141
Anal. Chem. 1003, 65, 141-147
Numerical Analysis of Elution Behaviors of Substituted Benzoate Anions in Ion Chromatography Naoki Hirayama’ and Tooru Kuwamoto Department of Chemistry, Faculty of Science, Kyoto University, Sakyo-ku, Kyoto 606-01, Japan
The Influence of sample Ion structures on the elutlons of the sample Ions In ion chromatography was studied for effective separation. I n this study, the differences in the retentlon tlmes of varied substltuted benzoate anions were analyzed numerically by dlvlding the substltuent effect Into three terms: an LFER-applkable effect tern, a sterlc effect term, and a ~ l o n a l e f f e ccorrectlon t tern. The LFER-appllcableeffect term was Introducedby putting the theory of h e a r free energy relatlonshlps(LFERs) Into the reiatlonshlpbetweenthe anlonexchange equlilbrlum constants (K.J and the dissoclatlon constants (K.) of the sample anlons. The sterlc effect term was Introduced by conslderlng the Influence of the hydrated slzes of the anions, which depend on the kind and number of the Introduced substituents. It is expressed by using newly deflned lnvarlabie rterlc effect lndkes. The posltlonal effect correction term was Introduced to correct the speclflc effect caused by the posltlon of the Substituents and expressed by udng newly defined podllonal effect correctlon Indlces. Furthermore, the numerical calculatlon of the retentlon times of these anions was done.
INTRODUCTION Ion chromatography (IC)’ is a very effective analytical method for the separation and simultaneous determination of many ions in an aqueous solution. It is widely used in the fields of environmental, clinical and food analyses, etc.2-4 The numerical analysis of the retention times of sample ions is very important in order to fix the optimum chromatographic condition. Several workers”31 have reported on the model and the method for the numerical analysis and prediction of the retention times. (1)Small, H.; Stevens, T. S.; Baumann, W. C. Anal. Chem. 1975,47, 1801-1809. (2)Gjerde, D. T.;Fritz, J. S. Ion Chromatography;2nd ed.; Huthig: Heidelberg, 1987. (3)Tarter, J. G.,Ed. Ion Chromatography;MarcelDecker: New York, Basel, 1987. (4)Haddad, P. R.; Jackson, P. E. Ion Chromatography; Elsevier: Amsterdom, 1990. (5)Gjerde, D. T.; Fritz, J. S. J. Chromatogr. 1979,176,199-206. (6)Van Os. M. J.: Slanina. J.; De Limy, C. L.; Hammers, W. E.; Agterdenbos, J. Anai. Chim. Acta 1982,i44; 73-82; (7)Haddad, P. R.; Cowie, C. E. J. Chromatogr. 1984,303,321-330. (8)Diop, A.; Jardy, A,; Caude, M.; Rosset, R. Analusis 1986,14,67-73. (9)Jardy,A.;Caude,M.;Diop,A.;Curvale,C.;Rosset,R. J . Chromatogr. 1988,439,137-149. (10)Hoover, T. B. Sep. Sci. Technol. 1982,17,295-305. (11)Jenke, D. R.; Pagenkopf, G. K. J. Chromatogr. 1983,269,202207. (12)Jenke, D. R.; Pagenkopf, G. K. Anal. Chem. 1984,56,85-88. (13)Jenke, D. R.; Pagenkopf, G.K. Anal. Chem. 1984,56,88-91. (14)Jenke, D. R. Anal. Chem. 1984,56,2674-2681. (15)Baba, Y.; Yoza, N.; Ohashi, S. J. Chromatogr. 1985,348,27-37. (16)Baba, Y.;Yoza, N.; Ohashi, S. J.Chromatogr. 1985,350,119-125. (17)Baba, Y.; Fukuda, M.; Yoza, N. J. Chromatogr. 1988,458,385394. (18)Baba, Y.; Ito, M. K. J. Chromatogr. 1989,485,647-655. (19)Baba, Y.; Kura, G. J. Chromatogr. 1991,550,5-14. (20)Rounds, M. A.; Regnier, F. E. J. Chromatogr. 1984,283,37-45. (21)Rojas, J.; Ballesteros, L.; Valcarcel, M. Microchem. J. 1986,34, 92-102. 00O3-2700/93/0365-0141$04.00/0
In the numerical analysis, it is the most important to consider the factors regulating the retention times of the sample ions, such as the concentration of the eluent ion or the ion-exchange equilibrium constant (KeJ between the sample ion and the eluent i0n.32 On the other hand, it is well-known that a series of equilibrium constants in one reaction of the chemical species having similar structures can often be expressed as the function of that of another reaction of an analogous species by using the theory of linear free energy relationships ( L F E R s ) . ~ Therefore, ~,~~ it is possible to treat a series of K,, as the function of a series of the dissociation constants (KJ of the sample species. Furthermore, an ion pair is formed between the sample ion and the functional group in the ion exchanger. It was previously reported that K,, increases with the decrease in the hydrated size of the sample ion.35-3sFrom this fact, the hydrated (solvated) size of the sample ion seems to be a very important factor in the numerical analysis. In this paper, the authors investigated the method for the numerical calculation of the retention times by using substituted benzoate anions, to which the LFERs are easily applicable,33?34as samples. It was found that the substituent effect of the retention times of these anions can be divided into three terms: an “LFER-applicableeffect term”, a “steric effect term”, and a ‘positional effect correction term”. The first is expressed as the function of the pK,, the second as that of the kind and number of the substituents, and the last as that of the position of the substituents.
THEORY The authors applied Leffler and Grunwald’s theory of extrathermodynamic relationships33 to numerically analyze (22)Stout, R. W.; Sivakoff, S. I.; Ricker, R. D.; Snyder, L. R. J. Chromatogr. 1986,353,439-463. (23)Drager, R. R.; Regnier, F. E. J. Chrornatogr. 1986,359,147-155. (24)Parente, E. S.;Wetlaufer, D. B. J. Chromatogr. 1986,355,2f+40. (25)Rocklin, R. D.; Pohl, C. A,; Schibler, J. A. J. Chrornatogr. 1987, 411, 107-119. (26)Senyavin, M. M.; Venitsianov, E. V.; Dolgonoaov, A. M. Zh. Anal. Khim. 1987,42,82-88. (27)Hodges, R. S.;Robert Parker, J. M.; Mant, C. T.; Sharma, R. R. J . Chromatogr. 1988,458,147-167. (28)Sasagawa, T.; Sakamoto, Y.; Hirose, T.; Yoshida, T.; Kobayashi, Y.; Sato, Y.; Koizumi, K. J. Chromatogr. 1989,485,533-540. (29)Whiteley, R. D.; Berninger, J. A.; Rouhana, N.; Wang, N. H. L. Biotechnol. Prog. 1991,7,544-53. (30)Maruo, M.; Hirayama, N.; Kuwamoto, T. J. Chromatogr. 1989, 481,315-322. (31)Hirayama, N.; Kuwamoto, T. J. Chromatogr. 1990,508,51-60. (32)Gjerde, D.T.; Schmuckler, G.; Fritz, J. S. J. Chromatogr. 1980, 187,35-45. (33)Leffler, J. E.; Grunwald, E. Rates and Equilibria of Organic Reactions; John Wiley & Sons: New York and London, 1963. (34)Inamoto, N. Hammett Rule; Maruzen: Toyko, 1983. (35)Marcus, Y.; Kertes, A. S. Ion Exchange and Solvent Extraction of Metal Complexes; John Wiley & Sons: London, 1969. (36)Gregor, H. P. J. Am. Chem. SOC.1948,70, 1293. (37)Gregor, H. P. J. Am. Chem. SOC.1951,73,642-650. (38)Reichenberg, D. In Ion Exchange; Marinsky, J. A,, Ed.; Marcel Dekker: New York, 1966;Vol. 1. 0 1993 American Chemical Society
142
ANALYTICAL CHEMISTRY, VOL. 65, NO. 2, JANUARY 15, 1993
estimate the steric effect caused by the hydration anddecided only by the kind and the number of the substituents. I‘’R,C~F (called G the positional effect correction term) is used to correct the steric effect caused specifically by the position of the substituents. On the other hand, IR,COO-FG (called the LFER-applicable effect term) shows the inductive and resonance effects usually considered in LFERs. The effects are regarded as independent of the steric effect. By using eq 6, the change (AGOR,~.)of the free energy in the anion-exchangereaction is shown as follows:
the elution behavior of substituted benzoate anions in IC. In this theory, the aromatic species (RY) is divided into the zone (R)containing the substituent and that (Y) containing the reaction cite. Each of these zones is regarded as contributing an additive term to the free energy and also as interacting with the other zone. The standard free energy ( G O R Y ) of RY is expressed by using the independent additive terms (GR,G y ) and the term ( I R , ~having ) the dimension of energy and resulting from the interaction between R and Y as follows: GORY
+ G y + IR,y = GR + G y + IRIy = GR
ACfoR,er= (Go, (1)
+ QOE-FG) + G c O s F G - Gco, +
IR(ICOO-FG - Icoo-) + if^,^^^^ + iff^,^^^^^ (7) where subscript “ex” denotes the anion-exchange reaction. By defining
where IR and I y are obtained by the assumption that I R ,is~ factorable into two terms related to R and Y. On the dissociation of substituted benzoic acid (RCOOH) RCOOH H+ + RCOO(2) the change (AGOR,J of the free energy is shown by using eq 1 as follows:
the substituent effect ( ~ R A G O ~on~ ) the anion-exchange reaction is expressed from eqs 4 and 7 as follows: bRAGoex
GCOOH+ IRICOOH) QoH+
+ Gcoo- - GcooH + IR(Ic0o- - IC,,,)
= ACfoR,ex-
A”Ro,er
= (IR- IRO)(ICOO-FG - Iced + if^,^^^^^ +
(3)
I
’
f
~
where subscript “a” denotes the dissociation reaction. Therefore, the substituent effect ( ~ R A G O ~on) the dissociation reaction of RCOOH is shown as follows:
I”R,COO-FC (10) Since I’R,COO-FG is determined by the kind and number of the Substituents (X), the term can be expressed as I’x. The steric effect of each substituent is considered to be independent, as follows:
bRAGoa= AGoR,a- AGO%,,
= (IR - 1%)(IC,, - IC,,,)
(4)
where subscript “Wshows the unsubstituted species. On the other hand, in the following anion-exchange equilibrium in the anion-exchange column of IC
(11)
E-FG + RCOO- F= E-+ RCOO-FG (5) where E- denotes the eluent anion having a charge of -1, and FG denotes the functional group on the anion-exchanger,the standard free enegy (GORCO~FG) of the retained anion (RCOOFG) cannot be expressed by using eq 1because the equation has no term for the hydrated size of the anion which governs the stability of the ion pair.39~~0 For the accurate analysis of G’RCWFG, eq 1has to be corrected by addingterms to estimate the steric effect, peculiar to an ion pair, governed by the hydrated size. In the theory of extrathermodynamicrelationships,33when there is more than one mechanism of interaction between two zones of the species, the use of separate interaction terms corresponding to respective mechanisms is allowable. Accordingly,in this paper, the two interaction terms independent of IR,COO-FG (ZIRICOO-FG) were added to eq 1 as follows:
where ifxi denotes steric effect factor of Xj and nx, is the number of Xi. On the other hand, I”R,CWFGis determined by the position of the substituents. If the steric effect of a meta substituent is used as the standard, the substituent effects of the ortho substituent, para substituent, and neighboring substituents have to be corrected. In other words, this term is divided into an “ortho-effect correction term” ( P 0 . x ) ,a “para-effect correction term” (I”,.x) and a “neighboring-effect correction term” ( P x - xas ~ follows: ,
IffR,COO-FG= Iffo.x+ IflP-x+ If’x-y
(12)
The correction of each substituent(s) is considered to be independent, aa follows:
G’RCOO-FG= G R + GCOO-FG + IRCOO-FG+ I’RCOO-FG+ Iff~,c00-~~ = G R + GCOO-FG+ IRICOO-FG + if^,^^^^^ +
I ’ f ~(6), where I ’ R , C O ~ F(called G the steric effect term) is used to (39) Kay, R. L.; Evans, D. F.; Matesich, S. M. F. In Solute-Solvent Interactions; Coetzec, J. F., Ritchie, C. D., Eds.; Marcel Decker: New York, 1976; Vol2, Chapter 10. (40) Okazaki,S.;Sakamoto, I. Soloents and Ions;San-ei: Kyoto, 1990.
~
~
~
~
~
where i~!o~xl, i / f P x , ,and i f ~ x l - xdenote l, positional effect correction factors and n,.xj, nP-x,,and nx,-x; are the numbers of respective substituent(s). By substituting eqs 11-15 into
,
~
~
ANALYTICAL CHEMISTRY, VOL. 65, NO. 2, JANUARY 15, 1993
149
correction indices” ( j ” o . ~ , , j ” p ~ , , $ ~ are , - ~defined ;) as follows:
eq 10
+ +
6RACoex= ‘CoO-FG - ‘CoO-6RA~oa I f x ‘COO-
- ‘COOH
+ I f f p x+ (24)
- ‘COO-FC - Zcoc-~,ACO, + C n x j x , + ‘COO-
- ‘COOH
I
~ n o . x , i ” o . x+J ~ n p x , i ” p x , + znxJ-x;iffx,-x; I
where XI denotes the fixed substituent. (In this paper, C1 was selected as XI.) In this case, eq 23 is rewritten as follows:
I
(16)
I
From the thermodynamic relationships
AGoR,a= -RT, In Ka(R)
(17)
AGoR,ex= -RT2 In K,,(R)
(18)
where T I and T2 show the temperatures, and from the following ion-chromatographic relationship32 tR’(R) = k’(R)to (19) where tRf(R) is the retention time of sample RCOO- anion, k’(R) is the capacity factor of RCOO-, Cap is the column capacity, w is the weight of anion exchanger in column, uo is the column dead volume, and t o is the void time, the following equation is obtained for the fixed eluting condition: log tR’(R) - log tRf(RO)= log Kex(R)- log Ke,(Ro)
By defining the “LFER-applicable effect constant* (-pf), the “adjusted steric effect factors” (ifxl),and the “adjusted positional effect correction factors” (iffo-xJ, iffp-xj,and iffx,-x,3 as follows -pf
=T1 ‘COO-FG T2
‘COOH
- ‘coo- ‘COO-
EXPERIMENTAL SECTION Apparatus. A Tosoh Model HLC-601 ion chromatograph system made up of a computer-controlled pump, a sample injector (lOOpL),and an oven was used. For the retentions of substituted benzoate anions, two columns (50-mm X 4.6-mm i.d.) packed with different anion exchangers were used. One of them was a Tosoh Model TSKgel IC-Anion-PW(polymethacrylate gel base, particle size 10 pm, capacity 0.03 mequiv/g), and the other was a Tosoh Model TSKgel IC-Anion-SW (silica gel base, particle size 5 pm, capacity 0.4 mequiv/g). The flow rate was maintained at 1.0 mL/min under a pressure of 30-60 kg/cm2. The column was placed in the oven regulated at 30 O C . The peaks were detected by a Tosoh Model UV-8Model I1 ultraviolet detector at a wavelength of 254 nm and recorded by a Shimadzu Model Chromatopack C-R1A. Eluents. Each stock solution of 1M hydrochloric acid (HCl), 100 mM perchloric acid (HC104),1M sulfuric acid (H2S04),and 1M acetic acid (HOAc)was prepared by diluting the concentrated acid. Each stock solution of 1 M oxalic acid (HoOx) and 1 M tris(hydroxymethy1)aminomethane (Tris) was prepared by dissolving the reagent in distilled water. Each eluent was prepared by mixing the acid stock solution and the Tris stock solution and by diluting the mixed solution with distilled water after which the eluent was deaerated. Standard Sample Solutions. Table I shows the substituted benzoates used in this st~dy.414~ The stock solutions (50 mM) of substituted benzoates were prepared by dissolvingeach sodium salt in distilled water or by mixing each acid and equivalent sodium hydroxide and then diluting the mixed solution with distilled water. Working standard solutions (0.2 mM) were prepared by diluting the stock solutions with distilled water. Calculation of the Factors and the Indices. When the substituents of the sample anions are only one kind (X), eq 23 is simplified as follows: log tRf(R)= (1% tR’(RO) - (-Pf)PKa(Ro))+ (-P’)PK,(R) + nxifx + no.xiffo.x+ npxiffpx + nx-xi x-x .If
-
ifx, (or i’fo.xl,i f f p x ji’fx,,x,3 , (22) 2.303R T, 1
where the symbol of -pf having a minus sign corresponds to the Hammett pu relationship, and by substituting them into eq 20, the following equation is obtained
= A + (-p’)pK,(R)
+ nxifx+ no-xiffo.x+ npxiffpx +
nx-xi”x-x (26) where A is a constant determined by the experimental condition. The values of -pf and A were regressively calculated from the retention times of unsubstituted and Xs-substituted benzoate anions by using the following equation:
log tRf(R)= A + (-p’)pK,(R)
In practice, the values of ifx,, iffo.x,iffpxJ,and iffxl-x; are changed by changing the eluent condition, such as the kind or concentration of the eluent anion or the pH, which influences the ionic strength of the mobile phase or the solvation of the sample anions in the column. However, these values are proportional to each other. Therefore, the invariable “steric effect indices” c i ) ~ , and ) the ‘positional effect
+ nx,i’x,
(27)
where XSis the substituent in which all the values of iffo.xs,iffpxs, and i”xs-xsare negligible. (In this paper, C1 was selected as XS.) After that, i’x, iff0.x,iffP-x,and iffx-x on arbitrary X were regressively calculated by using eq 26 with substituting the calculated values of -pf and A. Finally, j’x,j”,,.x, Tpx, and Tx-x were determined by using eq 24. (41) Dean, J. A.,Ed. Lange’s Handbook of Chemistry, 13th ed.; McGraw-Hill: New York, 1985. (42) Buckingham, J., Ed. Dictionary of Organic Compounds, 5th ed.; Chapman and Hall: New York, 1982. (43) Davis, M. M.;Hetzer, H. B. J. Phys. Chem. 1967,61, 123-125.
144
ANALYTICAL CHEMISTRY, VOL. 65, NO. 2, JANUARY 15, 1993
Table I. Substituted Benzoic Acids Used in This Study and Their Retention Times ( t R f ) column IC-Anion-PW no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 a
substituent
name benzoic acid o-chlorobenzoic acid m-chlorobenzoic acid p-chlorobenzoic acid 2,4-dichlorobenzoic acid 2,5-dichlorobenzoic acid 2,6-dichlorobenzoic acid 3,4-dichlorobenzoic acid o-nitrobenzoic acid m-nitrobenzoic acid p-nitrobenzoic acid 2,4-dinitrobenzoic acid 3,4-dinitrobenzoic acid 3,5-dinitrobenzoic acid o-anisic acid m-anisic acid p-anisic acid 3,4-dimethoxybenzoic acid o-methylbenzoic acid m-methylbenzoic acid p-methylbenzoic acid 3,4-dimethylbenzoic acid anthranilic acid m-aminobenzoic acid p-aminobenzoic acid 3,5-diaminobenzoic acid o-iodobenzoic acid m-iodobenzoic acid p-iodobenzoic acid salicylic acid m-hydroxybenzoic acid p-hydroxybenzoic acid B-resorcylic acid y-resorcylic acid protocatechuic acid a-resorcylic acid gallic acid
pK, (25 0C)41 4.20 2.88 3.83 3.99 2.68O 2.47O 1.59O 3.64* 2.18 3.46 3.44 1.43 2.82 2.85 4.09 4.08 4.49 4.44 3.90 4.27 4.36 4.41 4.79 4.79 4.85 5.30 2.86 3.86 4.00 2.98 4.08 4.58 3.29 1.30 4.48 4.04 4.19
tR‘/minC 2.42 2.14 9.87 9.43 7.11 6.83 2.69 34.74 1.69 6.83 7.03 4.69 30.89 15.49 1.03 3.62 3.28 2.31 1.59 3.91 3.85 6.39 3.01 1.66 1.67 1.13 3.64 24.08 24.58 15.62 3.40 3.03 17.06 92.89 3.29 4.99 3.71
log tR‘ 0.384 0.330 0.994 0.975 0.852 0.834 0.430 1.541 0.228 0.834 0.847 0.671 1.490 1.190 0.013 0.559 0.516 0.364 0.201 0.592 0.585 0.806 0.479 0.220 0.223 0.053 0.561 1.382 1.391 1.194 0.531 0.481 1.232 1.968 0.517 0.698 0.569
IC-Anion-SW tR’lminc 3.80 3.03 7.18 6.36 4.59 4.54 3.05 12.39 2.63 3.92 3.06 2.05 4.82 2.59 2.24 5.07 5.07 4.57 2.39 5.26 5.54 8.19 4.19 3.33 3.31 3.15 3.84 13.08 11.83 9.74 4.65 4.48 13.54 23.52 NEd 6.11 NEd
log tR’ 0.580 0.481 0.856 0.803 0.662 0.657 0.484 1.093 0.420 0.593 0.486 0.312 0.683 0.413 0.350 0.705 0.705 0.660 0.378 0.721 0.144 0.913 0.622 0.522 0.520 0.498 0.584 1.117 1.073 0.989 0.667 0.651 1.132 1.371 0.831
*
Reference 42. Reference 43. Eluent: 40 m M HC1-Tris, pH 7.0. Not eluted.
RESULTS AND DISCUSSION Table I shows the measured retention times of the substituted benzoate anions in 40 mM HC1-Tris (pH 7.0) as eluent. From these data, it was found that the log ~R’(R) is not obtained as a simple function of the pKa(R) and that none of the Hammett pu, pu+, pa-, and p 8 relationships are applicable to log ~R’(R)on IC. Therefore, it was necessary to consider the steric effect. Selection of XS. The selection of XS is necessary for the linear regression calculation of eq 27. In this paper, C1 was selected as XS because the substituent is relatively small and does not form hydrogen bonding and because the pKa values of many C1-substituted benzoates are easily available. By using data nos. 1-8 in Table I, the regression calculation of eq 27 was performed. The results and correlation coefficients ( r ) were as follows: For IC-Anion-PW log tR’(R) = -1.961
+ 0.562pKa(R) + 0.710+,
(28)
r = 0.986 For IC-Anion-SW
r = 0.963 Figure 1shows the comparison of the regressively calculated
values with the measured values on the IC-Anion-PWcolumn. From the facts, it was found that the selection of the C1 substituent as Xs is valid for the regression calculation. Calculation of the Factors and the Indices. The values of i’x, i”o.x, i’lP.x, and 1“’~-x were regressively calculated by were using eq 26. After that, jlx, jll0.x, jllP-x, and TX-X calculated by using eq 24 and the value of i’cl (=i’x[). Table I1 shows the calculated values. Several adjusted positional effect correction factors (i”,,.~,i’b.~,~”x-x)and positional effect correction indices o’ll0.x,jllP-x,jllx-x) were neglected because these calculated values were very small. Constancy of t h e Indices. Table I11 shows the results of the linear regression calculations of the indices calculated from the data in several eluents vs those indicesin the standard eluent, 40 mM HC1-Tris (pH 7.0). The indiceswere invariable in arbitrary eluent conditions, as shown in the table. Indices in Various Corrections. (a) Steric Effect Indices VX). The steric effect term is determined by the kind and number of substituents. As shown in Table 11, the value offx on X = NH2 (hydrophilic substituent)was negative because the hydrated size of this substituent is large. On the contrary, that on X = I (hydrophobic substituent) was relatively large because the hydrated size of this substituent is small. From these results, it was found that the values of f x are mainly determined by the hydration of X. (b) Positional Effect Correction Indices WPx,j l P x , ~”x-x,). Although the steric effect term does not contain the effect caused by the position of the substituents, the hydrated size or the charge distribution of the anion is affected by the
ANALYTICAL CHEMISTRY, VOL. 65, NO. 2, JANUARY 15, 1993 145
Table 11. Calculated Values of Adjusted Steric Effect Factors i'x, Adjusted Positional Effect Correction Factors P-x, P)Px, i"x-xl, Steric Effect Indices j'x, and Positional Effect Correction Indices i"px, PPx,~"x-x, (Eluent: 40 m M HC1-Tris, pH 7.0) substituent (X) i'x i1lO.x illP X 1'"X-x i'X Yo-x Yp-X Yx-x Column: IC-Anion-PW (A = -1.961, -p' = 0.562)
c1
NOz OCH3
CH3 2"
I
OH
c1
N0z OCH3 CHa 2"
I OH
0.710 0.808 0.227 0.138 -0.497 1.139 0.188
00 0.183 -0.552 -0.168 0.245 -0.224 1.381
0.314 0.138 0.154 0.124 -0.227 0.590 0.140
Column: IC-Anion-SW(A = -0.693,-pl = 0.305) 00 0' 0' 1.000' 0.302 00 0.240 0.439 -0.358 -0.125 -0.184 0.490 -0.243 00 0" 0.395 0.081 00 b -0.723 -0.185 0" b 1.879 0.686 -0.187 b 0.446
00 00
-0.273 00 00 00
-0.384
0" 0.249 -0.351 0" b b 0"
00 00 -0.385
00
1.000c
1.138 0.320 0.194 -0.700 1.604 0.265
0.258 -0.777 -0.237 0.345 -0.315 1.945
00 0.351 -0.494 00
' 0
b b
00 00
00
-0.541
00 0.962 -1.140 -0.774 0.257 -0,589 2.185
00
00
00 -0.398
0.764 -0,586
00 00 00
00
b b b
-0.596
Neglected. * Cannot be calculated. Defined as unity. Table 111. Relationship between the j'x, P-x, Ppx,and Values in Several Eluents and Those in the Standard Eluent Using a Linear Regression Calculation (Standard Eluent: 40 m M C1- (PH 7.0))
2.0 -K CI
-g 1.5
YX (or Yo.x,Y P xYX-d , - B + CYx (or Yo.x,Y P xYx-y) , (standard) column
IC-Anion-PW 2.0
2.0-
2.5 3.0 3.5 4.0 0.562pKa +0.710ncl C
B
,
h
c)
b,
-0" 1.5-
, I
I
eluent' 10 mM C104100 mM OAc50 mM Ox260 mM sod220 mM C180 mM C140 mM C1- (pH 8.0)' 40 mM C1- (pH 9.0)'
B 0.031 -0.006 -0.009 -0.013 0.005 -0.005
C
r
0.973 1.007 1.011 1.019 0.987 1.006 -0.001 0.998 0.002 0.998
b
IC-Anion-SW B
C
0.998 0.027 1.000 0.009 1.000 0.006 0.999 O.OO0 1.OO0 0.016 LOO0 0.007 1.000 1.000
I 0.970 0.986 0.987 0.992 0.984 0.995
'
b 0.996 0.999 LOO0 0.999 0.999 LOO0
pH = 7.0 except where specifically mentioned. Correlation coefficient. Calculated without usingj'oH,Yo-0~, andYPoH values.
,
I
0.5
70/'
~
,,
I'
2
,/
I
0.0-
I
,,I I
-0.5
I
,
I
I
-
,
,,
I
1
2
3
4
PK*
5
Flguro 1. Comparison of the regressively calculated log h'(R)of CIsubstltuted benzoate anions wlth the measured values. Eluent,40 mM HCI-Trls, pH 7.0. Column, TSKgel IC-Anlon-PW. (A) Relationship between log tR'(R) and (0.562pKJR) 0.710n~~).(6) Relationship between log fR'(R) and pKa(R). Key: (open circles) measured data (sample, shown In Table I); (solM Ilne, A) the regresslvely calculated result (eq 28); (broken lines, 6 ) the result obtained by substituting = (a) 0, (b) 1, and (c) 2, respectively, into eq 28.
+
steric interaction specifically caused by the position. Therefore, the positional effect correction term has to be carried out as a function of the position of the substituents. (i) Correction of the Ortho Effect v e x ) . When substituent X is situated in the ortho position of benzoate, some specific effects between X and COO-, such as steric repulsion, steric inhibition of resonance, and intramolecular hydrogen bonding, are considered. These effects cause a change in the hydrated size or the charge distribution.
The value of f ' o - 0 ~was positive and very large because the intramolecular hydrogen bonding between OH and COOinterferes with the hydration of the anion and results in the small hydrated size of the anion, as shown in Table 11. The value of j " o - ~ was ~ 2positive but relatively small because the intramolecular hydrogen bonding between NH2 and COO- is weaker than the bonding between OH and COO-. When X is relatively small and does not result in hydrogen bonding with COO- or COOH, such as C1, the value of y0.x is very small and negligible. (ii) Correction of the Para Effect v P x ) . In general, the steric effect of the p-X substituent is similar to that of the m-X substituent. i"p-x, which is used for the correction of the steric effect of p-X obtained by defining that of m-X as standard, is very small and negligible. However, when X is an electron-releasing substituent, such as OCH3 and OH, the value of j)lP.xis not negligible, as shown in Table 11. The para-electron-releasing-substituted benzoic acid has a dipolar resonance structure in water33 and is stabilized. The dissociation of the acid is depressed by the stability. As shown in Table I, the pK, values41 of p-anisic acid and p-hydroxybenzoic acid are actually 0.4-0.5 higher than those of the meta-substituted acids. On the other hand, the structure of the ion pair between the para-substitutedbenzoate anion and the functional group
146
ANALYTICAL CHEMISTRY, VOL. 65, NO. 2, JANUARY 15, 1993
i-I
0.5
1 /
index on PW
1.o
-1
Flgure 2. Relationshlp between the indlces of the IC-Anion-SW COlUmn and those of the IC-Anion-PW column. The eluent is the same as in Figure 1. Data: ( 0 )jfx (A, jfNOp);(A)j”,x, j’b.x, i’i-x (6,l’’-o,; C, j”NO,-NOz; D, /’’&H3); (0) E, U’NO~+ j ’ 6 N o 2 ) and F, (2/1No2+ j”NO,-NO,). The line is regressively calculated result (except for data A-F, eq 30).
in the anion exchanger is not a dipolar resonance structure. Therefore, for the expression of K,, using K,, it is necessary to correct the effect related to the dipolar resonance structure by using jlrP-x. (iii) Correction of the Neighboring Effect WX-X.). When the two neighboring substituents, X and X’, are relatively large, steric repulsion often occurs between them. This phenomenon is similar to the case of an ortho-substituted benzoate. Therefore the values of .?”O2-NO2 and ~ ” O C H ~ - O C H ~ were not negligible, as shown in Table 11. Indices of Different Ion Exchangers. In the two anion exchangers,IC-Anion-PWand IC-Anion-SW,having the same functional group, the diethylmethylammonium cation, the elution behaviors of sample anions are estimated to be similar to each other. Figure 2 shows the relationship between the indices of ICAnion-SW and those of IC-Anion-PW. The line was regressively calculated, except for NO^, j ” o . ~ ~ ’2” ,o ~ - N oand ~, j ” 0 . ~ as ~ 3follows: , fX
(or (or
j”p-x, j”x-xl)(IC-Anion-SW) = 1.159fx jl‘p.x,j l f x - x J )(IC-Anion-PW) - 0.018 (30) P
= 0.991
And it was found that the indices of the two anion exchangers are generally similar to each other. However, the data for YNO~, j ” 0 . ~ and ~ 2 ,. j ’ ” 0 ~ - ~were 0 ~ not near the line. It seems that the distribution of *-electrons in the benzene ring is changed by the resonance between NO2 and the ring on the same plane. Accordingly, the electrostatic interaction between the anion exchanger and benzene ring is changed. However, with o-NO2 or two neighboring NOz, the change in this distribution of *-electrons is interfered with by the steric inhibition of resonance. The sums of the steric effect index and the positional effect correction indices, (j”02 + y0.~02) and ( 2 j l N 0 2 + j ’ ” 0 2 - ~ 0 2were ), near the line, as shown in the figure. The reason for the unfitness of j ” 0 . ~is~not 3 known. Numerical Calculation of Retention Times. From the above, it was found that the retention times of the sample (substituted benzoate) anions in any eluent condition can be numerically calculated by substituting the values of A, -p’, and i’c, (=i’x,) and the indices of jlx, y0.x,YP.x,and ~”x-x, into eq 25. Figure 3 and Table IV show the comparison between the calculated and the observed retention times of substituted
1.5
calculated
Flgure 3. Relationshipbetween the calculated and observed log h‘(R) of substituted benzoate anions. Condltions: eluent, 5 mM NaOH; pH 11.7; column, TSKgel IC-Anion-PW.
Table IV. Comparison of the Calculated and Observed Retention Times of the Substituted Benzoate Anions
(Eluent: no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
5 mM
NaOH, pH 11.7. Column: IC-Anion-PW)
substituent
calcd tR’imin log tR’
1.93 6.14 6.00 4.38 26.30 16.14 1.21 3.81 3.52 2.55 1.79 3.97 4.43 6.30 3.27 1.95 2.10 1.27 3.73 19.95 23.60
0.286 0.788 0.778 0.641 1.420 1.208 0.082 0.581 0.546 0.407 0.252 0.599 0.646 0.799 0.515 0.291 0.322 0.103 0.572 1.300 1.373
obsd
tR’/min
log tR’
2.65 2.33 9.26 8.91 6.71 6.40 2.78 30.73 1.89 6.47 6.65 4.47 26.82 13.47 1.30 3.76 3.43 2.63 1.83 3.98 3.95 6.19 3.16 1.97 1.94 1.50 3.69 21.75 22.32
0.423” 0.367” 0.967” 0.950” 0.827” 0.806” 0.444” 1.488” 0.277 0.811 0.823 0.650 1.428 1.129 0.114 0.575 0.535 0.420 0.262 0.600 0.597 0.792 0.500 0.294 0.288 0.176 0.567 1.337 1.349
” Used for the calculation of the values of A (-1.760), -p’ (0.523), and i’cl (0.649) by using eq 27. benzoate anions in 5 mM NaOH eluent (pH 11.7) on the IC-Anion-PW column. The correlation coefficients ( r )of log ~ R ’ ( R and ) ~ R ’ ( R were ) 0.997 and 0.994, respectively. It was found that this method is very effectivein the numerical calculation of the retention times of substituted benzoate anions.
CONCLUSION In this paper, the authors found that the retention times of substituted benzoate anions in IC can be explained by dividing the substituent effect into three terms: an LFERapplicable effect term expressed by using the pK, values of the species, a steric effect term expressed by using the steric effect indices (j’x), and a positional effect correction term
ANALYTICAL CHEMISTRY, VOL. 65, NO. 2, JANUARY 15, 1993
expressed by using the positional effect correction indices o’llo-x,ypx, yx-x’).
LFER~are applicable not only to substituted benzoates but also to other aromatic or aliphatic compounds.33J4 By applying the LFERs, it is expected that the theory in this paper will be valid for the analysis of the retention times of other organic ions in IC.
147
ACKNOWLEDGMENT This work was financially supported by the Murata Science Foundation and the Nippon Life Insurance Foundation, No. C88110012.
RECEIVEDfor review March 18, 1992. September 30, 1992.
Accepted
