the data provided by wet chemical total oxidant analyzers in current use. Current oxidant analysis provides only an index or number that is a function of the oxidizing property of the atmosphere; it does not provide the concentration of any specific component. The number obtained is a complex function of the concentrations of all atmospheric components that are capable of oxidizing or reducing an iodide-iodine solution. Recent comparative studies (16, 17) in areas such as Philadelphia, Cincinnati, Denver, and Bayonne, N. J., have shown many periods during which ozone concentration measured by chemiluminescence is higher than the concentration of total oxidant, expressed as ozone, measured by coulometric or colorimetric methods. The value of a method designed to measure the oxidizing property of the atmosphere seems questionable when this method yields a total oxidant value lower than the value of the principal oxidizing component. The advantages of the chemiluminescence method for (16) M. 6.Richter, J. R. Smith, and L. A. Ripperton, (Research Triangle Institute, Research Triangle Park, N. C.), National Air Pollution Control Admin., USDHEW-PHS, Final Report Contract No. PH-22-68-30(1969). (17) J. A. Hodgeson, K. J. Krost, and R. K. Stevens, 156th National ACS Meeting, Atlantic City, N. J., (Sept., 1968), No. Watr 012.
measuring atmospheric ozone should be emphasized. This is the most sensitive method known for ozone detection, and it is also specific for ozone. The instrumentation can be quite simple and, except for the photomultiplier, approaches the ideals of a solid-state gas detector. The primary disadvantage of the method is the slowly decaying sensitivity of the surface, which requires periodic recalibration. In addition, the calibration source is a secondary standard, which requires checking from time to time by some absolute method. ACKNOWLEDGMENT
The authors are grateful to A. D. Baitsholts of Distillation Products Industries for providing us with materials and suggestions for many of the chemiluminescent surfaces used. RECEIVED for review June 25, 1970. Accepted September 14, 1970. Some of these results have been presented at the 156th National ACS Meeting, Atlantic City, N. J., September 1968 and at the 157th National ACS Meeting, Minneapolis, Minn., April 1969. Mention of commercial products does not constitute endorsement by the National Air Pollution Control Administration.
~
~
errnination of rornatic Sulfonates h~ o ~ a t o ~ ~ a p ~ y uaoiph N.s t e ~ Analytical Laboratories, The Dow Chemical Company, Midland, Mich. 48640 The ion exchange separation and determination of a number of aromatic sulfonates is described. With this technique, the use of a cross-linked polyalkyleneamine anion exchange resin minimizes the usual absorption of these compounds. Using a mixed aqueous organic solvent system (water-acetonitrile-methanol) and LiCl salt gradients of 0 to 0.5M, elution of mono- and disulfonates occurs with good efficiency and resolution. Utilizing relatively small columns (500 X 2 mm) and monochromated UV absorption detection, the sensitivity and specificity for these materials can be high. Under these conditions, mono- and &sulfonates are base-line resolved within 30-45 minutes. Additionally, separation of isomeric or homologous sulfonates (mono- or di-) is reported for benzenesulfonates, ~ ~ e n ~ ~ s u ~ ~ o biphenylsuifonates, nates, and diphenyla k E SEPARATION A N D DETERMINATION O f Organic SUlfOnateS has remained one of the more difficult problems faced by the analytical chemist. Recent years have seen a number of powerful techniques applied to the analysis of aliphatic and aromatic sulfonates. Siggia et al. ( I ) have described a technique employing the alkaline fusion of arylsulfonic acids and salts to yield a sulfite salt and the corresponding phenol. Utilizing gas chromatographic analysis of the resulting phenols and potentiometric titration of the base released on formation of formaldehyde-bisulfite addition product, various sulfonates
(1) S. Siggia, L. R. Whitlock, and J. C . Tao, ANAL.CHEM.,41, 1387 (1969). I802
e
could be analyzed with good precision and accuracy. However, certain substituents such as amino or hydroxy groups yielded air-sensitive or unstable fusion products. A chromatographic technique ( 2 ) utiIizing a non-ionic porous polystyrene resin gives separation of mono- and disulfonated aromatic compounds, However, the selectivity in the presence of unsulfonated materials or hydrophobic substituents is less than desirable. Similarly, continuous electrophoresis has been employed (3) for the analysis of biphenyl sulfonates. Utilizing an 8-hour electrophoresis and ultraviolet measurement of the separated mono- and disulfonates, determinations at the k l % level can be made. Again, however, the discrimination for isomeric or homologous species is insufficient to permit analysis of these mixtures. The formation of volatile derivatives followed by gas chromatographic separation and measurement (4, 5 ) has resulted in successful determinations of mixtures, but again, the presence of certain substituents has resulted in either unstable derivatives or multiply-derivafized compounds. Additionally, the presence of the sulfonates in an aqueous or semiaqueous matrix requires either careful preparation of standards or some preparatory clean-up of samples to avoid decomposition under the conditions used for derivatization. (2) (3) (4) (5)
M. W. Scoggins, and J. W. Miller, ANAL. CHEM., 40,1155 (1969). N. E. Skelly, ibid., 37, 1526(1965). J. J. Kirkland, ibid., 32, 1388 (1960). J. S. Parsons, J. Gas Chromatogr., 5,254 (1967).
ANALYTIBAL CHEMISTRY, VOL. 42, NO. 14, DECEMBER 1970
Siggia (I) has summarized the previously mentioned methods as well as some others which have specific applications. The method described here uses a combination of liquid column chromatography (LCC) utilizing an ion exchange resin and ultraviolet absorption for the separation and measurement of aromatic sulfonates. By means of gradient elution techniques, separation of monosulfonates from disulfonates can be accomplished in 30-40 minutes. Additionally, with careful choice of gradient and solvent, separation of isomeric and homologous sulfonates (mono- or di-) can be achieved.
PLASTIC /TUBING
PACKED COLUMN
EXPERIMENTAL
Ion Exchange Resin. The ion exchange resin used was purchased as Bio-Rex 5 , chloride form, from Bio-Rad Laboratories, Richmond, Calif. 94804. As received, the resin had an ionic capacity of 8.8 meq/gram (1.8 meq/cc). The 200-325 mesh material was additionally wet-sieved to yield 200/270 and 2701325 mesh fractions. This sieved material was washed with aqueous 4 M HC1, deionized distilled water, and anhydrous ACS grade methanol and stored as a slurry in methanol. For comparison purposes, a conventional polystyrenedivinylbenzene quaternary alkyl ammonium resin (Dowex 1-X2 anion exchange resin), chloride form, was also used and treated similarly. Sulfonates. pBhenolsulfonic acid (and o-phenolsulfonic acid impuritj) was a technical grade product from The Dow Chemical Companj , Midland, Mich. 48640. p-Diphenylsulfonic acid, p-hydroxy-p‘-diphenylsufonic acid, and p,p’diphenyldisulfonic acid were prepared in the research laboratories of The Dow Chemical Company as were p-diphenyl oxide sulfonate sodium salt, and p,p‘-diphenyl oxide disulfonate disodium salt. The 0- and p-aminobenzenesulfonic acids were obtained from Eastman Organic Chemicals as reagent grade materials. In most cases, the compounds as received were of undetermined purity with respect to the presence of other sulfonated materials (isomers primarily). Where the chromatogram indicated the presence of a significant (>5 %) impurity, the identity of the components was determined by infrared spectrometry and comparison of the UV spectra with published spectra. UV spectra were recorded in 1-cm cells on a Cary 14 spectrophotometer. Infrared spectra were recorded on a PerkinElmer Model 337 spectrometer. The sulfonates studied were generally soluble in a mixture of water and methanol. In some cases, some acetonitrile was necessary to dissolve enough material to give 1 % solutions for injection into the chromatographic column. Solvents. The methanol and acetonitrile used jn this work were ACS grade reagent solvents. Mixtures were made by volume: 1 :1 :3 water :acetonitrile :methanol was prepared by mixing equal volumes of the three liquids. Chromatographic Monitor. The liquid chromatograph used for this ion exchange work is shown in Figure 1 in schematic form. The photometer for monitoring light absorption of the column effluent was constructed from a Beckman DU monochromator and a Gilford (Gilford Instrument Labs, Oberlin, Ohio 44074) Model 222 photometer light source supply and detector. With this arrangement, an output for recorder presentation linear in absorbance was available as well as the capacity for variable calibrated absorbance offset for the recording of large peaks without the necessity for range change. Additionally, the use of a monochromator permitted absorbance measurements over the range of 220 to 400 nm. This facility allows maximizing general sensitivity for aromatic sulfonates (240-280 nm) or optimizing selectivity as in the case of 0- and p-phenolsulfonic acid, which have quite dif-
CELL
Figure 1. Schematic diagram of liquid column chromatograph See text for description of components ferent UV spectra (6). The solvent and eluant were pumped with a Milton-Roy rniniPunip, Model 196-31. Inlet and outlet tubing for the pump was 0.022-in. i.d. and made of Teflon (Du Pont) (Chromatronix, Inc., Berkeley, Calif. 94710). Chromatographic columns and associated fittings were purchased from Chrornatronix. In general, the columns and tubing could be used up to flow rates of 1.5-2.0 ml per minute without leakage when 500- X 2-mm columns of 44-52 micron (270/325 mesh) resin were used. The cell was a Gilford model 203 flow-through cell assembly. Although the cell volume of the 1-cm cells is 125 gl, the flow pattern through the cell is reasonably efficient at flow rates of 0.3 ml/min or greater. The absorbance output of‘the detector was recorded on a Honeywell Electronik 194 recorder with 50-mV span. The recorder was equipped with a 10-speed chart drive and a 60:l jump speed to give a convenient presentation of elutions from 30 seconds to several hours. Columns and Packing. The “micro-bore” Chromatronix columns were packed with 2701325 mesh resin by pouring a slurry of the resin in the 1 :1 :1 solvent into a column filled with solvent and allowing the resin to settle with gentle suction. The pump was then connected and the resin compacted at 1 ml/min flow rate. The necessary resin to completely fill the column was then added aia a syringe. When packed in this manner, 500- X 2-mm columns had a void volume of about 50 %. Samples were applied in several ways. For most work, 10g1 samples were injected onto the resin bed using a 25-pl Precision Sampling Corp. (Baton Rouge, La. 70815) series C “pressure-lok” syringe. For quantitative measurements of peak height or area using the offset capability, 10-p1 portions of sulfonate solutions were injected into the solvent stream with a Chromatronix MSV-1-10 sample injection valve with a IO-pl bore. When tested with repetitive injections of a sulfonate solution, the precision of peak heights observed was observed to be ~ 0 . 8 %(rsd). In general, quantitation was accomplished by measurement of peak heights rather than areas. RESULTS
The performance of the resin can generally be described as good. The relatively high capacity (2.8 meq/ml resin bed) coupled with the intermediate base character of the ion exchange sites gives high efficiency to the column. This large (6) F. Langmaier, E. Muck, and D. Kokes, Collect. Czech. Chem. Commun., 24, 2066 (1959).
ANALYTICAL CHEMISTRY, VOL. 42, NO. 14, DECEMBER 1970
1803
HO 0
0.4
0.2
-
-
a
-
0 ~ ” ” ’ 0 0.1 0.2 03 0.4 0.5 0.6 0.7 O B 0.9 1.0
1.2
1.1
v ,cm / sec
Figure 2. Column efficiency as a function of Bow velocity Conditions: Column, 500 X 2 mm Bio-Rex 5, chloride form; solvent1:1:1 water :acetonitrile:methanol; UV detection at 250 nm,; sample size, 10 p1 (a) Unretained solute (benzene) (b) Diphenylsulfonates (K = 15-50)
20
56-
40
30
TIME, MINUTES
Figure 4. Ion exchange chromatography of p-hydroxyp ‘-diphenylsulfonate Conditions: Column, 500 X 2 mm Bio-Rex 5, C1-; flow rate, 1.0 ml/min; A, 250 nm; 1-cm cell path; 0.4 abs. unit full scale; solvent, 1 :1:1 water:acetonitrile:methanol; sample, 20 pg (20 pl); gradient, 1.OMLiCl into 75 ml of solvent
0.20
015
IO
Ez
W
4 3 LL
e
W LL
010
z
8 W
eK
005 u
0
10
20
30
40
0
50
TIME, MINUTES
Figure 3. Ion exchange chromatography of p-diphenylsulfonate Conditions: Column, 500 X 2 mm Bio-Rex 5, C1-; flow rate, 1.0 ml/min; A, 250 nm; 1-cm cell path; 0.4 abs. unit full scale; solvent, 1:1:1water :acetonitrIle:methanol; sample, 10 pg of sulfonate (5 pl); gradient, 0.33M LiCl into 40 ml of solvent
I
I
IO
20
I 30
I 40
I 50
I
60
TIME, MINUTES
Figure 5. Ion exchange chromatographic separation of diphenylsulfonates
capacity has the desirable effect of providing a large number of ion exchange sites within one plate distance. Under gradient elution conditions, the self-sharpening effect (7) which is enhanced by the use of concentration gradients results in improved performance by a factor of 2 to 5 (decreased plate height). Figure 2 shows the conventional vanDeemter plot of H (plate height) as a function of linear velocity of the mobile phase. As can be seen from curve a, unretained solutes exhibit significant band-spreading. Curve b of Figure 2 illustrates the decrease in plate height which can be obtained under gradient elution conditions for three sulfonated diphenyl compounds. The decrease in plate height shown here is typical of the improvement obtained on other sulfonates. The advantages of gradients in liquid-solid adsorption column
chromatography have been admirably demonstrated by Snyder (8,9). Although the shape of the gradients formed with the use of the device shown in Figure 1 does not produce the most desirable shape of gradient (semi- or full-logarithmic), it is extremely useful with respect to developmental studies because of the ease of solvent and/or concentration changes. Figure 3 shows the results obtained for the ion exchange chromatography of p-diphenylsulfonate. Also indicated
(7) F. Helferich, “Ion Exchange,” McGraw-Hill, New York, 1962, p 425.
(8) L. R.Snyder, J. Chramatogr. Sci., 7,395 (1969). (9) Ibid.,7,595 (1969).
1804
Conditions: Column, 500 X 2 mm Bio-Rex 5, CI-; A250 nm, 0.4 abs. unit full scale; 1-cm cell path; flow rate, 0.5 ml/min; solvent, 1:1:1 water:acetonitrile:methanol; sample, 10 pg each component; gradient, 1.OM LiCl into 40 ml of solvent
ANALYTICAL CHEMISTRY, VOL. 42, NO. 14, DECEMBER 1970
I
-0.30= i
-0.25
OH
L JL
- 0.20kb t W V 2 m 4
ci
-015
a
8
w n
_ _ 0.10 E
v) 0
m
s
a
J
-00.05u
0
10
20 30 TIME, MINUTES
40
50
Figure 8. Ion exchange chromatographic behavior of aminobenzenesulfonates I
I
I
20 30 TIME, MINUTES
IO
0
I
I
40
50
Figure 6. Ion exchange chromatography benzene- and phenolsulfonic acids
Conditions: Column, 500 X 2 mm Bio-Rex 5, CI-; flow rate, 1.0 ml/min; A, 250 nm; 0.4 abs. unit full scale; solvent, 1:1:1 aater:acetonitrile:methanol; eluant, 0.1M LiCl in solvent (no gradient); sample, 10pg each component
of
Conditions: Column, SO0 X 2 mm Bio-Rex 5, C1-; flow rate, 1.0 ml/min; X, 250 nm; 0.4 abs. unit full scale; 1-cm cell path; sample, 10 pg each major component; gradient, LOA4 LiCl into 50 ml 1:l:l solvent
Table I. Efficiency and Resolution of Aminosulfonate Elution Elution conditions Gradient Constant strength Compound H R H R Dimethylorthanilic acid 0.21 0.91
-0.40
0.78
Orthanilic acid /
”
CH3
/’ u W
-0.30:
- 0.25 2
/’
z m a
W LL
0:
am 4
/
0
IO
/
/
/
/’
20
30
40
0.64
0.56
70.35
z 3 W
/’
0.15
50
TIME, MINUTES
Figure 7. Ion exchange chromatographic behavior of aminobenzenesulfonates Conditions: Column, 500 X 2 Bio-Rex 5, Cl-; flow rate, 1.0 ml/min; 0.4 abs. unit full scale; 1-cm cell path; X, 250 nm; solvent, 1:l:l water :acetonitrile:methanol; gradient, 0.66 M into 50 ml solvent; sample, 10 pg orthanilic 20 fig dimethyl orthanilic acids
+
is the shape of the gradient produced. Figure 4 shows the effect of a gradient with about twice the slope on the elution of the p-hydroxy-p I-diphenylsulfonate. The separation of a mixture of three diphenylsulfonates is shown in Figure 5 using a gradient whose slope is still greater. The base-line resolution of the components permits an easy isolation of similar separated unknown materials for subsequent examination (UV, IR, NMR, etc.). Figure 6 shows the chromatographic behavior of benzene-
sulfonic acid and phenolsulfonic acid. The small peak prior to elution of the p-phenolsulfonic acid in curve b is due to ophenolsulfonic acid shown to be present to the extent of 5 by independent UV examination. The separation of two aminobenzenesulfonates shown in Figure 7 demonstrates the ability of this technique to resolve closely related homologs. The equivalent separation of the same two compounds under constant eluant strength is shown in Figure 8. Using the notation of Scott (IO), the plate heights and resolutions for these two separations are shown in Table I. In this case, each chromatogram represents an acceptable separation from an analytical point of view. In fact, for all but extremely precise determinations, the slightly faster elution would probably make the constant strength elution preferable. In spite of the significant difference in plate height for the different elution conditions, the high capacity of this resin permits the above efficient separation in quite acceptable times. An indication of the resolving ability of this resin for isomers of aminobenzenesulfonates is shown in Figure 9. Although the separation is not complete (R = 0.30), this demonstrates again the potential for separation of aromatic sulfonates by this technique. A comparison of this method using the cross-linked polyalkylene-amine resin and the elution of the same materials from the conventional styrene-divinylbenzene matrix resin
x
(10) R. P. W. Scott, “Gas Chromatography, 1960,” Butterworth, London, 1960, p 423.
ANALYTICAL CHEMISTRY, VOL. 42, NO. 14, DECEMBER 1970
1805
w
z m
8 m
4
IO
20
30
40
50
60
TIME, MINUTES
Figure 9. Ion exchange separation of aminobenzenesulfonate isomers Conditions: As in Figure 4
Table 11. Comparison of Bio-Rex 541 and Dowex 1-X2-ClColumn volume for elution (LiCI), molarity Bio-Rex Dowex Biox-Rex Dowex Components 5 1-x2 5 1-x2 Monosulfona te 5 25 0.03 0.28 Disulfonate 15 70 0.16 0.58
is shown in Table I1 where the elution of a mono- and disulfonated chlorinated alkylated diphenyloxide species (Dowfax 6A1 brand) in terms of the volume and concentration of eluant is given. DISCUSSION
Kirkland ( 1 1 , I Z ) has demonstrated the extremely fast and efficient separations that can be obtained using pellicular ion exchange materials. Similarly, Horvath and Lipsky (13) have indicated the advantages to be gained from minimizing intra-particle diffusion in high efficiency column chromatography. The small amount of published work on the ion exchange behavior of aromatic ions can probably be interpreted as the result of the difficulties in absorption of these materials into the conventional polystyrene resin matrix. Although ion exchange chromatography has been variously used for the separation of organic anions and cations, the problem of absorption of the hydrophobic portion of the molecule into the polymeric resin has restricted the application of ion exchange for organic separation. Feitelson (14) has (11) J. J. Kirkland, J. Chromafogr.Sci., 7 , 361 (1969). (12) J. J. Kirkland, Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Cleveland, Ohio, March 1969, paper No. 137. (13) C . Horvath and S.R. Lipsky, “Advances in Chromatography,”
A. Zlatkis, Ed., Preston Technical Abstracts, Evanston, Ill., 1969, p 268. (14) J. Feitelson, “Ion Exchange: Vol. 2,” J. Marinsky, Ed., Marcel Dekker, New York, 1969, p 135. 1806
*
discussed in some detail the types of forces which can be involved in the interaction of the resin with organic ions. Although the van der Waal’s interactions can become important as the number of cavities and absorption sites in the resin increases, it is the difference between the attractive forces of the solvent (mobile phase) and the resin (stationary phase) that determine the distribution coefficient under any given conditions. Samuelson (15) has outlined the type of intuitive approach which can explain but only qualitatively predict the behavior of organic species in solution in the presence of ion exchange resin. Nevertheless, the ion exchange separation of amino acids (16) has developed into a major analytical technique. Peterson (17) has reported distribution coefficients for several aromatic sulfonates which are larger than the bisulfate and bisulfite constants by an order of magnitude. This gives some indication of the extent to which aromatic absorption in the polyaromatic resin can influence the ion exchange process. As indicated in Table 11, even under these mixed solvent conditions, the absorption of aromatic sulfonates is appreciable relative to the polyalkyleneamine ion exchange resin. The advantages of small particle size and high accessibility of the solutes to the stationary phase will not be discussed in any detail here. Horvath and Lipsky (13), Kirkland (18, 19), Snyder (20), Sie (21), and Knox (22) have all discussed in some detail the various aspects of current high-efficiency liquid column chromatography. Although several studies using different solvent systems comprised of varying rations of water, acetonitrile, and methanol were investigated, the 1 :1 :1 mixture used here gave the best performance with respect to peak shape, ease of elution (minimum concentration of LiCI), and solubility for the wide range of sulfonates studied. As Samuelson ( I S ) has noted, the amount of water (or conversely, alcohol) plays an important part in determining the relative amount of absorption of solutes into ion exchange resins. In this system, too, the relative amounts of solvents influence the relative elution of these sulfonates. In general, as the amount of acetonitrile increased, the elution is facilitated. The influence of acetonitrile may be due to the high polarity of this molecule and its ability to mediate the short range dispersion forces responsible for absorption by the resin matrix. The use of this cross-linked polyalkylene amine resin with relatively high capacity and efficiency permits the use of the 500- X 2-mm analytical columns for semipreparative purposes as well. Since the total capacity of the column is about 3 milliequivalents, IO-mg quantities of sulfonate mixtures can be applied to the colunm and the separated components collected. The lithium co-ions may be removed by passage through a column packed with Dowex SOW-X8-hydrogen form cation exchange resin and the solvent removed by vacuum evaporation. In this manner, sufficient material for subsequent examination or standardization may be obtained without resorting to separate preparative scale columns.
RECEIVED for review June 19, 1970. Accepted September 10, 1970. (15) (16) (17) (18) (19) (20) (21) (22)
0.Samuelson, ibid., p 167. S.Moore and W. H. Stein,J. Biol. Chem., 192,663 (1951). S. Peterson, Ann. N . Y. Acad. Sci., 57,144 (1953). J. J. Kirkland, J. Chromatogr. Sei., 7, 7 (1969). J. J. Kirkland, ANAL.CHEM., 41,218 (1969). L. R. Snyder,J. Chromatogr. Sci., 7,352 (1969). S. T. Sie and N.VandenHoed, ibid., p 257. J. H. Knox and M. Saleem, ibid., p 614.
ANALYTICAL CHEMISTRY, VQL. 42, NO. 14, DECEMBER 1970
