Anal. Chem. 1983, 55, 851-854 (9) Arpino, P. A ; Guiochon, G. J . Chromatogr. 1982, 251, 153. (10) Zander, A. I.; Hieftje, G. M. Appl. Spectrosc. 1982, 35, 357. (11) Carnahan, ,J. W.; Mulligen, K. J.; Caruso, J. A. Anal. Chim. Acta 1981, 130, 227. (12) Beenakker, C. I. M. Spectrochim. Acta, Parts 1976, 316, 483. (13) Beenakker, C. I. M. Spectrochim. Acta, Parts 1977, 328, 173. (14) Van Dalen, J. P.J.; De Lezenne Coulander, P.A,; De Galan, L. Spectrochlm. Alcfa, Pari 8 1978, 336, 299. (15) Snelleman, W.; Rains, T. C.; Yee, K. W.; Cook, H. D.; Menis, 0. Anal. Chem. 1070, 42, 394. (18) Van D a h , J. P. J.; De Lezenne Coulander, P. A.; De Galan, L. Anal. Chim. Acta 1977, 94, 1.
85 1
(17) Slaats, E. H.; Markovski, W.; Fekete, J.; Poppe, H. J . Chromatogr. 1981, 207, 299.
RECEIVED for review June 28, 1982. Resubmitted Decernber 20,1982. Accepted January 25,1983. This paper was presented at the Symposium Detection in High Perform,ance Liquid Chromatography, January 19827 Amsterdam, The Netherlands. 19-209
Determiination of Inorganic Anions by Ion Chromatography wiith Ultraviolet Absorbance Detection Rlchard J. Williams Allied Corporati&n, Chemical Research Laboratory, Morristown, New Jersey 07960
This work describes the appllcatlon of varlable wavelength UV detectlon In sorles with the normal conductlvlty detector, In ion chromatography for the determlnatlon of lnorganlc anions. This comblnatlon of detectors greatly Increases the amount of informatlon that can be collected on a glven sample. The appllcatlon of IJV detection has the foliowlng advantages: (1) ald In the identtflcatlonof unknown peaks, (2) use In resolving overlapplng peaks, (3) help In ellmlnatlng problems assoclated wlth the carbonate dlp, (4) reductlon of problems assoclated wlth ion excluislon In the suppressor column, (5) ablllty to detect anlons not normally detected by the conductlvlty detector, e.g., stilflde and arsenlte.
The technique of ion chromatography (IC) has gained wide acceptance for the analysis of inorganic ions in a variety of aqueous matrices (1-5). The technique involves separation of the ions of interest on a low capacity ion-exchange column followed by a suppressor column and a conductivity detector. The suppressor column consists of a high-capacity ion-exchange column opposite in type to the separator column. The sole purpose of the suppressor column is to decrease the background conductance of the eluent, usually by neutralization, so that the ionic species of interest can be detected with sufficient sensitivity by the Conductivity detector. Cation analysis is accomplished with millimolar acidic eluents, while anion analysis is accomplished with a mixed millimolar NaHC03/Na2C03eluent. Although UV detection is the most common form of detection in HPILC, it has been considered unsuitable for IC. This has come about from the widely held belief that most inorganic ions lack suitable chromophores for UV detection (6-8). This is not always the case, especially when dealing with inorganic anions. Buck et al. showed that many inorganic ions exhibit strong absorption below 220 nm (9). In a recent paper, Reeve was able to separate many of the common inorganic anionri on a cyane-bonded silica column using a spectrophotometer for detection a t 210-220 nm (IO). Leuenberger et al. determined nitrate and bromide in foodstuffs after separation on an amino-bonded silica column, using a UV detector at 210 nm (12). Bouyoucos and Armentrout combined a UV detector with the normal IC conductivity detector to monitor dialkylated organophosphorothionic acids which were obscured conductometrically by chloride (12). The 0003-2700/83/0355-085 1$01S O / O
author used the same combination of detectors to detect aromatc and unsaturated sulfonic acids by IC (13). The aim of this work was to examine the potential of combining a UV detector with the conductivity detector for the detection of inorganic anions separated by IC.
EXPERIMENTAL SECTION A Dionex Model 14 ion chromatograph (Sunnyvale, CA) was used throughout the study. Chromatograms were recorded on a Honeywell (Fort Washington, PA) dual pen strip chart recorder. The chromatographic conditions are summarized in Table I. All of the columns were obtained from Dionex. A 100-/.~Lsample loop was used for all injections. A laboratory Data Control (Rivieria Beach, FL) Spectromonitor I1 variable wavelength UV-Vis detector was used along with the normal conductivity detector. All of the inorganic anions, as sodium or potassium salts, !were obtained commercially and used as received. Eluents were made from ACS reagent grade chemicals, except the 0.01 N HCl eluent which was made from Ultrex hydrochloric acid (J. T. Balker). Water was purified with a Milli-Q system (Millipore Corp.). RESULTS AND DISCUSSION The UV detector can be inserted into the system at two locations: (1) directly after the separator column (unsuppressed eluent, position l),(2) after the suppressor column (suppressed eluent, position 2). Figure 1shows the absorption spectra of both the unsuppressed and suppressed standard eluent, 0.003 M NaHC03/0.0024 M Na2C0,, and illustrates that the two positions are not equivalent. The suppressor column, besides decreasing the high background electrical conductance of the eluent, also decreases the high background absorbance of the eluent in the 190-220 nm region. This allows UV detection in the 190-220 nm region where most inorganic anions absorb most intensely. Thus, for maximum sensitivity, the preferred position of the UV detector is after the suppressor column. The suppressor column also decreases the background absorbance of the other common IC eluents such as 0.002 M Na2C03/0.002 M NaOH, 0.006 M Na2CO9, and 0.005 M NazB40,.10H20. Figure 2a shows a separation of seven common anions monitored using the conductivity detector. Figure 2b was obtained with the UV detector after the suppressor column. As is illustrated in Figure 2, nitrite, bromide, and nitrate absorb strongly in the UV, while fluoride, phosphate, and sulfate do not show appreciable absorption above 190 nm. Chloride absorbs weakly in the W region below 200 nm. Note that nonabsorbing anions are sometimes observed by the UV 0 I983 American Chemical Society
852
ANALYTICAL CHEMISTRY, VOL. 55, NO. 6, MAY 1983
Table I. Chromatographic Conditions
B
C
D
3 X 250 mm L-20 X 250 mm anion'
6 X 250 mm anion'
9 x 250 mm
0.006 M Na,CO,
H2O
4 x 140 mmb 4 X 140 mmC 0.01 M HCl
2.3 mL/min
0.9 mL/min
0.9 mL/min
A
precolumn separator column suppressor column
4 X 50 mm anion 4 X 250 mm anion 6 X 250 mm aniona
eluent
0.003 M NaHCO, 0.0024 M Na,CO, 2.3 mL/min
flow rate
3 X 150 mm L-20 6
' Anion suppressor column consists of a high capacity cation-exchange column in H' form. Postsumressor column. 10
Halide suppressor column.
4 1
1
0.8
a I
I
0 3 t
-'i
,
,
,
..J,#,
..-
I
l
l
4
8
12 1 6 20 24
1
1
l
b
,
,-.___ b, ,
SOD 280 1 8 0 270 280 250 240 230 120 210 200 i s 0
Wavelength (mm)
Figure 1. Absorption spectra of standard eluent, 0.003 M NaHCO,/ 0.0024 M Na2C03: (a)unsuppressed eluent (-); (b) suppressed eluent
-
(- -).
detector as negative peaks, as is shown in Figure 2b for phosphate. Table I1 contains a more extensive list of UV active anions and also lists the detection limits for many of the anions. Detection limits are generally in the sub-partsper-million range. No attempt was made to optimize the chromatographic conditions for a particular anion. Out of the 27 anions listed in Table 11, all but eight absorb in the UV above 190 nm. This demonstrates that the commonly held assumption regarding the lack of suitable chromophores by inorganic ions for UV detection is incorrect. This misconception has probably come about from the fact that although many anions absorb in the 190-220 nm region, few of them show an absorption maximum above 190 nm. This has limited the analytical usefulness of direct UV absorbance measurement for the determination of inorganic anions. However, the lack of an absorption maxima does not severely limit the analytical usefulness when the UV spectrometer is combined with a separation technique such as ion chromatography. The UV detector, in combination with the conductivity detector, offers a number of advantages in IC. Figure 3 shows the separation of sub-part-per-million iodate and bromate monitored using both detectors. The iodate and bromate peaks monitored with the conductivity detector (solid line chromatogram) were distorted by the so-called carbonate dip (14). The size of the carbonate dip is controlled by the eluent (both its concentration and composition), the size of the injection loop, and the size of the suppressor column. The peak distortion caused by the carbonate dip makes both qualitative and quantitiative analysis more difficult. Little or no carbonate dip is observed in Figure 3 (broken line chromatogram) with the UV detector after the suppressor column (position 2). However, when the UV detector is placed directly after
,
,
4
8
#
I
,
,
1 2 16 20 Minutes
Figure 2. Ion chromatogram of seven common anions (Table I , column A): (a) conductivity detector (-), peak 1, 3 ppm F-; peak 2, 4 ppm CI-; peak 3, 10 ppm NO2-;peak 4, 50 ppm PO:-; peak 5, 10 ppm B r ; peak 6 , 3 0 ppm NO3-; peak 7,50ppm SO:-; (b) UV detector (---) at 192 nm, positlon 2; peak 2, 4 ppm CI-; peak 3, 10 ppm NO2-; peak 5, 10 ppm Br-; peak 6, 30 ppm NO3-. 1
$&x
I
I{
, I'Ij $,
Yj1J
2
1;
It
h(qSeO,aa N, not UV active, 2 X noise level.
3
-
2.4 2.4 2.6 3.0 3.0 3.8 4.3 4.6 5.4 5.9 6.6 6.7 6.9 7.8 8.4 10.8 10.9 12.2 12.6 13.0 13.3 13.4 14.1 15.2 15.2 18.8 19.6 21.2 22.8 27.2
Y, UV active,
UV detector position
wavelength, nm
chromatographic condition Table I
Na
2c
Yb
Id
195 205 195 195 195 195 195 192 195 195 195 195 195 195 195 195 200 195 195 195 200 195 195 195 195 195 195 200 195 195
A A A A A A A A C A B A A B A B D A B A D A A A A B B D A A
UV active
UV detection limit,e ppm
Y Y N N Y Y Y Y N
0.08 0.2
2 2 2 2 2 2 1 2 2 2 2 2 2 2 1 2 2 2 1 2 2 2 2 2 2 1 2 2
0.16 2 0.1
N
Y Y N Y Y Y Y Y Y Y Y Y Y Y N Y N Y 2, after suppressor.
0.5 0.15 0.2 1.5 2 0.1 1.2 0.3 0.1 4 0.4 0.4 15
1, before suppressor. e 100 p L injection, peak height
-
A I
carbonate dip if a UV detector can be used. Other UV active anions that elute in or near the carbonate dip include chlorite, chloride, formate, sulfide, and arsenite. Figure 4a shows the UV and conductivity peaks from a 30 ppm nitrate injection while Figure 4b shows the UV and conductivity peaks from a 30 ppm chlorate injection. Both species have the same retention time and peak shape (tailing), and with just the information from the conductivity detector it would be implossible to identify which species is present. With both detectors, it is possible to obtain a UV/conductivity peak height ratio, and this information can be used as an aid in identifying unknown peaks. Additional UV/conductivity peak height ratios can be obtained by varying the wavelength. Identification ie then made by running a standard solution of the suspected species (same retention time) and comparing the peak height ratios at the same wavelengths as the unknown species. The ion chromatographic determination of weak-acid anions is complicated by ion exclusion in the suppressor column, resulting in faster elution and sharper peaks, directly proportional to the degree of exhaustion of the suppressor column (6, 15). A 10 ppm nitrite standard showed a 37% increase in peak height over an 8-h period when monitored with the conductivity detector, while only a minor 2% increase in peak height was observed over the same time period, by using the UV detector in position 1. This demonstrates another advantage of the 'UV detector and shows that the preferred position for the UV detector is directly after the separator column (positioin 1) for UV-absorbing weak-acid anions. The UV detector can also be used in some cases to resolve overlapping peaks. Figure 5 shows the ion chromatogram of 10 ppm nitrite in the presence of 1000-2000 ppm chloride. Determination of the nitrite peak by using the conductivity detector is complicated by both the ion-exclusion effect and
iB53
A
b
-12
16 20 Minutes
I
12
I
I
16
20
Minutes
Flgure 4. Separation of nltrate and chlorate by ion chromatography (Table I, column A): (a) conductivity detector (-), 30 ppm NO3- IUV detector (---) at 195 nm, position 2; (b) conductivity detector (-), 30 ppm CI0,- UV detector (---) at 195 nm, position 2.
incomplete resolution between the large chloride peak and the much smaller nitrite peak. As was previously shown, the ion-exclusion interference can be eliminated for UV-active anions by placing the UV detector between the separator and suppressor columns (position 1). In addition, the problem of overlapping peaks can sometimes be resolved spectrophotometrically by proper choice of wavelength, as shown in Figure 5 (broken line chromatogram). Sulfide cannot be detected under normal ion chromatographic conditions. It is converted into hydrosulfuric acid (H,S) in the suppressor column. Hydrosulfuric acid is a very weak acid that does not ionize sufficiently to be detected with the conductivity detector. The UV detector, however, is able to detect sulfide a t low levels, as is illustrated in Figure Eia.
854
ANALYTICAL CHEMISTRY, VOL. 55, NO. 6, MAY 1983
4
8
Mlnules
Flgure 5. Nitrite determinatlon in the presence of 1000-2000 ppm CI(Table I , column A): UV detector (---) at 210 nm, position 1; conductivity detector (-); peak 1, 1000-2000 ppm CI-; peak 2, 10 ppm
NOa-.
MINUTES
Flgure 7.
Separation of arsenite and arsenate by ion exclusion chromatography (Table I , column D): conductivity detector (-); UV (---) at 200 nm, position 1; peak 1, 15 ppm AsO?-; peak 2, 30 ppm ~s0,3-.
out; then a second arsenate determination is carried out after a casutic hydrogen peroxide oxidation has converted arsenite to arsenate. Arsenite also can be separated by IC using standard conditions (Table I, column A). It elutes early, near the carbonate dip. For this application the UV detector should be placed between the separator and suppressor columns.
Mlnutes
Mlnules
Flgure 6, Sulfide separation:
(a) Ion chromatographic conditions (Table I, column A): conductlvity detector (-); UV detector (---) at 215 nm, position 1; peak 1, 10 ppm S"; peak 2, SO3"; peak 3, S O:-. (b) Ion exclusion chromatographic conditions (Table I, column C): conductivity detector (-); UV detector (---) at 192 nm, position 1; peak 1, SOS2-;and SO.,'- peak 2, SO?-;peak 3, 15 ppm S2-.
Since sulfide is a weak-acid anion, the UV detector was placed between the separator and suppressor columns (position 1). Figure 6a also shows the presence of sulfite and sulfate due to oxidation of the sulfide ion. Ion exclusion, besides being a source of chromatographic interference, can also be used to separate weak acids (7, 16, 17). Figure 6b shows the separation of sulfide by ion exclusion chromatography (IEC), using a strong cation-exchange column and water as the eluent. Peak 1is a combined peak resulting from both sulfite and sulfate, as monitored with the conductivity detector. Peaks 2 and 3 correspond to sulfite and sulfide, respectively, as monitored using the UV detector. Since sulfide can be separated by either IC or IEC, the choice of separation mode will depend on the sample matrix. Figure 7 shows the separation of arsenite (AsO$-) and arsenate (As0& by IEC using 0.01 M HC1 as the eluent. Arsenous acid (H3As03) is a very weak acid and cannot be detected at low levels by the conductivity detector. However, like sulfide, it is easily detected with the UV detector. The simultaneous determination of arsenite and arsenate is now possible. Previously, the IC determination of arsenite was done indirectly by difference (2). This required two arsenate determinations: an initial arsenate determination is carried
CONCLUSIONS Conductance is still the preferred mode of detection in ion chromatography. However, the use of a UV detector, in series with the conductivity detector, greatly increases the amount of information obtainable on a given sample and can eliminate many of the problems associated with the use of a suppressor column. R e g i s t r y No. Tetrafluoroborate, 14874-70-5.
LITERATURE CITED Mulik, J.; Puckett, R.; Wllllams, D.; Sawicki, E. Anal. Lett. 1976, 9 , 653-663. Hansen, L. D.; Rlchter, B. E.;Rollins, D. K.; Lamb, J. D.; Eatough, D. J. Anal. Chem. 1979, 5 1 , 633-637. Dulski, Thomas R. Anal. Chem. 1978, 5 1 , 1439-1443. Wetzel, R. A.; Anderson, C. L.; Schlelcher H.; Crook, G. D. Anal. Chem. 1979, 5 1 , 1532-1535. Siemer, D. Anal. Chem. 1980, 52, 1874-1677. Small. H.: Stevens. T. S.: Bauman. W. C. Anal. Chem. 1975, 4 7 , 1801-1609. Pohl, C. A.; Johnson, E. L. J. Chromatogr. Scl. 1880, 18, 442-452. MacDonald, J. C. Am. Lab. (Fairfisld, Conn.) 1979. 11. 45-55. Buck, R. P.; Singhadeja, S.;Rogers. L. B. Anal. Chem. 1954, 26, 1240-1242. Reeve, R. N. J. Chromatogr. 1979, 177, 393-397. Leuenberger, U.; Gauch, R.; Rieder, K.; Baumgartner, E. J. Chromatogf. 1980, 202, 461-466. BoU~ouCos,S. A.; Armentrout, D. N. J. Chromatogr. 1980, 189, 61-71. Williams, R. J. Paper presented at 22nd Rocky Mountain Conference on Analytical Chemistry, Denver, CO: Aug 1960: Paper No. 23. Stevens, T. S.; Davls, J. C.; Small, H. Anal. Chem. 1981, 5 3 , 1466-1492. Koch, W. F. Anal. Chem. 1979, 5 1 , 1571-1573. Tanaka, K.; Ishizuka, T.; Sunahara, H. J. Chromatogr. 1879, 174, 153-1 57. Tanaka, K.; Ishiruka, T. J. Chromatogr. 1880, 190, 77-83.
RECEIVED for review July 21,1982. Resubmitted and accepted January 19, 1983. Portions of this work were presented as Paper No. 308 at the 1981 Pittsburgh Conference, March 9-13, 1981.
