HAROLD 8.
CASSIDY
CHROMA TOG RAP^ is a method by means of whi::h
mixtures may be separated so that under favorable
conditions the components are obtained pure. But
since there are so many kinds of mixtures, and so many
different processes that may be used in analyzing them,
it is necessary to look a t the entire realm of analysis
in order t o see clearly the special features that dis-
tinguish chromatography and the various chroma-
tographies.
There are three kinds of analysis: qualitative, quanti-
tative, and relationship (or structural) analysis. These
can be sufficiently clearly marked off from each other
under normal conditions that they can be used for
classification. (There are situations in which it may
be difficult to make a clear distinction between any
two, but these are not of particular importance here.)
Chromatography is a form of qualitative analysis. As
with all qualitative methods chromatography may he
made quantitative, but the method, in its strictly
chromatographic aspects, only separates mixtures
and tells neither how much of each component is pres-
ent nor the relationship of that component to the
others in the original mixture. It is important to
notice this because the apparatus and procedures of
chromatography are essentially very simple. The
quantifying accessories are usually of necessity com-
plicated. They are important and necessary parts of
the apparatus, but are only adjuncts to the chromatog-
raphy. This has t o be emphasized because when this
distinction is not made the chromatographic parts of
analyses often disappear under the tremendous variety
and complexity of the associated instrumentation,
and what is essentially simple from a functional point
of view becomes hidden in a mass of contrivances.
Qualitative analysis (as well, also, as quantitative
and relationship analysis) may be applied to all kinds
of mixtures. These may be classed for convenience
as mixtures of bulk-level matter, as molecular mixtures,
and as submolecular level compositions. These classes
are not rigidly defined, since some large molecules
approach particulate bulk (colloidal) matter in size,
and since separations of ions are much more con-
veniently classed as molecular than as submolecular
Presented as part of the Symposium an Vapor Phase Chroma-
tography before the Division of Analytical Chemistry a t the
129th Meeting of the American Chemical Soci~ty,Dallas, 1956,
snd taken largely from the manuscript of a new book on chrome
tagraphy now in preparation.
'Communication No. 1370 from the Sterling Chemistry
Laboratory, Yale University.
Yale University, New Haven, Connecticut
species separations. This classification is for the
purposes of utility rather than for the files of the
formalist.
Chromatography belongs among the methods appli-
cable to molecular mixtures. For this reason it is
wrong to speak of chromotography as a filtering proc-
ess. Filtration applies to the separation of bulk
particulate matter (see Table 1) and it is misleading
to speak of filtering molecular mixtures apart in a
chromatography apparatus. This means, also, that
in order to investigate a substance chromatographically
it must be reduced to a molecular mixture. This is
sometimes not ensured, and may account for some of
the nonmoving material occasionally found at the top
of a column or a t the site of the initial zone on paper,
whence the mention here.
There are in general two types of qualitative analy-
tical methods for separating n~olecular mixtures:
the chemical and the physical. Through both of these
methods the object is to convert the molecular mixture
into new mixtures of more desirable compositions, or
in the most favorable cases into pure const~ituents,
which are separable by bulk procedures. For example,
a chemical separation of a molecular mixture of silver
as the nitrate from sodium nitrate might depend on
plating out the silver, inducing it to deposit on a bulk
electrode that may later be removed mechanically from
the system. An example of a physical method is
found in distillation. The mixture to be separated
is partially volatilized under conditions of temperature
(and reflux, perhaps) that produce a vapor of different
composition from the liquid, then these two phases are
mechanically separated, usually by condensiug the
vapor separately.
Put in general terms, the homogeneous molecular
mixture (a phase) is converted to a heterogeneous
system, the phases of which can then be separated by
mechanical means. This is typical of most of these
separations at the molecular and submolecular levels.
The distinction between chemical and physical classes
(Table 1) is convenient and not absolute. In the
former, the interactions that lead to separation are
chemical in nature; in thelatter, physical. Chromatog-
raphy belongs among the physical methods, though as
Tswett pointed out in the early 1900's there is no real
reason why there may not be developed chromatog-
raphies that depend on chemical reactions. Some of
these are, indeed, known.
Among the qualitative methods that depend pri-
VOLUME 33, NO. 10, OCTOBER, 1956 483

TABLE 1
The Field of Analvsis
Analysis
Q~sl!~stive I .
Quantitat~ve Relatio!nship
May be a plied to
f
Mixtures of bud-level materials
I
Molecular1 m~xtnres
. ~ubmolecular-levelbornpositions
I
Methods of sepktion involve: Methods of separation rely on Methods of sepmsdon involve breaking
Sorting, filtering, mechanical
1 of chemical bonds
I
operations of all kinds physical interxtions ChemicalIresctions
II
I I 1
Distribution Thdmal 1.
Gravitat~onal Electrophoretio Electromagnetic Mass spectrographic
hetweenphases analysis analysis analysis analysis analysis (some
kinds)
(See ~ a 6 l 2)
e

marily on physical interactions me can list a group DEFINITION OF CHROMATOGRAPHY

that utilizes distribution between phases (Table 1).
Chromatography falls in this .group. The group
(Table 2 ) contains two classes of distribution systems:
those that involve two bulk phases; and those that
involve a bulk phase and a thin phase or film. The
first kind includes distillation, absorption, extraction,
crystallization, and sublimation. The second kind
includes all the chromatographies. It should be
pointed out that although only distributions between
two phases are classified here, the classification may
be extended t o more phases. Further, as with the
previous categories, there are regions of overlap.
For example it may be difficult to decide in certain
cases how thick a film must grow to be before it is
classed as a bulk phase. Also, the classification does
not include combinations of two or more methods into
one, as in extractive distillation. Such possible
combinations can be derived from these tables.
The nature of chromatography is not fully delineated
until one further characteristic is brought out. The
pairs of phases of all kinds, listed in Table 2 , can he
brought into contact (so that the distribution can
occur) in two basically different ways: in batchwise
(which includes for our purposes cocurrent and cascade)
and in differential countercurrent contacting. It is
the second of these methods of contacting that charac-
terizes chromatography.
It can he understood, from Table 2, why it is that
"chromatography" is such a tremendously extensive
and varied field. Whereas among the systems that
are produced by distribution between two bulk phases
that are contacted in a differential countercurrent
manner each system has a different name: distillation
under reflux, extraction in a packed tower, etc., when
a bulk phase and a thin phase are utilized all the
systems are called chromatographic. The field of
chromatograpy is therefore comparable in extent to
that which includes all the others.
We are now in a position to define chromatography
and to classify the different kinds of chromatography
under the generic definition. Chromatography com-
prises a group of methods for separating molecular
mixtures that depend on distribution between a bulk,
usually mobile phase and a thin, usually stationary
phase. These are brought into contact in a differential
countercurrent manner. I n most chromatographic
methods, the bulk mobile phase flom over the thin
stationary phase. I n all cases the molecules of the
mixture to he separated distribute between the two
phases. Those that tend more to the stationary phase
are retarded with respect to those that tend more to
enter the mobile phase; the latter thus move away, in
the stream of mobile phase, from the former and become
separated. The mobile phase may be a solvent,
developer, eluent, or displacer depending on its function.
When the bulk mobile phase is a gas, or vapor, and
the thin stationary phase is a liquid film held stationary
on a support, the system comprises "gas-liquid partition
chromatography" (GLPC) or "gas partition chromatog-
raphy" (GPC). When the bulk mobile phase is a gas
and the stationary phase is at the surface of an adsorb-
ent, the system is "gas adsorption rhromatography"
(GAC). The two may not always be distinguishable
with absolute certainty, in which case a more non-
committal term "gas chromatography" may be used.
When the hulk mobile phase is a liquid, and the thin
stationary phase is a liquid film, a gel, or imbibed layer,
the system is "liquid-liquid partition chromatography,"
or "partition chromatography," a special rase of which
is "paper partition chromatography." If the sta-
tionary phase is furnished by the surface of a solid
adsorbent, the system comprises classical "adsorption
chromatography," or Tswett-column analysis. "Ion-
exchange chromatography" falls also in this group.
These may not be clearly distinguishable under all
484
TABLE 2
Cross-table of Phase-pair Distributions, and the Methods that They Generate
--
Bulk phases
Gas or vapor ( G )
Liquid ( L )
! Distillation,
Absorption
Solid ( S )
I Sublimation
Gs
T h i n phases
Liquid flm or "mobile" inkrfaeial filni ( M )
! Gas partition chroma-
tography ( G P C )
M-G
In this rase an artificial interface, a membrane or barrier, must be put between the phases.
circumstances, for, as Tiselius pointed out, if an ad-
sorbed film becomes thick enough the conditions for
partition chromatography may arise.
Finally, if the hulk phase is liquid and the thin film is
the interfacial region formed at the surface of bubbles
of gas, or droplets of liquid passing through the hulk
phase, we have foam or emulsion rhromatography, as
the case may he.
These six types comprise all the kinds of chromatog-
raphy that differ because of the phases employed.
There are, however, differences in technique that
enable a further differentiation between chromato-
graphic methods. I n principle, all of the above kinds
of chromatography may be carried out by the tech-
niques of frontal analysis, development and elution
analysis, displacement and carrier displacement analy-
sis, and gradient elution analysis. Frontal analysis
is an old method of using a bed of adsorbent. The
mixture to be treated is passed into the bed until
break-through occurs, when a front of the least strongly
retarded component of the mixture issues from the
bed of stationary phase, followed by a front containing
also the next least strongly held component, and so on,
with, in favorable cases, one front for each component
(including the solvent). The very old method of
"capillary analysis" belongs in this category. In
development analysis, an amount of the mixture to
be analyzed that is small relative to the capacity of
the bed of stationary phase is applied to the bed. The
developer then separates the components of this mixture
out, as described above, to form zones of the com-
ponents in the bed. (The bed of stationary phase
may, of course, he a strip or sheet of paper.) In
elution analysis these zones are made to issue from the
bed one after the other. Displacement and carrier dis-
placement analyses employ more strongly adsorbed
substances, added t o the mixture and already present
in the mixture, t o displace less strongly adsorbed ones,
so that the zones move along the bed and out of it in a
compact array. Gradient elution analysis employs a
JOURNAL OF CHEMICAL EDUCATION
Bulk phases
Lipid (L) Solid ( S )
L-L
Extraction
absorption
S-L
Crystallization
S-S
(Enfleurage)
M-L
Liquid-liquidpartitionchromatog~shy,
foamand emulsion chromstograpk
developer with gradually increasing eluting power to
crowd the rear part of each zone toward the main body
of the zone, thus decreasing the ill effects of tailing.
These names were given to the methods by Tiselius
and his co-workers, who invented the displacement
techniques.
I t has been pointed out that in all kinds of chromatog-
raphy the distribution process that leads to a separa-
tion occurs between a bulk phase and a thin phase in
differential countercurrent contact. A consequence
of this method of contacting is that there is never any
over-all equilibrium in the system (though a steady
state may be achieved temporarily and locally).
A further conseqnence is that as one phase is a thin
film, and hence a region of somewhat limited freedom
for the molecules that enter it, geometrical factors
may become important in influencing the separation.
This is particularly so when the thin film is a t the
surface of a solid adsorbent, for here not oniy is the
film thin, often of molecular thickness, hut also the
surface of the adsorbent is practically always inhomo-
geneous in terms of interactive sites as well as of topol-
ogy. This may explain the very subtle differentiations
that occur between molecules admitted to the stationary
phase and those largely excluded from it.
GAS CHROMATOGRAPHY
A special feature of gas adsorption and gas partition
chromatographies is that the distribution picture is
greatly simplified, especially in the latter case, com-
pared with the chromatographies that employ hulk
liquid phases. Where the mobile phase is a hulk
liquid, the separation depends on competitive inter-
actions between the molecules of the mixture and those
of the mobile and stationary phases. (The competi-
tions involve, also, possible associative effects.) Fur-
ther, there is an additional competition for the station-
ary phase by the molecules of the mobile phase. While
these three competing interactions add a tremendous
range of sensitivity and subtlety to these chromato-
VOLUME 33, NO. 10, OCTOBER, 1956
graphic methods, they do complicate prartice and
theory. In the gas-liquid partition method two of
thesc competing interactions virtually vanish. The
carrier gas (the bulk mobile phase) is usually chosen
to he virtually inert toward both the stationary phase
and the mixture to he separated. Hence there are
only the interactions between the molecules of the
mixture and those of the stationary phase to be ron-
sidered in theory and practice, a t least to a good first
approximation. (For this reason the mutual inter-
artions of different kinds of molecules in the mixture
may make their effects more noticeable.) It there-
fore is easier to see, in gas-liquid partition chroma-
tography, the operation of dispersion, dipole-dipole,
polarizability, and hydrogen bonding interactions as
they operate in the separations of diffcrent mixtures
on different stationary phases.
The gas chromatographies are heing rapidly de-
veloped, and a great service would he done to workers
in this field, as well as to newcomers, if suitable no-
menrlature were agreed upon. At present, a t least
485
four different names are heing used for gas partition
chromatography, and two for gas adsorption chroma-
tography The inventors of the partition method,
Drs. A. T. James and A. J. P. Martin, called it gas-
liquid partition chromatography, a name shortened
by the Shell Development Group to GLPC. In the
discussion during the last session of the Dallas sym-
posium it seemed generally agreed that suitahle names
for the t.echniques are "gas partition chromatography,"
and "gas adsorption chromatography," with the more
general term "gas chromatography" to include
To these names can he added the terms that qualify the
terhniques of operation, if that should be necessary,
namely frontal analysis, elution analysis, displacement
analysis, and gradient analysis. In the last case the
type of gradient, whether concentration, temperature,
or whatever else, may have to be specified. General
willingness to use these names would contribute to the
rlarity of the growing literature on gas chromatography.
Letter from tho chairman of the Symposium, Dr. Charlo8 R.
Willinghnm.