Enhancement of Retention by Ion-Pair Formation in Liquid
`Chromatography with Nonpolar Stationary Phases
`
`Csaba Horvath,* Wayne Melander, Imre Molnar,’ and Petra Molnar
`
`Chemical Engineering Group, Department of Engineering and Applied Science, Yale University, New Haven, Connecticut 06520
`
`In lon-pair reversed-phase chromatography, the retention of
`lonized analytes on a nonpolar bonded stationary phase Is
`enhanced by the presence of a “hydrophobic” counterion
`(hetaeron) in the mobile phase. Either ion-pair formation In
`the mobile phase with relatively strong retention of the complex
`or the conversion of the stationary phase Into an lon-exchanger
`may explain the phenomenon. Analysis of the pertinent
`equilibria shows that the observed hyperbolic or parabolic
`dependence of the capacity factors on the hetaeron con-
`centration cannot shed light on the mechanism. The ex-
`perimental data obtained for the retention of catecholamines
`by using C,-C,,. alkyl sulfates and other similar hetaerons in
`a wide concentration range, however, could be mechanistically
`interpreted from the chain length dependence of the pa-
`rameters for the relationship between the capaclty factors and
`hetaeron concentration. Although the results clearly dem-
`onstrate that in the system investigated, lon-pair formation
`governs retention, ion-exchange mechanism can be operative
`under certain conditions. Changes In retention upon addition
`of salt to the eluent are treated both theoretically and ex-
`perimentally. The effect of organic solvents on the behavior
`of the chromatographic system Is discussed in view of the
`proposed theory.
`
`According to the popular notion, the selectivity in chro-
`matographic separations is determined bythe differencesin
`the equilibrium distribution of eluite molecules, i.e. analytes,
`between the stationary and mobile phases. Quite frequently,
`however, secondaryequilibria between the eluites and certain
`species present in the eluent can drastically affect retention
`(1). In a recent paper (2) we analyzed theeffect of protonic
`equilibria in the mobile phase on retention in liquid chro-
`matography with nonpolar stationary phases. The selectivity
`of the chromatographic system for ionogenic eluites was shown
`to be greatly influenced bytheir dissociation constant and the
`hydrogen ion concentration in the eluent because the binding
`of protons increases and decreases the retention of weak acids
`and bases, respectively.
`Recent work has demonstrated that retention of charged
`eluites on nonpolar bonded stationary phases can be aug-
`mented by the presence of suitable counterions, which have
`a substantial hydrophobic moiety, in the mobile phase (3-5).
`This technique is often referred to as “soap” or “ion-pair
`reversed-phase” chromatography. The counterions used in
`the mobile phase belong in the group of detergents such as
`alkyl sulfonates and sulfates or tetraalkylammonium com-
`pounds, and ion-pair formation between the eluite and
`counterion is assumed to be responsible for the increase in
`retention (6).
`This approachis particularly interesting because the use
`of nonpolar bonded stationaryphases for liquid chromato-
`graphic separations has a wide currency (7).
`In most cases
`
`‘Present address, K. G. Dr. Herbert Knauer, Hegauer Weg 38,
`Berlin 37, West Germany.
`
`octadecyl-silica is the stationary phase and hydro-organic
`mixtures with methanolor acetonitrile as well as neat aqueous
`or organic solvents are used as eluents. The technique is
`simple and can be used for the separation of a wide variety
`of substances. For this reason it is the most popular method
`in high performance liquid chromatography. The mechanism
`of retention is believed to be the same as involved in the
`so-called hydrophobic effect (8) and we recently adapted the
`solvophobic theory (9) to treat the interaction of eluites with
`the hydrocarbonaceous functions of bonded phases in a
`rigorous thermodynamic fashion (2, 10). We have shown that
`the magnitude of retention is governed by the effect of the
`solvent on these species and their adducts.
`The increasing popularity of using an ion-pair forming agent
`in the mobile phase to increase the retention of oppositely
`charged eluites prompted us to investigate the fundamental
`aspects of this technique. Our treatment, however, is quite
`general and applicable to a variety of cases where the retention
`of an eluite on nonpolar bonded phases is enhanced by
`complex formation with a componentof the eluent. For sake
`of convenience we propose to call the complexing agent
`hetaeron, a term derived from the Greek work for companion
`(eratpov), Thus “hetaeric” chromatography would denote a
`technique in which a certain concentration of a complexing
`agentis intentionally maintained in the mobile phase in order
`to affect the selectivity of the chromatographic system by
`secondary equilibria. The name should berestricted to
`situations where the eluite-hetaeron complex is formed in the
`mobile phase and distributed between the two phases. Of
`course, secondary equilibria may change the properties of the
`stationary phase. Indeed, it was recently suggested (11) that
`in ion-pair reversed-phase chromatography the stationary
`phase acts as a dynamically coated ion exchanger because of
`the adsorption of the detergent ions. By using extensive
`experimental data and applying the solvophobic theoryfor
`the interpretation of the results, we intend to demonstrate
`in this article that in the situations examined the mechanism
`of the chromatographic process entails ion-pair formation in
`the mobile phase and binding of the neutral complex to the
`stationary phase.
`
`THEORY
`
`Phenomenological Treatment. In orderto shed light on
`the relationship between the capacity factor, which is a
`convenient measure of retention, and the equilibrium con-
`stants, which govern the retention on nonpolar bonded
`stationary phases in the presence of various concentrations
`of a complexing agent (hetaeron) in the mobile phase, the
`process first will be treated phenomenologically.
`Figure 1 illustrates the various equilibria which are involved
`in the chromatographic process. The eluite, E, whose retention
`is of interest, can interact with the hetaeron, H, to form a
`complex, EH, which is bound to the hydrocarbonaceous ligand,
`L, of the stationary phase to form LEH. Alternatively, the
`hetaeron maybindfirst to the stationary phase and then form
`LEH.In addition the species E and H can individually form
`the adducts LE and LH with the ligands of the stationary
`phase. It is assumed that binding of the eluite and the heteron
`2030
`CUREVAC EX2030
`ANALYTICAL CHEMISTRY, VOL. 49, NO. 14, DECEMBER areSees
`a
`Page 1
`ge 1
`
`

`

`Ko
`
`E+ H
`
`EH
`
`E+H
`
`Ko
`
`+L
`
`Ky + Ky
`
`+L
`
`K3
`
`Ke
`—_—_—
`
`——
`
`LE +H
`
`Ks
`
`LEH
`
`E + LH
`
`Figure 1. Schematic illustration of the equilibria Involved in the
`chromatographic process with nonpolar bonded stationary phases and
`a complexing agent in the mobile phase. The meaning of the symbols
`is: E, eluite; 4, complexing agent (hetaeron); L, hydrocarbonaceous
`ligand bound to the support; K, to K, are the corresponding equilibrium
`constants
`
`to the stationary phase ligands takes place independently.
`The equilibrium constants are expressed by the following
`set of equations in which the species concentrations in the
`mobile and stationary phase are denoted by the subscripts
`m and s, respectively. The concentration of L is defined as
`the accessible ligand concentration in the stationary phase.
`
`K, = [LE],/{E]m{L],
`K,= [EH] m /[E]mlH]m
`K3= (LH), /(7)m(L],
`K,= [LEH],/(EX],,[L],
`Ks = (LEH],/(LH],[E]n
`Ke = (LEX],/(LE)[A]
`
`(1)
`(2)
`(3)
`(4)
`(5)
`(6)
`
`The capacity factor of the eluite, k, is defined in the usual
`wayas
`
`k= 9({LEH], + [LE],)/({E]m + [EH]m)
`
`(7)
`
`where ¢ is the phase ratio, i.e. the ratio of the volume of
`stationary phase to the volume of the mobile phase in the
`column.
`In chromatographic practice when the column is equili-
`brated with the eluent the hetaeron concentration in the
`mobile phase is constant. If [E] << [H],,, only a negligible
`fraction of the hetaeron is in the form of a complex so that
`the hetaeron concentration can be considered invariant and
`we can write
`
`[Hm = [4]
`
`(8)
`
`The assumption that complex formation occurs with the
`hetaeron already bound to the stationary phaseyields, by the
`combination of Equations 1-3, 5 and 7-9 the following ex-
`pression for the capacity factor.
`
`k= ¢(L)(K, + K3K;[H])/(1 + K,[H])(1 + K3[4))
`(11)
`
`A third possible combination uses Equations 1-3 and 6-9 and
`yields
`
`k= @[L)(K, + K,Ko(H])/(1 + K,[H])(1 + K3[H])
`(12)
`
`Equation 12 would imply that the eluite first binds to the
`stationary phase and then forms a complex with the hetaeron.
`Equations 10-12 all express the dependenceofthe capacity
`factor on the hetaeron concentration and have the general
`form
`
`k= (ko + B[H))/(1 + K2[H])(1 + K3[H])
`
`(13)
`
`where kpis the capacity factor of the eluite in the absenceof
`hetaeron, Ky is the association constant for the eluite and the
`hetaeron, K, is the binding constant of the hetaeron to the
`stationary phase and B is the product of the two equilibrium
`constants as shown in Equations 10-12. A plot of & vs. [H]
`according to Equation 13 yields a parabola provided 1/K,[H]
`<1/Ko.
`If either K.[H] << 1 or K,[H] << 1, Equation 13 can be
`written as
`
`k = (ko + B[A])/(1 + P[A))
`
`(14)
`
`where P can be either K, or Ky. Equation 14 is the equation
`of a rectangular hyperbola, for the dependenceof the capacity
`factor on the hetaeron concentration. Both parabolic and
`hyperbolic dependenceof the capacity factor on the hetaeron
`concentration have been observed in ion-pair reversed phase
`chromatography(5, 6).
`When the complex is an ion-pair, both the eluite and the
`hetaeron haveto be fully ionized for the above treatment to
`be valid. Whereas the hetaeronis usually a strong electrolyte
`in ion-pair chromatography, the eluites are often weak bases
`or acids and therefore the pH of the eluent can have an
`influence on the retention. The corresponding protonic
`equilibria can readily be incorporated into the above model
`as will be shown by the example of a weakly basic eluite.
`The protonation of the neutral eluite, E°, is characterized
`by its acid dissociation constant, K,, related to the equilibrium
`
`E° + H* 2 EH*
`
`(15)
`
`Both forms, E° and EH*, can bind to the ligands of the
`stationary phase according to the following equilibria
`
`Since the extent of binding of the eluite by the stationary
`phase is expected to be small and the total ligand concen-
`tration [L]; is conserved, we may write that
`
`Eo + L.2@ LE®
`
`and
`
`[L]r = [L], + (LA), = [L]
`
`(9)
`
`EH* + L2 LEH*
`
`(16)
`
`(17)
`
`There are several ways to evaluate from Equations 1-6 a
`combination of the equilibrium constants which govern the
`chromatographic process.
`If we assume that the eluite is
`bound by the stationary phase as its complex with the he-
`taeron, which is formed in the mobile phase, then the
`combination of Equations 1-4 and 7-9 yields the following
`expression for the capacity factor
`
`k= $(L)(K, + K,K,[H])/(1 + K,[H])(1 + K3[4])
`(10)
`
`2296 e ANALYTICAL CHEMISTRY, VOL. 49, NO, 14, DECEMBER 1977
`
`The equilibrium constants corresponding to Equations 16 and
`17 are denoted by K,° and K,, respectively.
`In this case, however, only the protonated eluite molecules
`can form an ion-pair [HEH] with the hetaeron-counterion and
`bind as a complex, [LHEH], to the stationary phase. Con-
`sequently, mass balanceyields the following expression for
`the capacity factor
`
`
`[LHEA], + [LE], + [LEH*],
`se
`° El, + [EH], + [HEA]
`CUREVAC EX2030
`CUREVAC EX2030
`Page 2
`Page 2
`
`18
`
`no
`
`

`

`Following the previous approach and using Equations 1-4
`and 15-17 to substitute the equilibrium constants into
`Equation 18, we obtain for the capacity factor the expression
`
`KiK,
`(K, + [H*] + K,K,[H])
`(1+ A, + &,[H])(1 + K3[4])
`[H*]
`
`k=@[L]
`
`(19)
`
`Equation 19 can also be written in a form similar to that of
`Equation 13, but in this case the magnitude of the parameters
`would also be dependenton theacid dissociation constant of
`the eluite and the hydrogen ion concentration in the mobile
`phase.
`Enhancement Factor. Experimental data show that in
`ion-pair reversed phase chromatography the dependence of
`the capacity factor on the detergent concentration often
`follows hyperbolic behavior (6) such as represented by
`Equation 14. According to this expression the two limiting
`values of the capacity factor are ky and B/P at zero and at
`sufficiently high hetaeron concentrations, respectively. The
`ratio of the two quantities gives the highest possible am-
`plification of the capacity factor due to the presence of the
`hetaeron. It is termed the enhancement factor, n, and given
`by
`n = B/RoP
`
`(20)
`
`where ky is the capacity factor of an eluite in the absence of
`hetaeron, P is either the stability constant of the eluite—
`hetaeron complex or the equilibrium constant for the binding
`of the hetaeron to the stationary phase and the physical
`meaning of B also depends on the particular mechanism which
`governseluite retention in the presence of the hetaeron. We
`shall see later how 7 can be used for both the elucidation of
`mechanism and the practical selection of a hetaeron.
`Mechanistic Implications of the Solvophobic Theory.
`In view of the preceding section, on a closer examination of
`the process, the retention of the eluite in ion-pair chroma-
`tography on nonpolar bonded phases can occur either by
`“dynamic ion-exchange’, i.e., ion-pair formation takes place
`between the eluite and the hetaeron bound tothe stationary
`phase, or by ion-pair formation in the mobile phase and
`binding of the complex to the nonpolar stationary phase.
`Since a phenomenological approach cannot distinguish be-
`tween the two cases on the basis of the dependence of the
`capacity factor on the hetaeron concentration, we shall use
`the solvophobic theory to estimate the relative magnitude of
`the equilibrium constants on the basis of the molecular
`properties of the hetaeron and eluites.
`The solvophobic theory was developed to describe the effect
`of solvent on chemical phenomena (9) and has successfully
`accountedfor inter alia, the solubility of small nonelectrolytes
`in water and other solvents (72) and the effect of solvent
`variation on reaction rates for several different chemical
`reactions (13). It is not restricted to water as the solvent and
`expresses the energetics of the solvent effect in terms of, at
`least in principle, measurable properties of the solute and
`solvent, unlike other theories for the hydrophobic effect.
`We have recently adapted this approach to quantitatively
`treat the effect of eluite and eluent properties on chroma-
`tographic retention using polar solvents, especially water, and
`a nonpolar stationary phase (2, 10). In our model, we assumed
`that the chromatographic process entails a reversible asso-
`ciation of a solute, S, with the hydrocarbonaceousligand, L,
`of the stationary phase to form a complex SL. The logarithm
`of the corresponding equilibrium constant for a nonionized
`solute, K,°, was expressed for fixed column and eluent
`properties at a given temperature by Equation 47 in Ref. (10)
`
`which for our purpose can be written in a simplified form as
`
`InK,;°+~a-b+cAA
`
`(21)
`
`wherea, b, and c can be regarded as constants dependent upon
`solvent and column properties; b also depends on the eluite
`properties such as dipole moment,polarizability and molecular
`volume.
`AA is the difference between the molecular surface
`area of the complex, Ags,, and those of the eluite, As, and the
`hydrocarbonaceous ligand, A,, so that
`
`AA =Ag, —- Ag - AL
`
`(22)
`
`For the capacity factor of an ionized solute, the following
`simplified expression can be derived from Equation 20g of Ref.
`(2)
`(23)
`In K,=a'+b' f(Z)+ c'AA
`where a’, b’, and c’ are again solvent and column dependent
`parameters. Theeffect of charge on the solute moleculeis
`represented by the function /(Z) which goes approximately
`as the absolute value of the product of charges on the ion and
`its counterion.
`The simplified expressions in Equations 21 and 23 allow
`us to make some qualitative and semiquantitative statements
`regarding the constants lumped together in the enhancement
`factor as far as the mechanism of ion-pair reversed-phase
`chromatography is concerned.
`It is recalled that the pa-
`rameters Xo, B, and P in Equation 20 are directly related to
`the equilibrium constants defined in Equations 1-5.
`As ky is the capacity factor of the eluite in the absence of
`hetaeron, in view of Equations 1, 10, and 14 we may write that
`
`In Ro = In(@[L]K;) = a’ + b’ f(Z) + cAA
`
`(24)
`
`Obviously the value of ky is independent of any kind of
`ion-pair formation. On the other hand, the meaning of the
`constants B and P in Equation 20 is dependent on the actual
`mechanism of the chromatographic process.
`The energy of any electrostatic interaction and hence the
`logarithm of the corresponding equilibrium constant depends
`upon the product of the charges on the interacting species.
`Thus, for the ion-pair formation in the mobile phase the
`equilibrium constant, K>, in Equation 2 can be expressed by
`
`In K,= f'(ZpZy) + const.
`
`(25)
`
`where Zy and Zy are the charges on the eluite and hetaeron,
`respectively, Similarly, in the case of dynamic ion-exchange,
`represented by Equation 4, the equilibrium constant, K;, for
`the interaction between the eluite and the hetaeron bound
`to the stationary phase can be expressed by
`
`In K, =f" (Zp Zz) + const.
`
`(26)
`
`The other equilibrium constants of interest, K; and K.,
`correspond to the binding of the hetaeron and the complex
`to the stationary phase ligands, as shown by Equations 3 and
`4, respectively. According to the solvophobic theory, both
`equilibrium constants can be expressed as a function of the
`decrease in the molecular surface area upon binding of the
`species to the stationary phase. Thus, by using Equation 23,
`we can write for the equilibrium constant representing the
`binding of the charged hetaeron, that
`
`In K3=a' + b' f(Zy) + c'AA;
`
`(27)
`
`On the other hand, the equilibrium constant for the binding
`of the neutral ion-pair can be expressed from Equation 21 as
`In Ky4= a-b+ cAA,
`(28)
`2030
`CUREVAC EX2030
`ANALYTICAL CHEMISTRY, VOL. 49, NO. 14, DECEMBER SOREY4S:
`Page 3
`age 3
`
`

`

`Table I. Relationship between Hetaeron Properties and
`the Parameters of Equation 14 as Predicted for the Two
`Limiting Mechanismsin Ion-Pair
`Reversed-Phase Chromatography
`
`Parameter
`k
`B
`
`P
`
`Ion-pair formation occurs in the
`
`mobile phase
`
`issita
`hydrophobic surface
`area
`(carbon number)
`charge type
`(P= K,)
`
`stationary phase
`gets
`hydrophobic surface
`area
`(carbon number)
`hydrophobic surface
`(carbon number)
`area
`
`+ charge type
`(P= K;)
`Bik ,P
`hydrophobie surface
`charge type
`area
`(carbon number)
`
`According to the previous discussion the meanings of AA;
`and AA, in Equations 27 and 28 are different and given by
`
`OA3= Apy-~ Ay - AL
`
`and
`
`AA, = Ang, ~- AL ~ Anz
`
`(29)
`
`(30)
`
`where A; is the molecular surface area of the species i denoted
`by the subscripts.
`Wecan express the enhancementfactor, 7, by two different
`combinations of the equilibrium constants. In thefirst case,
`assuming dynamic ion exchange we obtain from Equations
`11, 14, 20, 24, and 26 that
`
`In n = In (K,;/K,) = const. +
`f"'(ZeZy) — f(Ze) — C(Aren — Az, ~— Aux) (81)
`
`In the second case, when ion-pair formation in the mobile
`phase dominates, the enhancement factor can be expressed
`by using a similar combination of the pertinent equations as
`
`In n = In (K4/K,) = const. -
`f(Zz) + c(Ange, — Aue + Ag ~ Axe)
`
`From the results presented earlier (2) we know that
`
`Ag, ~Ag-A, «Ag
`
`(382)
`
`(33)
`
`Since this relationship is expected to hold forall species under
`investigation we can write for the last term in Equation 32
`that
`
`Aye, — Aye + Ag ~ Ate & (Anz — Az)
`
`(34)
`
`In other words the last term in Equation 32 depends upon
`the difference in the surface area of the complex and eluite.
`If the complex is formed to maximize the electrostatic effect,
`this difference is very nearly the surface area of the hetaeron,
`An, alone. Hence, if the retention proceeds primarily by the
`formation of ion-pairs in the mobile phase, the enhancement
`factor will depend upon the surface area of the hetaeron as
`
`logn «Ay
`
`(35)
`
`Consequently, in the case of normal alkyl sulfates or sulfonates
`log 7 is expected to be proportional to the carbon numberof
`the hetaeron.
`On the other hand, no such a dependence is expected in
`the “ion-exchange” mechanism. Indeed, the relevant hetaeron
`property, as seen from Equation 30,is a function of its charge,
`f(ZyZ_), only. In view of these relationships, the analysis of
`
`2298 e ANALYTICAL CHEMISTRY, VOL. 49, NO. 14, DECEMBER 1977
`
`
`
`Table II. Relationship between Eluite Properties and the
`Parameters of Equation 14 as Predicted for the Two
`Limiting Mechanisms with Ion-Pair Formation
`
`Parameter
`
`mobile phase
`
`stationary phase
`
`Ion-pair formation occurs in
`
`Rk,
`
`B
`
`charge and
`hydrophobic
`surface area
`charge and
`hydrophobic
`surface area
`ve
`charge and
`hydrophobic
`surface area
`
`charge and
`hydrophobic
`surface area
`charge and
`hydrophobic
`surface area
`charge (P= K,)
`charge
`
`P
`Bik,P
`
`experimental data obtained with hetaerons containing the
`sameionic groups butdifferent alkyl chain lengths can shed
`light on the actual mechanism of the process. The dependence
`of the enhancementfactor on the properties of the hetaeron
`andeluite is shown in Tables I and II, which summarize the
`conclusions of this approach.
`EXPERIMENTAL
`
`A Model 601 (Perkin-Elmer, Norwalk, Conn.) high pressure
`liquid chromatograph was used with a Model 7010 sampling valve
`(Rheodyne, Berkeley, Calif.), a Model FS 770 (Schoeffel,
`Westwood, N.J.) variable wavelength detector at 254 nm, and a
`Perkin-Elmer Model R-56 recorder were used. Partisil 1025 ODS
`(Whatman, Clifton, N.J.) columns packed with 10-um octadec-
`yl-silica containing about 5% (w/w) carbon were used in the study
`of hetaeron behavior. The chromatogram in Figure 2 was obtained
`with a LiChrosorb RP-18 column (Rainin, Boston, Mass.) packed
`with 5-um octadecyl-silica. All columns were 250 mm long and
`had 6.4 mm o.d. and 4.6 mm i.d. Most experiments were carried
`out by isocratic elution using neat aqueous 5 X 10 M phosphate
`buffer, pH 2.5, and the hetaerons werealso dissolved in this buffer.
`In most cases the flow rate and the column temperature were 2
`mL/min and 40 °C, respectively. Some experiments were carried
`out with hetaerons dissolved in mixtures of methanol and the
`above mentioned phosphate buffer as the eluent.
`Catecholamine derivatives were obtained from Aldrich
`(Milwaukee, Wis.) or Schwartz/ Mann (Orangeburg, N.Y.) and
`reagent grade H,;PO, and KHPO, were supplied by Fisher
`(Pittsburgh, Pa.). The alkyl sulfates and hexylsulfonate used were
`Eastman products (Rochester, N.Y.), whereas the alkyl phosphates
`were gifts from Hooker Chemicals Corp. (Buffalo, N.Y.). The
`perfluorated carboxylic acids were purchased from Aldrich.
`Methanol was “distilled in glass” from Burdick and Jackson
`(Muskegon, Mich.).
`Retention times were measured from the distance between the
`injection point and the peak maxima on the chromatogram. The
`mobile phase hold-up times were measured as described previously
`(2) and the capacity factors have been calculated in the usual way
`(14).
`The analysis of the data was performed on a PDP 11/10
`minicomputer equipped with a RX0Q1 floppy disc, a VT55 display
`unit, and a Deewriter. The computer program used for parameter
`estimation by the least squares method was written in BASIC
`language.
`The symbols used in this study for the sample components are
`as follows: DOPA, 3,4-dihydroxyphenylethylamine (dopamine);
`EP, 1-(3,4-dihydroxyphenyl)-2-(methylamino)ethanol
`(epi-
`nephrine, adrenaline); OP, 1-(4-hydroxyphenyl)-2-aminoethane
`(octopamine); NE, 2-amino-1-(3,4-dihydroxyphenyl) ethanol
`(norepinephrine, noradrenaline); DOS, 3,4-dihydroxyphenylserine.
`RESULTS AND DISCUSSION
`
`A typical chromatogram in Figure 2 illustrates the sepa-
`ration of certain catecholamines on octadecyl-silica in the
`absence and in the presence of n-octylsulfate in the neat
`aqueous phosphate buffer used as the eluent. As the chemical
`nature and concentration of the alkyl sulfate have great
`CUREVAC EX2030
`CUREVAC EX2030
`Page 4
`Page 4
`
`

`

`
`
`°
`
`20
`40
`HEXYLSULFATE [mM]
`
`60
`
`
`
`
`
`
`Figure 4, Dependence of the capacity factor of charged catecholamine
`derivatives on the concentration of n-hexylsulfate in the neat aqueous
`mobile phase. Conditions are given in Figure 3
`
`o
`
`20
`40
`60
`OCTYLSULFATE [m M|
`Figure 5. Dependenceof the capacity factor of charged catecholamine
`derivatives on the concentration of n-octylsulfate in the neat aqueous
`mobile phase. Conditions are given in Figure 3
`
`NMET
`
`NE
`
`DA
`
`CT
`
`E
`
`t
`
`06
`
`DA
`
`ME
`
`.O3-OCT
`E
`
`NE
`
`E
`S
`
`< v
`
`<
`eoa
`°
`%
`
`0123
`
`(arene
`oO
`4
`8
`12
`16
`MINUTES
`
`Figure 2, Chromatogramsillustrating the effect of ion-pair formation
`with n-octylsulfate in the eluent on the separation of catecholamines
`by reversed-phase chromatography. Column, 5 um LiChrosorb RP 18;
`flow rate, 2.0 mL/min; temp., 70 °C;inlet pressure, 2200 psi. Eluents:
`A, 5 X 107 M phosphate in water, pH 2.2; B, 5 X 10°? M phosphate
`and 3 X 107° M octylsulfate in water, pH 2.2
`
`100
`
`40
`60
`80
`BUTYLSULFATE [mm]
`Figure 3. Dependence of the capacity factor of protonated catechol
`amine derivatives on the concentration of n-butylsulfate in the eluent.
`Column, 10 um Partisil ODS; flow rate, 2.0 mL/min; temp., 40 °C; inlet
`pressure, 400 psi; eluent, 5 X 10° M phosphate in water, pH 2.55,
`containing various concentrations of the hetaeron
`
`
`
`°
`
`20
`
`influence on the retention of the eluites, experiments were
`carried out in a wide range of conditions in order to shed light
`on the chromatographic process in view of the preceding
`theoretical anaysis. In addition to alkylsulfates of different
`chain lengths, hexylsulfonate as well as butyl- and amyl-
`phosphates were also employed.
`Equation 13 predicts that the observed capacity factor will
`initially increase with increasing hetaeron concentration
`followed by a monotonic decrease at high hetaeron concen-
`trations. However, if over the experimentally accessible range
`of hetaeron concentrations, either the binding of hetaeron to
`
`the stationary phase, K,[H], or the extent of ion-pair formation
`in the mobile phase, K;[H], is negligible, the capacity factor
`can be expressed by Equation 14. In this case the capacity
`factor first rises and eventually becomes practically inde-
`pendentof the hetaeron concentration. Therefore, if Equation
`14 holds over the experimental range of hetaeron concen-
`tration, a plot of & vs. [H] yields a rectangular hyperbola.
`The capacity factor of four catecholamines and the amino
`acid, 3,4-dihydroxyphenylserine, as a function of the con-
`centration of various n-alkyl sulfates in the eluent is shown
`2030
`CUREVAC EX2030
`ANALYTICAL CHEMISTRY, VOL. 49, NO. 14, DECEMBER SHREY.aS
`Page 5
`age 5
`
`

`

`
`
`Sr
`
`DECYL
`
`
`
`CAPACITYFACTOR
`
`20
`40
`ALKYLSULFATE [mM]
`Figure 7. Plots of the capacity factor of adrenaline vs. the hetaeron
`concentration for various n-alkylsulfates. Conditions are given in Figure
`3
`
`60
`
` °
`
`1.0 0.8
`
`
`© DECYLSULFATE
`
`% HEXYLSULFATE
`
`@ BUTYLSULFATE
`
`=|
`
`O- 0.6
`°
`|lx,
`
`0.2
`
`P [#1]
`
`Figure 8. Normalized plot of the capacity factor data obtained for
`dopamine using three different n-alkyl sulfates as the hetaerons. The
`theoretical curve calculated from Equation 41 is given by the solid line.
`The data were obtained under conditions described in Figure 3
`
`hexylsulfonate, butyl- and amylphosphates in neat aqueous
`eluents, the relationship between the capacity factors mea-
`sured with these eluites and the hetaeron concentration was
`found to conform well to Equation 13.
`We noted in the theoretical section that the dependence
`of the capacity factor on the hetaeron concentration alone does
`not shed light on the actual mechanism of the process.
`However, as one changes from one hetaeron to another, both
`representing the sametype of compounds, predictions of the
`effect on the capacity factor can be made byrecourse to
`solvophobic theory as shown in Tables I and II. If ion-pair
`formation occurs in the mobile phase, the enhancementfactor
`will depend strongly on the hydrophobic area of the hetaeron
`CUREVAC EX2030
`CUREVAC EX2030
`Page 6
`Page 6
`
`
`
`
`
`
`
`DECYLSULFATE [mm]
`Figure 6. Dependence of the capacity factor of charged catecholamine
`derivatives on the concentration of n-decylsulfate in the neat aqueous
`mobile phase. Conditions are given in Figure 3
`
`in Figures 3-6. The alkyl chain of the hetaerons ranges from
`butyl to decyl groups and the upperlimit of the concentration
`range was usually determined by thesolubility of the hetaeron
`in the neat aqueouseluent.
`Inspection of the data shows that with butyl-, hexyl-, and
`decylsulfates the capacity factor in most cases rises with
`increasing hetaeron concentration to a constant value from
`which it does not decline significantly. On the other hand
`when decylsulfate is used, the capacity factor increases to a
`maximum from which it rapidly decreases with further in-
`crease in the hetaeron concentration. Thus, the qualitative
`predictions are supported by the data as is also illustrated by
`the dependence of the capacity factor of adrenaline on the
`concentration of the hetaerons in Figure 7.
`In order to test the validity of Equations 13 and 14, the data
`shown in Figures 3-6 were analyzed bya least-squarefit. The
`data obtained using butyl-, hexyl-, and octylsulfates did not
`converge by using Equation 13 but they did fit Equation 14.
`On the other hand, the data obtained using decylsulfate could
`be well fitted to Equation 13, whereas thefit to Equation 14,
`was very poor.
`The qualityof the fit can be illustrated by a generalized
`plot of the data. Equation 14 is divided by kp, the expression
`can be rearranged to obtain
`
`(k/Ron) — (1/n(1 + PLA])) = PLA]/(1 + PLA])
`
`(86)
`
`The RHS of Equation 36 represents a normalized capacity
`factor corrected for the effect of unconjugated eluite binding.
`Whenit is plotted against the normalized hetaeron con-
`centration, P[H], a rectangular hyperbola should be obtained.
`Sucha plot of data obtained with butylsulfate, hexylsulfate,
`and decylsulfate as hetaerons and DOPAas theeluite is shown
`in Figure 8.
`In the case of other acidic hetaerons such as
`
`2300 e¢ ANALYTICAL CHEMISTRY, VOL. 49, NO. 14, DECEMBER 1977
`
`

`

`100Fr
`
`50
`
`T
`
`nh o
`
`ENHANCEMENT
`
`FACTOR,7 ° T
`
`4
`
`6
`
`8
`
`10
`
`
`
`CARBON NUMBER OF THE HETAERON,N,
`
`Figure 9. The dependence of the enhancement factor of catecholamine
`derivatives on the carbon numberof n-alky! sulfates as the hetaerons.
`The straight line obtained by least squares analysis fits the expression,
`log 7 = 0.225 (40.0317) N., where 7 is the enhancementfactor and
`N, is the carbon number. Theintercept is zero within experimental
`error
`
`and the charge of eluite butit will be independentofthesize
`of the eluite. On the other hand, if ion-pair formation occurs
`on the stationary phase,i-e., in the case of the ion-exchange
`mechanism, the enhancementfactor will depend on the charge
`of the hetaeron and the charge and size of the eluite.
`The predictions based on the solvophobic treatment of
`chromatographic retention support the concept that in our
`experiments soap chromatography proceeds through the
`formation of ion-pairs in the mobile phase followed by ad-
`sorption onto the nonpolarstationary phase. The logarithm
`of the enhancementfactor evaluated with the least-squares
`parameters of Equation 14 is linear in the carbon numberof
`the hetaeron and largely independentofthe size of the eluite
`as illustrated in Figure 9. This is predicted for the ion-pairing
`mechanism according to Equation 35, whereas no dependence
`on the hetaeron size but a dependence on the molecular area
`of the eluite is expected for the ion-exchange mechanism. The
`constant P of Equation 14 should be equal to K, and inde-
`pendentof hetaeron size in the ion-pair mechanism butwill
`depend uponthe charge of both eluite and hetaeron. It will
`also be dependent upon the charge type since the distance
`of closest approach of the twoionsis implicitly included in
`the function f(ZpZy) in Equation 25 and 26.
`The mean values of the pertinent parameters as obtained
`by least squares analysis are shown in Table III. The standard
`deviation of the values is also indicated. The capacity factor
`of the eluite in the absence of hetaeron, Xo, is not shown
`becauseit is not a property of the hetaeron. It is seen from
`the data in Table III that the enhancementfactorsare fairly
`independent of the eluite, but they strongly depend and, in
`fact, increase exponentially with the carbon numberof the
`hetaeron in agreement with the data shownin Figure 9. The
`only exception to this rule is 3,4-dihydroxyphenylserine which
`was not entirely in the cationic form at the eluent pH em-
`ployed. However, we can estimate the enhancement factor
`of the cationic form by usi