes are about 20 mm The migration of the mobile

phase is based on the phenomenon of capillary forces

This chromatographic method is described, in detail, in

other sections of this volume

This method is characterized by many limitations

which either can cause unsatisfactory separation of a

mixture of substances or lead to long development

times (even up to several hours) or, sometimes, makes

use of solvents of high viscosity impossible The

efficiencies of such chromatographic systems are also

rather low

Discussion

In conventional TLC, the velocity of chromatogram

de-velopment depends on the dimension of stationary-phase

particles, viscosity of the mobile phase, distance from

the start line of the mobile phase, and other parameters

Therefore, there is no possibility of regulation of

resolu-tion by change of migraresolu-tion velocity of mobile-phase

flow; the distance between the solvent reservoir and the

solvent front ( ) varies with time (t) according to

where k is a constant that depends on the

chromato-graphic system (mobile phase and adsorbent) and the

size of sorbent particles constituting the layer and

presents a parabolic relationship

As the most popular planar liquid chromatographic

technique, TLC uses a vapor phase of solvent above

the sorbent layer, which has an important influence on

the resolving power

Many of the inconveniences of TLC are avoided in

overpressured layer chromatography (OPLC), which

Copyright © 2002 by Marcel Dekker, Inc All rights reserved

is a logical extension of the theory and practice oped in high-pressure liquid chromatography (HPLC)which can now be used in the field of planar liquidchromatography This extension offers some excep-tional advantages to a chromatographer OPLC is, inpractical terms, a planar HPLC technique OPLC inte-grates many of the benefits of TLC, high-performanceTLC, and HPLC This technique corresponds to anHPLC column having a relatively thin, wide cross sec-tion and using a pressurized ultramicrochamber withstandard chromatoplates Eluent is forced into the sor-bent layer by the means of a pump which enables de-velopment of chromatograms with forced flow of themobile phase (more precise penetration into micro-pores) The eluent migrates against the sorbent resist-ance imposed by external pressure on the sorbent sur-face, and the vapor phase is excluded In the case ofOPLC, there is a linear relationship between the dis-tance ( ) of the solvent front from the starting point

devel-and the migration time (t):

where k is a constant that depends on the rate of

sol-vent flow, on the externally applied pressure, and onthe size of the particles constituting the layer In prin-

ciple, k is constant throughout the development and

independent of the rate of solvent migration InOPLC, the parameter is also used to describe theposition of the separated analyte and, in this case, thevalues do not depend on the starting distance.The parameters characterizing chromatographic

systems in TLC, such as average plate height (H ), duced plate height (h), and theoretical plate number (N ), are calculated in a similar way in OPLC, but, prac-

re-tically, they do not depend on either average particlediameter ( ) or the start distance ( ) In OPLC, thestart distance has no influence on the efficiency of sep-aration, and the average plate hight is nearly constant

on a layer of exceptionally fine particles, even over alonger development distance Thus, the major advan-tage of OPLC over other planar techniques lies in thisfact We can say that OPLC permits relatively largeplate numbers to be obtained and can be applied morefavorably in the case of smaller particles

vents the eluent from flowing off the chromatoplate in

an unwanted direction According to the technique ofchromatogram development used (unidirectional, bidi-rectional, circular, on-line, off-line, parallel coupledmultilayer, serial coupled multilayer), all four margins

of a plate, three margins, two opposite margins can beimpregnated, or they can be left uncoated

In linear development of a chromatogram, tional or bidirectional developments of the chromato-gram are possible Similarly, as in liquid column chroma-tography, there are possible, in this case, either on-line oroff-line techniques of sample application, separation, anddetection, as well as various modifications (e.g., partlyoff-line method) Bidirectional development can also bevertical Using vertical bidimensional development, ap-plying different eluents, components of complex, difficultmixtures can be separated The separation of such mix-tures is also possible by means of this technique usingmultiple automatic development of chromatogram

unidirec-In OPLC, the changes in composition of the eluentgive good possibilities for special separation tech-niques such as isocratic and gradient separation Thechoice of mobile phase can be effectively and quicklydetermined using the optimization model Prisma ac-cording to Nyiredy This is a three-dimensional modelthat correlates the solvent strength and the propor-tion of eluent constituent, which determines the se-lectivity of the mobile phases according to Snyder’ssolvent classification

The newest apparatus for overpressured layer matography, “Personal OPLC BS-50” manufactured

chro-in Hungary, is shown chro-in Fig 2 Generally speakchro-ing, itconsists of the separation chamber and a liquid deliv-ery system The separation chamber contains the fol-lowing units: (a) holding unit, (b) hydraulic unit,(c) troy layer cassette, and (d) drain valve The appa-ratus also has a pumping system for eluent delivery andfor the hydraulic liquid delivery The entire apparatusand the total chromatogram development process arecontrolled by a computer system The apparatus andthe method of chromatogram development is charac-terized by high reproducibility of results and chro-matographic parameters External pressure (5 MPa),eluent flow rate and its volume, and development timecan be automatically programmed The OPLC-BS-50chamber works in off-line and on-line systems Usingthis equipment, it is possible to separate 70 –100 oreven more samples at the same time, depending, ofcourse, on the chosen technique of the OPLC process.This OPLC method is charaterized by high preci-sion for determination of retention parameters and re-producibility of results

Basic Instrumentation for OPLC

As far as OPLC is concerned, the method, in principle,

differs from conventional TLC in the design of the

equipment that is used The first attempts at

construc-tion of chromatographic pressure chambers were

made in the beginnings of the 1960s However, only at

the end of the 1970s, Tyihák, Mincsovics, and Kalász

were successful in construction of a well-operating

OPLC chromatograph called Chrompress 10

(maxi-mum pressure permitted in this chamber was 1.0 MPa)

and, later, in the 1980s, Chrompress 25 (2.5 MPa)

(La-bor MIM, Hungary) and the most modern Personal

OPLC BS-50 (OPLC-NIT, Budapest, Hungary) (5.0

MPa) In Poland, during 1980 –1990, a pressure

thin-layer chromatograph was constructed (Cobrabid,

War-saw); however, due to its narrow range of operating

pressures (0.8 MPa), it was not widely used

We will describe only the most up-to-date OPLC

sys-tem This fully automatic OPLC system allows

separa-tion of mixtures on an analytical and on a

semiprepara-tive scale The fundamental separation process occurs

on a chromatographic plate (constructed of glass or

alu-minum foil) with sorbent (Fig 1) covered and

com-pressed by a special polyethylene or Teflon foil,

pres-sured by water In this way, a flat, thin chromatographic

column is created This technique also requires a special

chromatoplate which is sealed at the edges, which

pre-Fig 1 Cassette of foil-backed layer for linear development:

1  cover sheet, 2  sorbent layer, 3  eluent puncture, 4 

eluent trough, 5  sample application site.

Separation of components which the former TLC

techniques failed to achieve

Smaller attainable plate height over longer

migra-tion distances

Rapid separation and high resolution for industrial

control

Optimization of resolution as a function of solventvelocity, development distance, and temperaturePossibility of using high-viscosity eluents and poorlywettable stationary phases

Possibility of both quantitative evaluation andpreparative applications

Efficient separation of multicomponent samplesDifferent development modes: unidirectional, bi-dimensional, continuous on-line, and off-lineLong migration distances on fine-particle layerswith short development times

No air interactionsMinute consumption of developing solventProgrammable operating system

Some applications include analytical and tive analyses in all types of biological, biochemical,pharmaceutical, clinical, forensic, food, and environ-mental laboratories

prepara-Suggested Further Reading

Kaiser, R E., Einfuerung in die HPLC, Huethig,

Hei-delberg, 1987

Kaiser, R E and R I Rieder, in Planar phy, Vol 1 (R E Kaiser, ed.), Huethig, Heidelberg,

Chromatogra-1986, p 165

Mincsovics, E., K Ferenczi-Fodor, and E Tyihák, in

Handbook Thin-Layer Chromatography (J Sherma

and B Fried, eds.), Marcel Dekker, Inc., New York,

1996, p 173

Nyiredy, Sz., C A J Erdelmeier, and O Sticher, in

Proc Int Symp TLC with Special Emphasis on Overpressured Layer Chromatography (OPLC) (E.

Tyihak, ed.), LABOR MIM, Budapest, 1986, p 222.Nyiredy, Sz., S Y Meszaros, K Dallenbach-Toelke, K

Nyiredy-Mikita, and O Sticher, J High Resolut Chromatogr Chromatogr Commun 10: 352 (1987) Róz˙yio, T K., R Siembida, and E Tyihak, Biomed Chromatogr 13: 1 (1999).

Ruoff, A D and J C Giddings, J Chromatogr 3: 438

Tyihák, E., E Mincsovics, P Tetenyi, I Zambo, and H

Kalász, Acta Horticult 96: 113 (1980).

Fig 2 Personal OPLC BS-50 apparatus: 9,10  Teflon cover

sheet of cassette, upper and lower, respectively; 11a  tube

driving eluent from pressure gauge to chamber inlet connector;

11b  end connector for tube 11a; 11c  chamber–eluent–

inlet connector; 12a  tube for eluent outlet of the chamber;

12b  end connector for tube 12a; 12c 

chamber–eluent–out-let connector; 20  display; 21  function keys; 22  numeric

keys; 23  pressure gauge for eluent; 24  eluent switching

valve; 25  tank holder; 26  eluent tanks A, B, and C; 27 

pump head for hydraulic liquid delivery; 28  pump head for

eluent delivery; 29  hydraulic liquid and eluent connecting

tubes; 30  connecting stub for syringe to fill up eluent pump;

31  middle hole of T distributor for fitting tubes in case of

two-directional development; 32  left hole of T distributor for

fitting tubes in case of two-directional development; 33  right

hole of T distributor for fitting tubes in case of two-directional

development; 34  hole for piston rinsing against deposition;

35  syringe.

Liquid chromatography (LC) was the first

chromato-graphic mode to be developed in the beginning of the

twentieth century For almost 70 years, it was

em-ployed without major modifications until the end of

the 1960s when an instrumental version of liquid

chro-matography was finally produced Before this

mile-stone, LC was performed mainly in large-bore glass

tubing packed with large-diameter solid particles To

differentiate the instrumental version developed in the

late 1960s from the noninstrumental, usually referred

as the “classical” version, the former was named

high-pressure liquid chromatography and, later,

high-per-formance liquid chromatography (HPLC) Because

HPLC used smaller particles as the stationary phase,

the columns had to be packed at higher pressures in

or-der to obtain a more stable bed required by the higher

pressures used in these techniques Altogether, HPLC

offered a much higher efficiency (number of plates)

than “conventional LC” and, as a consequence, higher

resolution (separation power) as well The standard

HPLC columns used in the 1970s consisted of particles

of 5 –10 µm packed in stainless-steel tubing of

4.0 – 4.6 mm inner diameter (i.d.) and 15 –25 cm long

Typical flow rates under these conditions are 1–2

mL /min In a typical quality control laboratory (8 h a

day; 5 days a week, 20 days a month; 12 months a year),

more than 100 L of chromatographic solvent are

gen-erated in a 1-year period Most of these solvents are

highly toxic to man and the environment, requiring

special waste storage, transportation, and final

dis-posal As a consequence, a miniaturization of the

HPLC techniques using less solvent became important

immediately after its development

A major step in the miniaturization of HPLC

col-umns was done early in 1967 by Horváth and

co-work-ers [1,2], when investigating the parametco-work-ers that

influence the separation of nucleotides in a 1-mm-i.d

column These columns were then named microbore

columns A further step in the miniaturization process

was done in 1973, by Ishii and co-workers, by

separat-ing polynuclear aromatic hydrocarbons (PAHs) in a

0.5-mm-i.d PTFE column The term micro-LC wasthen introduced to differentiate this technique fromHPLC, which uses larger-bore columns [3 –5] Shortlyafter, Scott and Kucera published several articles deal-ing with microbore (1-mm-i.d columns) LC [6,7]

In spite of the fast development in its early days(late 1960s and early 1970s), the miniaturization ofHPLC followed a slow progress until recently, with thedevelopment of LC – mass spectrometry (MS) usingelectrospray-type interfaces

Capillary Liquid Chromatography

Capillary liquid chromatography (CLC) is a mode ofHPLC that deals with columns having internal diame-ters equal to, or smaller than, 0.5 mm This number islimited by the internal diameter of the fused-silica tub-ing commercially available, which is the most populartubing used in this area The CLC columns are usually

15 – 60 cm long, having internal diameters 0.5 mmand being either coated or packed with the stationaryphase Due to the small inner diameter of the CLC col-umns, this technique is more demanding in instrumen-tation than HPLC, particularly with respect to the sol-vent delivery, sample introduction, and detectionsystems

Sample Introduction

Because the column inner diameter is small, theamount of stationary phase is also very small and, as aconsequence, the amount of sample that can be intro-duced into the column without overloading is verysmall (typically a few nanoliters) In most cases, thepreferred sample introduction system consists of an in-jection valve containing an internal loop smaller than0.1 mL

Pumping System

Because the eluent flow rate is relatively small cally a few microliters per minute), the pump used todeliver it to the column is critical There are two ma-

(typi-2 Packed Capillary Liquid Chromatography

jor approaches being used: pumps capable of

deliver-ing flow rates in the range of few microliters per

min-ute (usually syringe-type pumps) or reciprocating

pumps using a flow splitter In both cases,

repro-ducible flow rates are hard to obtain using

commer-cially available pumps

Detectors

Almost all detectors currently used in HPLC have

been evaluated to be used in CLC The major

modification required, in most cases, is a decrease in

the detector cell volume in order to accommodate the

small sample volume without considerable peak

broad-ening Ultraviolet-visible (UV-vis), fluorescence,

elec-trochemical, mass spectrometric, and several other

de-tectors have been successfully used with CLC

Columns

Capillary LC columns can be generally made in two

different ways: wall-coated open tubular (WCOT) or

packed columns Coating the internal wall of the

tub-ing (usually fused silica) with a thin film of a

solvent-re-sistant polymer makes WCOT columns and is the same

technology as used for capillary GC columns Usually,

cross-linked or immobilized phases are preferred in

or-der to avoid stationary-phase removal by the eluent

The major drawback of these columns is that they have

to be made with an internal diameter smaller than 20

mm in order to be highly efficient for complex

separa-tions, thus justifying their use instead of the packed

capillary columns [8] This places great demands on

the instrumentation, the eluent quality, the sample

preparation step, and so forth, thus making it

imprac-tical at this moment As an alternative, the packed

cap-illary columns using the technology already available

to prepare HPLC columns is less instrument

demand-ing and have been gaindemand-ing more acceptance, every day

becoming the preferred form of CLC

Advantages of Capillary Liquid Chromatography

The advantages of CLC are consequences of its

minia-turization [9] Due to its miniaturized size, it requires

much less stationary phase than does ordinary HPLC

and, as a consequence, more expensive phases can be

used to prepare the columns This includes chiral

phases, experimental new materials, expensive

bio-compounds, and so forth In the same way, the amount

of mobile phase is very small, thus leading to a savings

in buying, storing, and discarding the solvent, allowing

the use of expensive eluents such as deuterated vents and chiral modifiers such as cyclodextrins, tran-sition metals, and so forth In many cases, the totalamount of mobile phase in one separation is just 10 mL(1 mL /min; 10-min run); this explains why this tech-nique is sometimes referred to as “one-drop chroma-tography.” The amount of sample injected is also verysmall, so it becomes an important technique when thesample size is critical, such as in biomedical studies(brain, spine liquid, newborn tests, etc.), forensicchemistry (fire debris, explosives, blood residues), en-vironmental analysis, and several other applicationfields

sol-Other advantages of CLC, when compared toHPLC, includes its higher permeability [10], chemicalinertia, easier coupling to other separation andidentification systems such as mass spectrometry, gaschromatography, and nuclear magnetic resonance, andthe possibility of making longer columns, thus achiev-ing more plates (efficiency) and resolution

Figure 1 shows a chromatogram of a separation ofPAHs using a packed capillary column As can be veri-fied, a good separation is obtained with minuteamounts of stationary phase, mobile phase, andsample

Limitations of Capillary Liquid Chromatography

In spite of the several advantages over HPLC, CLC hasnot yet achieved its maturity as a separation technique

to be used worldwide, particularly as a routine nique for quality control laboratories Among the lim-itations still hindering the further development ofCLC, one of the most critical ones is the very limitedavailability of commercial equipment dedicated to this

tech-17.8

-2.9

35.3 tempo (min)

solid-technique Even so, most systems are simple

adapta-tions of parts already used for HPLC, by just

decreas-ing their sizes and volumes without specifically havdecreas-ing

CLC in consideration Therefore, in order to become a

routine technique as its counterpart in gas

chromatog-raphy, capillary liquid chromatography still has to have

a broader interest for the instrument manufacturing

companies in the technique before it will spread out

beyond the academic environment Those who have

worked with packed columns in gas chromatography

in the 1960s have already seen this same history

Acknowledgments

Professor F Lanças wishes to thank FAPESP

(Fun-dação de Apoio à Pesquisa do Estado de São Paulo)

and CNPq (Conselho Nacional de Desenvolvimento

Científico e Tecnológico) for financial support to his

7 R P W Scott and P Kucera, J Chromatogr 169: 51 (1979).

8. H Menet, P Gareil, and R Rosset, Anal Chem 56:

Particle Size Distribution: A Key

Property of Particulate Samples

Particle size distribution analysis was considered, in

one of the latest Pittsburgh Conferences, as one of the

most outstanding trends in analytical science This is

not an overstatement, as most of the real samples of

analytical interest occur either in dispersed form or in

dispersed matrices Just for argument’s sake, in

indus-trial applications the characterization of the size of

sample particles is routine and is an essential part of

the overall quality control procedures In the medical

field, for particles used to carrier drugs, the size is a

critical performance factor (e.g., liposomes) In the

food industry, the alcoholic yield from fermentation of

starch, and even the taste of chocolate, depends on the

size of particles of which these samples are composed

Before discussing our method for determining

par-ticle size, it is necessary to briefly review the definition

of size distribution If all particles of a given system

were spherical in shape, the only size parameter would

be the diameter In most real cases of irregular

par-ticles, however, the size is usually expressed in terms of

a sphere equivalent to the particle with regard to some

property Particles of a dispersed system are never of

either perfectly identical size or shape: A spread

around the mean (distribution) is found Such a spread

is often described in terms of standard deviation

How-ever, a frequency function, or its integrated

(cumula-tive) distribution function, more properly defines not

only the spread but also the shape of such a spread

around the mean value This is commonly referred to

as the particle size distribution (PSD) profile of the

dis-persed sample

An examination of technical literature and trade

publications indicates that a wide variety of

instru-ments are commercially available for PSD analysis [1]

The classical methods are based on either electrical

properties (e.g., the Coulter Counter® principle) or

optical properties (e.g., laser scattering) of the analyte

However, none of these techniques are separation

methods Because particulate dispersions are often

Copyright © 2002 by Marcel Dekker, Inc All rights reserved

highly complex in terms of the polydispersity index,multimodal size distribution, and density, it is hardlypossible, without the use of separative methods, to ob-tain an accurate determination of their size distribu-tion Among separative chromatography-like meth-ods, one can consider hydrodynamic chromatographyand field-flow fractionation (FFF) The application toPSD of a subset of the latter family of methods is thetopic of this article

PSD by FFF

Field-flow fractionation is a broad family of liquidchromatographic-like techniques which have beenshown, over more than 20 years, to be able to frac-tionate and characterize high-molecular-weight spe-cies in a size range spanning five orders of magnitude,from macromolecules to micron-size particles [2] FFFhas been demonstrated to be a rapid method for thedetermination of the mean diameters and polydisper-sities of particulate samples When compared to stan-dard methods for PSD analysis, the main advantage ofFFF lies in the fact that FFF is a separation methodwhich has some common features with liquid chroma-tography The output from an FFF experiment is afunction of the detector signal versus the retentiontime of the analyte Whereas, in liquid chromatogra-phy, such an analytical response is referred to as the

chromatogram, in FFF it is commonly defined as the fractogram However, when it is compared to classical

liquid chromatography, the existence of a direct tionship between retention and some physical proper-ties of the analyte, such as the size, is a fundamentalfeature of most FFF techniques The theory of FFF re-tention has been fully explained elsewhere, [2, andreferences therein] as well as in other entries of thisencyclopedia What is important to focus on here isthat, in FFF, particle size determination of the analytecan be obtained by means of a direct numerical con-version of the retention scale, whereas the relativeamount of separated analyte is, as in the case of chro-matograms, in some way proportional to the signal in-tensity The basic procedures for the two conversions

rela-is the topic of threla-is entry

the fractogram the well-known, approximate sion that is valid for highly retained samples in GrFFF[3]:

expres-(1)

where (cm) is the diameter value corresponding to

the ith data point of the fractogram, the retention

of the GrFFF channel, w (cm) is the channel thickness

and g is the so-called hydrodynamic correction factor,the knowledge of which is, therefore, required for PSDanalysis

In practice, PSD curves can be obtained directlyfrom the experimental, digitized peak (fractogram),once it is converted to a function of particle di-ameter with the use of Eq (1) The frequency func-tion of particle size is expressed as [3]

(2)

where is the mass concentration of the

analyte at the ith digitized point and and arethe differences in retention volume and particle diam-

eter between the ith and the (i 1)th digitized points,respectively The incremental quantity can be cal-culated for any given by Eq (1)

As far as the conversion of the analytical response

is concerned, the most used detectors in GrFFF havebeen, until now, conventional ultraviolet (UV) detec-tors commonly used for HPLC With this type of de-tector, the amount of particles with diameter is pro-

portional to the detector response at the ith point With

particulate samples, in fact, because of UV detector

op-tics, the response is a turbidity signal read within an

angle between the incident light and the photosensor(i.e., usually smaller than 10°) rather than the ab- sorbance This turbidity signal can be assumed to be di-

rectly proportional to the sum of all cross-sectional eas of the particulate sample components at any time.The validity of the above assumption, in the case of par-ticles which are about 10-fold larger than the incidentwavelength, is discussed elsewhere [6] The mass fre-quency function can thus be expressed as [7]

1

V r, i

Gravitational FFF: An Economical Device

for PSD Analysis of Micron-Size Dispersions

Here, we treat the case of PSD analysis of particulate

systems of micron-size range (i.e., with a size

distribu-tion extending above 1 mm) Since 1994, in our

labora-tories, this topic has been dealt with by means of a

low-cost subset of sedimentation FFF (SdFFF), the

gravitational field-flow fractionation (GrFFF)

tech-nique [3] GrFFF had already been applied to the

frac-tionation of a variety of micron-size dispersion, either

inorganic as commercial chromatographic supports [4]

or biological, as cells and parasites [5] In no cases,

however, had PSD been performed through GrFFF

As an SdFFF subset, GrFFF requires the application

of a sedimentation field that, in this case, is simply

Earth’s gravity applied perpendicularly to a very thin,

empty channel with a rectangular cross section The

big advantage of GrFFF, compared to other

tech-niques for the characterization of particulate matter,

lies in its very low cost ($50 for a homemade

chan-nel) and easy implementation in a standard

high-pres-sure liquid chromatography (HPLC) system (the

chan-nel can simply replace the standard HPLC column)

The GrFFF channel employed here can be easily built

as described elsewhere [3 – 9] It is basically a

ribbon-like capillary channel which consists of two

mirror-pol-ished plates, of either glass or plastic material (e.g.,

polycarbonate) which are clamped together over a thin

sheet of either Teflon or Mylar from which the channel

volume has been removed Simplicity and economy of

use make it possible for laboratories that are not

spe-cialized in PSD analysis to perform dimensional

char-acterization of supermicron particles dispersions with

limited effort and cost

We shall show here that GrFFF is capable of

per-forming reliable, quantitative PSD analysis of

particu-late matter Some basics of the overall procedure will

be overviewed and the relevant questions presented

In fact, in order to obtain a PSD by means of GrFFF,

the conversion of the retention time axis into the

ana-lyte size axis is necessary For the same reason, the

de-tector signal axis must be converted into mass (or

con-centration) of the fractionated analyte

Procedure and Discussion

From a GrFFF Fractogram to a PSD

The diameter scale can be obtained from retention

co-ordinates (i.e., the retention time axis) by applying to

analysis for a sample of silica particles commonly used

as the stationary phase in HPLC (5 mm LiChrospher,

Merck) is reported in Fig 2a For the sake of

compar-ison, the PSD of the same sample obtained by laser

dif-fraction is reported in Fig 2b We can observe that

PSD resolution is higher in GrFFF than in laser

dif-fraction, where just a histogram is obtained On the

other hand, accuracy is comparable when the

experi-mental distribution moments (i.e., the percentiles

indi-cated as ) are compared to the nominal

val-ues given by the manufacturer We must point out that

differences in distribution moment values as high as

10% are commonly reported when different,

uncorre-lated techniques for PSD studies are compared [1]

Quantitative Particle Size and Sample

Amount Distribution in GrFFF

We have derived an original method by which

quanti-tative particle size and sample amount distribution

(PSAD) in GrFFF can be obtained by applying to Eq

(2) a derivation of the Lambert–Beer law in

flow-through systems [7] If compared to standard PSD, a

PSAD thus represents a distribution of the real mass of

the analyte as a function of size, rather than a

func-tional expression only proporfunc-tional to mass.

For particle dispersions in the micron-size interval,

which is the typical application range of GrFFF, it has

been demonstrated that the sample amount exiting the

detector cell, can be expressed as

d10,

Where (min) is the peak area, F is the flow

rate, b (cm) is the cell thickness, and K is the tal extinction coefficient of the particulate sample Ifthe extinction coefficient can be assumed to be ap-proximately constant, as in the case of particles whosesize is at least 10 times higher than the incident wave-length [6], the detector reading, expressed as “ab-

to-sorbance” at the ith point is related to the real bidity signal from the detector by the equation [7]

Fig 1 Scheme for the conversion from a fractogram to a

fre-quency function of particle size.

(a)

(b)

Fig 2 Comparison between normalized, frequency functions

of size of silica particles for HPLC packing (5 mm; Chrospher, Merck, Darmstad); nominal distribution percen-

line): curve 2 (dashed line): curve 3 (dashed line):

(a) GrFFF/PSD: Sample load: 100 mg; channel: mobile phase: Milli-Q water/ Triton 0.1% (v/v)/ 0.02% (w/v); flow rate:

UV detection: 330 nm; experimental g  0.70 (b) Laser tion PSD (Malvern MASTERSIZER ® , Malvern Instruments Ltd., UK).

diffrac-0.004 cm 3 /min; 1.010 

d50 3.7 mm,

d10

1F m % 2

just a subset, as noted earlier A description of thehydrodynamic and other secondary effects on GrFFFretention is far above the introductory nature of thisentry However, just to introduce the reader to the pos-sibility of obtaining a direct conversion of retention tosize by predicting g, the above-mentioned model can

be used for relating retention volume to particle meanelevation during elution as follows [10]:

(8)where (cm) is the distance of the center of the par-ticles from the accumulation wall and

is an empirical function It was shown that der optimized experimental conditions, in a properlydesigned GrFFF system, a balance between secondaryeffects of forces other than hydrodynamic forces cangive negligible effects [9] In this way, particle elevation

un-is predictable In thun-is case, also, the value of g can beestimated, thus allowing for the direct conversion ofretention time to analyte size without previous calibra-

tions (standardless) This possibility of calculating g is

a task still in progress and it will open more promisinguses of GrFFF for dimensional analysis of suspendedparticulate matter, because PSD can be obtained in the

“single-run” mode (i.e., without previous calibration).This could be a significant enhancement in the futureevolution of GrFFF/PSD In fact, in the GrFFF/PSD example in Fig 2a, the conversion from retention

to size was performed only by means of an tal evaluation of the parameter g, with a calibrationplot formerly obtained with standards [3]

experimen-Acknowledgments

Giancarlo Torsi, Dora Melucci, Andrea Zattoni, andGabriele Berardi of the Department of Chemistry “G.Ciamician,” Bologna, Italy, are duly acknowledged

5 A Bernard, B Paulet, V Colin, and Ph J P Cardot,

Trends Anal Chem 14: 266 (1995).

6. P Reschiglian, D Melucci, and G Torsi, matographia 44: 172 (1997).

where is the analyte concentration at the ith

point of the fractogram; that is,

Therefore, the turbidity signal can be expressed as

(6)and the PSD expression for the frequency function,

which is, in Eq (3), just proportional to particle mass,

can be transformed into a real function in mass directly

from Eq (6), thus giving [7]

(7)Integration of Eq (7) yields the cumulative distri-

bution which gives, plotted as percent distribution, the

size distribution percentiles Moreover,

once it is related to the injected sample amount, it gives

the cumulative distribution of analyte mass as a

func-tion of size, with its asymptotic value giving the total

sample recovery Some examples of GrFFF/PSAD of

silica samples used as HPLC column packing are

re-ported in Ref 7

Direct Conversion of Retention to Size; Secondary

Effects

As shown earlier, the direct conversion from retention

time to particle diameter values [Eq (1)] requires that

the correction factor g is predicted or experimentally

estimated It is known that, in GrFFF, g can be

influenced by either hydrodynamic or other effects as

those due to the mobile phase [8] and the channel

walls’ nature [9] All of these effects can influence

par-ticle size determination by GrFFF

We have been developing an approach to the

evalu-ation of the second-order effects, which act on GrFFF

retention, and to the prediction of the correction

fac-tor g Among these effects, prominent are those due to

hydrodynamic forces which lift the analyte particles

away from the accumulation wall during their elution

GrFFF really shows a significant dependence of

reten-tion (and, thus, of the parameter g) on the flow rate:

The higher the flow rate, the higher the lift and,

there-fore, the lower the retention In order to evaluate

par-ticle lift and, thus, parpar-ticle retention, the semiempirical

model given by Williams et al has been applied

[10 –12] This model is known to predict particle

eleva-tion from the accumulaeleva-tion wall in sedimentaeleva-tion

field-flow fractionation (SdFFF), of which GrFFF it is