376
Anal. Chem. 1985, 57,376-378
steric problems due to the silica surface should be present with either of these isomers. After separation of the ortho isomer from the commercially available mixed isomers, synthesis of silica-immobilized 8quinolinol from the various fractions gave the following results: 90% ortho, 41 pmol/g capacity; 97% meta para, 237 pmol/g capacity; and 97% of all isomers, 220 pmol/g. As expected, the ortho isomer gave much lower capacities than the meta and para isomers or the entire mixture of isomers. This is due to the steric effect of the silica surface on the reactive amine functional group as we have speculated previously ( I , 2). The bulky 8-quinolinol molecule cannot couple effectively due to these steric problems. Somewhat disappointing, however, was the relatively small increase in capacity observed when performing the synthesis with a fraction containing primarily meta and para isomers. This indicates that probably a large amount of the silane is attached to the silica surface, but the bulky 8-quinolinol moiety “covers up” more than one of the bound aromatic amine groups. Therefore, it appears that the limit to capacity (if the synthetic conditions and properties of the silica carrier are held constant) occurs due to the sheer size of the 8quinolinol molecule. The presence of unreacted silane (either ortho or unreacted meta or para), however, provides an explanation for the larger capacities obtained from carbon determinations vs. metal ion (copper(I1)) studies (I) since a large
+
amount of organic matter is attached to the silica surface but is not subsequently coupled to 8-quinolinol. In fact, comparison of capacities obtained by elemental analysis and copper(I1) uptake indicates that only one out of two or three of the aromatic amine groups becomes coupled to 8-quinolinol.
Registry No. SiOz, 7631-86-9; (0-aminopheny1)trimethoxysilane, 93383-46-1;(m-aminophenyl)trimethoxysilane, 70411-42-6 (p-aminophenyl)trimethoxysilane,33976-43-1; 8-quinolinol, 148-24-3. LITERATURE CITED (1) Marshall, M. A.; Mottola, H. A. Anal. Chem. 1983, 55, 2089-2093. (2) Marshall, M. A.; Mottola, H. A. Anal. Chim. Acta 1984, 158, 369-373.
(3) Analabs Guide to ChromatographicPhases, PUB-N10; Analabs: North Haven, CT, 1981.
’
Present address: Monsanto Co., Corporate Research Center, 800 N. Lindbergh Blvd., St. LOUIS,MO 63167.
Monte A. Marshall’ Horacio A. Mottola* Department of Chemistry Oklahoma State University Stillwater, Oklahoma 74078
RECEIVED for review September 11, 1984. Accepted October 22, 1984.
Separation of Methyl Violet 2B by High-speed Countercurrent Chromatography and Identification by Californium-252 Plasma Desorption Mass Spectrometry Sir: Many dyestuffs are sold as mixtures designed to produce a subtle shade of color for some industrial process ( I ) . Characterization of their colored components by paper chromatography or TLC is usually a simple matter, at least when the dyes are not closely related. Exact chemical identification of the components is another matter since many dyes are complex organic salts that do not lend themselves to analysis by the techniques of gas chromatography or ordinary mass spectrometry. Because of their complexity and their tendency to form aggregates in solution, even ‘HNMR may prove difficult ( 2 ) . We recently encountered a sample of Methyl Violet 2B (1, Rl-R, = CH3, = H) (Aldrich Chemical Co., Milwaukee, WI)
Gz
CI -
1
whose 262Cfplasma desorption mass spectrum (PDMS) (3), in agreement with previous secondary ion mass spectrometry (SIMS) results (4),suggested it was largely the higher homologue, Crystal Violet (1, R1-R6 = CH3). Apparent loss from the cetion of methylene was explicable partly (3) on the basis of contamination by the lower homologue. Since these compounds differ from one another only by N-methyl groups, separation was anticipated to be difficult (indeed, chromatography on silica gel using chloroform/acetic acid/O.l N
hydrochloric acid (2/2/1) provided a chromatogram showing only one spot, R, = 0.64) and recourse was made to the high-speed countercurrent chromatography technique (coilplanet centrifuge) recently developed by Ito and co-workers (5).
EXPERIMENTAL SECTION Separation of the dye Methyl Violet 2B (Lot No. D11097, Aldrich Chemical Co., Milwaukee, WI) (6 mg in 5 mL of moving phase) was effected on a high-speed countercurrent chromatograph (PC, Inc., Potomac, MD) using a multilayer Teflon tubing (1.6 mm i.d.) coil with a revolutional radius of 4 in. and a rotational radius range from 2 to 2.5 in. The solvent system consisted of preequilibrated chloroform/acetic acid/O.l N hydrochloric acid (2/2/1). The coil was charged with the aqueous layer (stationary phase) and the chloroform layer (moving phase) pumped through from the internal head end of the multilayer coil at 240 mL/h with a Milton Roy pump at a setting of 50%. A total of 84 8-mL samples were collected in a fraction collector, analyzed, using a Perkin-Elmer Model 552 spectrophotometer, and finally evaporated to dryness. The 252CfPDMS system has been previously described in detail (6). Samples in methanol were electrosprayed onto a 1.7 cm radius aluminized Mylar disk and placed in the spectrometer. Spectra were accumulated for 20-30 min each and analyzed with a Perkin-Elmer Model 3220 computer.
RESULTS AND DISCUSSION Purification of the dye was effected using phases consisting of acetic acid, chloroform, and 0.1 N hydrochloric acid (2/2/1). The separation (Figure 1)revealed the presence of a t least seven dyes. Mass spectra (PDMS) of the first four compo-
0003-2700/85/0357-0376$01.50/00 1984 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 57, NO. 1, JANUARY 1985
XlO
1
-
1
30r
MW
I
FRACTIONS 5 5 -
1m-
-
344
(TETRAMETHYL)
M W 372
177
Y x
.
INCREASING HYDROPHILICITY
377
FRACTIONS 59 70 303
,345
ELUATE FRACTION (&nl)
Figure 1. Countercurrent chromatogram of Methyl Violet 2 8 certified (Lot No. 01 1097, Aldrich).
FRACTIONS 81
m
m
-L_lhi!372iC'l
MASS
FRACTIONS B 1 5
Figure 3. PosMve ion californium-252 in PDMS of fractions 55-84 from countercurrent chromatogram of Methyl Violet 28.
Ym
1358 IC'I
lma -
-
methyl group from the neutral base (or its radical ion) produces the cation of the lower homologue. In support of this hypothesis, the spectrum of a sample of the highly purified dye, C r y s d Violet (identical with fractions 6-11 above) shows this same loss. This ion therefore must not by itself be taken to imply presence of a lower homologue. The dye in ftactions 55-64 (Figure 3) showed a similar 581 nm) to that of the first four absorption spectrum (A, dyes but the PDMS revealed the molecular weight of the cation to be 119 amu higher (m/z 491) than the hexamethyl derivative. A simple explanation entails substitution of one of the methyl groups by a fourth (dimethy1amino)befizyl radical (2), perhaps a byproduct of the manufacturing process.
FRACTIONS 15 18
urn-
A
-
FRACTIONS 20 25
"i
3301c'l
m
FRACTIONS 34 40
303
3M
NICH,),
4SO
0
CI-
MASS
Figure 2. Positive ion caiifornlum-252 PDMS of fractions 6-40 from countercurrent chromatogram of Methyl Violet 28.
nents (Figure 2) revealed a series of homologous cations at mlz 372,358,344, and 330. The first two, responsible for the major portion of the dye (Figure 1))correspond to the expeckd hexamethyl and pentamethyl derivatives while the last two are the tetra- and trimethyl derivatives. The latter two peaks are somewhat broad, possibly due to the presence of several isomers, e.g., tetramethyl (R1-R4 = methyl, R5R6 = H or R1R2R3R5= methyl, R4Rs = H). The triarylmethane dyes elute in the expected order of increasing polarity since the stationary phase is largely aqueous 0.1 N hydrochloric acid. Their absorption maxima (acidic CHC13) lie at 587,592,573, and 568 nm, respectively. Besides the abundant cation peaks in their PDMS, the spectra of the first eluting three dyes show an ion corresponding to the cation minus methylene (Figure 2). This peak is unlikely to result from methylene loss from the cation, and we ascribe its presence to hydrogenation, possibly by hydride ion, of the easily reduced C=N+ linkage during the complex condensed state ionization process. Subsequent loss of a
The remaining peaks were likewise collected but their mass spectra suggest that they are quite different in constitution from the simple triarylmethanes (Figure 3); no structures are assigned at this time. Besides the excellent separations achieved here for relatively minor structural changes, we feel that the high-speed countercurrent chromatography technique has much to recommend its use in the preparation of samples for PDMS and other mass spectrometric methods, viz., relatively small volumes of ordinary solvents are employed, the sample comes in contact only with a continuous 1 mm i.d. Teflon tube, extensive searching for optimum chromatographic conditions is usually not necessary (partition coefficients between 2.0 and 0.5 are effective) and quantities approaching phase saturation may be separated.
ACKNOWLEDGMENT The authors wish to thank Yochiro Ito (Laboratory of
378
Anal. Chem. 1985, 57,378-379
Technical Development, National Heart, Lung, and Blood Institute) for helpful discussions.
(6) Macfarlane, R. D. Anal. Chem. 1083, 55, 1247A-1264A and refer-
Registry No. 1 (RI-R6 = Me3, H3), 93454-66-1;1 (R1-R6 = Me4,H2),93454-67-2; 1 (R1-R5 = Me, R6 = H), 603-47-4; 1 (R1-& = Me), 548-62-9; 2, 93454-65-0; methyl violet 2B, 8004-87-3.
Henry M. Fales* Lewis K. Pannell Edward A. Sokoloski
LITERATURE CITED (1) Conn, H. J. “Biological Stains”; Williams & Williams Co.: Baltimore, MD, 1961; pp 106, 146. (2) Hiranaka, P. K.: Kleinman, L. M.; Sokoloski, E. A,; Fales, H. M. Am. J. Hosp. Pharm. 1075, 32, 926-930. (3) Pannell, L.; Sokoloski, E. A.; Fales, H. M.; Tate, R . submltted for publication in Anal. Chem. (4) Scheifers, S. M.; Verma, S.; Cooks, R. G. Anal. Chem. 1083, 55,
2260-2266. (5) Ito, Y.; Conway, W. D. Anal. Chem. 1084, 56, 534A-554A.
ences therein.
Laboratory of Chemistry National Heart, Lung, and Blood Institute Bethesda, Maryland 20205
Peter Carmeci PC, Inc. 11805 Kim Place Potomac, Maryland 20854 RECEIVED for review September 5, 1984. Accepted October 23, 1984.
AIDS FOR ANALYTICAL CHEMISTS Considerations for Polishing Glassy Carbon to a Scratch-Free Finish Duane E. Weisshaar and Theodore Kuwana*
Department of Chemistry, T h e Ohio State University, Columbus, Ohio 43210 Currently a major research endeavor in this laboratory is the correlation of electrochemical behavior with the physical properties of glassy carbon electrode surfaces (1-6). A primary concern in these studies is the preparation of highly polished, reproducible surfaces. Until recently the usual procedure was to polish the electrode with successively smaller sizes of alumina on a polishing pad or cloth which produced scratch-free surfaces. It has been observed in this laboratory that electrodes polished on such pads tend to be less active electrochemically (at least for some reactions) than electrodes polished on a glass plate (6). However, polishing electrodes on a glass plate with a slurry of alumina and water often results in visible scratching of the electrode surface. A severe example of this scratching is depicted in Figure 1. This scratching is caused by large agglomerates of alumina that are inherent in the alumina preparation process (7,8). Recently, so-called, “deagglomerated” aluminas have become available. These aluminas are produced by an alternate process that minimizes the formation of agglomerates during production (8). We have observed that when Ni is electrodeposited on glassy carbon, the Ni tends to deposit along scratches (5). For our purposes it is imperative that the electrode surfaces be free of such scratches. A comparison of the polishing ability of deagglomerated aluminas from two different sources and the “regular” grade aluminas was undertaken to find a suitable source of alumina that would minimize this problem. The purpose of this paper is to pass the results of this comparison to others who may have similar polishing requirements. The polishing procedure described below is used routinely in this laboratory for the preparation of metal oxide dispersed electrodes (1-3) and polyvinylacetic acid coated electrodes (4-5). Research toward a fundamental understanding of the electrochemical behavior of these electrodes is continuing and will be published separately.
EXPERIMENTAL SECTION A ground glass plate (Barnes Analytical Division, Stamford, CT) was polished on a padded lapping wheel (Ecomet 111,Buehler,
Table I. Alumina Samples TestedQ crystal
particle size, Mm
structure
1.0
C Y
0.3
a Y
0.05
source Ab
source Bb
R, S, D R, S,D R, S, D
D, s D D
R, regular; S, suspension; D, deagglomerated. Sources available on request.
Ldt., Lake Bluff, IL) for 15 min each with 600 grit aluminum oxide and 1.0 km alumina. The pad was thoroughly flushed with water between steps. This process did not produce a smooth surface, but the roughness was reduced drastically. The plate, as received, was quite rough and polishing on it without prior smoothing of the plate produced a glassy carbon surface with a high density of scratches. Three GC-10 Tokai glassy carbon electrodes (Tokai Carbon Co., Ltd., Tokyo Japan) of approximately 1 cm2 were used to evaluate the polishing ability of the various samples of LO-, 0.3-, and 0.05-pm alumina listed in Table I. Each electrode was polished in a slurry of alumina and double distilled water on the plate using a circular motion with relatively heavy hand pressure for approximately 30 s. A fresh aliquot of alumina was used to polish each electrode after which the polishing plate was rinsed with distilled water and wiped with a tissue. The alumina suspensions were diluted with double distilled water to give a slurry of about the same consistency as with the powders. The relatively heavy hand pressure was necessary to break up loose agglomerates that otherwise would scratch the surface. All photomicrographs were obtained with a Model PME-100-FF Olympus Tokyo Inverted Bench Model Metallographic Optic Microscope (LECO Corp., St. Joseph, MI). The alumina sample code that will be used here has the format: type-source. The types of alumina employed in this study were: regular (R), suspension (S), and deagglomerated (D). The two sources of alumina will be referred to as A and B. (The sources will be provided upon request.) Thus, S-B will refer to an alumina suspension from source B. The term “visible”, as used herein, refers to artifacts that are visible to the naked eye such as the scratches in Figure 1.
0003-2700/65/0357-0378$01.50/00 1984 American Chemlcal Society