the use of smaller particle size resin and more efficient stirring. The present work has revealed the occurrence of ion-pair sorption of the hydrochloride salts of organic ammonium ions on Amberlite XAD-2 resin. The detailed mechanisms of this sorption and its liquid chromatographic application are also under investigation.
LITERATURE CITED (1) G.Ibrahim, S.Andryanskas, and M. L. Bastos, J. Chromatogr., 108, 107 (1975). (2) A. K. Burnham, G. V. Calder, J. S. Fritz, G. A. Junk, H. J. Svec, and R. Willis, Anal. Chem., 44, 139 (1972). (3) M. D. Grieser and D. J. Pietrzyk, Anal. Chem., 45, 1348 (1973). (4) C.-H. Chu and D. J. Pietrzyk, Anal. Chem., 46, 330 (1974). (5) F. F. Cantwell, Anal. Chem., 48, 1856 (1976). (6) D. Dyrssen, Svensk. Kem. Tidskr., 64, 213 (1952). (7) J. A. Christiansen, Acta Chem. Scand., 16, 2363 (1963). (8) D. Ratajewics and 2. Ratajewics, Chem. Anal. (Warsaw),16, 1299 (1971). (9) M. P. Komar, Ind. Lab. ( U . S . S . R . ) ,34, 617 (1968). (IO) H. J. C. Tendeloo, A. E. Mans, and G. Dehoogh, Red. Trav. Chim. Pays-Bas, 67, 395 (1948). (1 1) H. J. C. Tendeloo, A. E. Mans, and G. Dehoogh, Red. Trav. Chim. Pays-Bas, 68, 253 (1949). (12) A. E. Mans and G. J. Vervelde, Recl. Trav. Chim. Pays-Bas, 7 1 , 977 (1952). (13) Z.Stolkova, Lisfy Cukrov., 67, 207 (1951). (14) F. F. Cantwell and D. J. Pietrzyk, Anal. Chem., 46, 344 (1974). (15) F. F. Cantwell and D. J. Pietrzyk, Anal. Chem., 46, 1450 (1974). (16) F. F. Cantwell, Ph.D. Thesis, University of Iowa, December 1972. (17) J. Kielland, J . Am. Chem. Soc., 59, 1675 (1937). (18) R. A. Robinson and R. H. Stokes, “Electrolyte Solutions”, 2nd ed., Butterworths, London, 1959. (19) C. F. Hiskey and F. F. Cantwell, J . Pharm. Sci., 57, 2105 (1968). (20) L. F. Blackwell, A. Fischer, I.J. Miller, R. D. Topsom, and J. Vaughan, J . Chem. SOC., 1864, 3588. (21) H. M. McNair and E. J. Bonelli, “Basic Gas Chromatography”, Varian Associates, Palo Alto, Calif., 1968,p 52.
:L5 30
nOHx
io4
Figure 6. Experimental titration curves of rn-nitrophenol with sodium hydroxide in 50 mL of 0.10 M sodium bromide in the presence (0)and absence (0)of 5.0 g of Amberlite XAD-2 resin
ion pair Q’A- is strongly sorbed by the resin.
CONCLUSIONS Heterogeneous titrations in the presence of XAD-2 permit the accurate determination of acids of the BH’ charge type which are too weak to be titrated in a homogeneous aqueous solution. Such titrations show promise as an alternative to nonaqueous titrations. The opposite sign of the pH shift (ApH) of HA and BH’ charge-type acids should make possible differentiating titrations of acids with similar strength but different charge type. A study of this application is under way. The principal limitation to the method is the relatively long equilibration times required after the addition of each increment of titrant. This time can be reduced, however, by
I
RECEIVED for review February 4,1977. Accepted May 2,1977. This work was supported by the National Research Council of Canada and the University of Alberta.
CORRESPONDENCE
Preparation and Low-Voltage Mass Spectrometry Sensitivities of Methylated Polynuclear Aromatic Hydrocarbons
.
Sir: At the Grand Forks Energy Research Center (GFERC), the liquefaction of low-rank coal by the CO-Steam process ( I ) is being studied. In this procedure, lignite or subbituminous coal is reacted with synthesis gas at nominally 450 “C and 4000 psi. At GFERC, the composition of the oil produced by this process is being studied in detail using mass spectrometry as the primary method. Mass spectrometry as a quantitative tool for analysis of oils and petroleum distillates has been utilized for over 25 years. The analysis of aromatic hydrocarbon mixtures by low voltage mass spectrometry is widely employed by the petroleum industry (2, 3) and is now finding application to analysis of coal liquids ( 4 , 5 ) . In order to obtain accurate measurement of concentrations in mixtures of aromatic and heteroaromatic compounds, sensitivities must be known. The work by Lumpkin and Aczel (6) allowed the prediction of sensitivities for many aromatic compounds which were not available from any source, and the decrease in sensitivity upon alkylation was measured for many compound types. The observed lowering of sensitivity seemed reasonable, because the addition of one or two long side chains increased the molecular weight, 1280
ANALYTICAL CHEMISTRY, VOL. 49, NO. 8 , JULY 1977
but did not significantly lower the ionization energy. However, most aromatics in coal liquids are either unalkylated or contain 1 or 2 methyl groups. The low voltage mass spectrometry sensitivities of these compounds, therefore, are not predictable on the basis of the work of Lumpkin and Aczel. Very few of the compounds needed for sensitivity measurement are available commercially. The parent polynuclear aromatics (PNA) can be purchased from a variety of sources, but methyl and dimethyl derivatives of these aromatics are not available commercially. A rapid simple procedure (7) has been reported for synthesis of partially hydrogenated PNAs and partially hydrogenated heterocyclics. The synthesis of standard compounds is generally avoided because the syntheses are often time consuming and laborious. The methylation of PNAs could be difficult because of the high reactivity of the aromatic substrate. An ordinary Friedel-Crafts reaction using methyl halides or alcohols is not a satisfactory procedure, since polymerization of the polynuclear aromatics occurs in the presence of an acid catalyst. In contrast, a two-step synthesis for methylation of PNAs is described here. Bromomethylation is effected in the first step
Table I. Summary of PNAs Reacted, Reaction Conditions, and Products Formed
+_
Reaction time mmol PNA reacted ( 5 mmol)
mmol paraformaldehyde
Acenaphthene Acenaphthene Fluorene Fluoranthqne Chrysene
25 3 15 15 15
at
of
25
PNA unreacted
Range of methylation products obtained
0 0.2 0.1 2.5 4.5
di-hexa mono-tri mono-hexa mono-tri mono-di
"C
(h) 2 0.5 2 2
16
Table 11. Low Voltage MS Sensitivity Factors, h f , Determined for Methylated PNAs (Values are 5% relative)
using hydrogen bromide and trioxymethylene in acetic acid. Without isolating the intermediate, reduction to the hydrocarbon is performed with zinc dust.
EXPERIMENTAL Synthesis of Methylated PNAs. Ten mmol of the PNA (Aldrich or Eastman) was weighed into a 250-mL flask and dissolved ih 20 mL of glacial acetic acid. Twenty mL of 32% (w/v) hydrobromic acid in acetic acid (Fisher Scientific Co.) was introduced, and with stirring, 5-15 mml of trioxymethylene was added. The quantity of trioxymethylene depended on the reactivity of the PNA toward bromomethylation and on the degreee of methylation desired (see Table I). The reaction mixture was stirred magnetically for 0.5 to 16 h, and at the end of the desired reaction time, an additional 5 mL of hydrobromic acid in acetic acid was added. Zinc dust was charged in small portions until vigorous evolution of hydrogen had ceased, then an additional quantity of zinc dust (0.5 to 1.0 g) was added to ensure complete reduction of the bromomethyl aromatic. The reaction mixture was poured into 200 mL of distilled water and extracted with three 50-mL portions of toluene. Toluene was removed from the reaction product using a rotary evaporator, and pure compounds were isolated by preparative GC. The chromatograph was a Varian Aerograph Model 2740 equipped with a 6-ft by '/4-inch o.d., 4-mm i.d. glass column packed with 3% SE 30 on Varaport 30. The oven temperature was programmed from 12e150 "C up to 250-275 "C at 10 "C/min depending on which PNA reaction mixture was separated. Measurement of MS Sensitivities. The compounds isolated by preparative GC were recombined to give a mixture containing approximately equal concentrations of the parent and each homologue. This mixture was then quantitatively analyzed using a Varian Aerograph Model 2740 gas chromatograph with a FID and a Spectra Physics System I integrator. The relative concentration of each homologue was then calculated. After measurement of concentrations, the mixture was introduced into the mass spectrometer using a heated glass inlet at 325 "C. The mass spectrometer was an AEI MS 30 (single beam) with a DS 50 data system. The voltage used for low voltage work in studies at GFERC has been 10 V, therefore sensitivities were measured at this ionizing voltage. The repeller voltage was zero. A t least six spectra were recorded using the data system, and intensities at each mass were averaged over the spectra chosen. The sensitivities of the methyl homologues were calculated relative to the parent hydrocarbon designated as having unit sensitivity. Isomers were assumed to have the same low-voltage sensitivity (6).
RESULTS AND DISCUSSION Table I summarizes the compounds in this study, their relative reactivities, and the products formed from each. The reactions proceeded smoothly with minimum side reactions except for fluorene, pyrene, and phenanthrene. Pyrene and phenanthrene were quite unreactive in the system described, partly because of low solubility. Fluorene formed dimers and alkylated dimers; mass spectrometric analysis of the unseparated reaction mixture showed peaks at about double the molecular weight observed for the methylated fluorenes. Chrysene gave only a low yield of products.
Compound Acenaphthene parent monomethyl dimethyl trime thyl tetramethyl pentame thyl hexamethyl Fluorene parent monomethyl dimethyl trimethyl tetramethyl pen tame thyl hexame thyl Fluoranthene parent monome thyl dimethyl trime thyl Chrysene parent monomethyl
kf
hb
1.00 0.96 1.01 0.96 0.99 1.05 1.10
1.12 1.08
1.00 0.85 0.70
0.62 0.59
0.60 0.66
1.13 1.08 1.11 1.18 1.23 1.18 1-00 0.83 0.73 0.70 0.71
0.78
1.00 0.97 0.99 0.87
1.32 1.28 1.31 1.15
1.00
1.19 1.09
0.92
a h f = (C,/C,) x (Ip/Zm), where C, and C, are the concentrations of methylated PNA and parent PNA respectively, and I , and I , are the low voltage ion intensities. kN is the sensitivity factor calculated relative to naphthalene.
Only about one hour of laboratory time was required for performing the preparation. If mixtures were to have been used unseparated, the total time of obtaining the needed compounds would have been minimal, but preparative GC was somewhat time-consuming. Preparative GC provided well defined portions of individual homologues, since the boiling points and the GC retention times were quite different for the respective methyl isomers. In some cases, however, 2 or 3 peaks were observed in the GC separation for a given methyl homologue. In those situations, isolation of individual isomers would have been possible; however, this was not attempted in this study since homologues of a PNA exist in a coal liquid as.a mixture of isomers. The mass spectrometric sensitivities measured in this study are summarized in Table 11. The dependence of low voltage mass spectrometric sensitivity on alkylation (6) for a single long alkyl chain to a PNA does not hold true for methyl groups. In fact, most data previously obtained for sensitivities for mono- and dimethyl homologues of aromatic compounds showed a slight augmentation of the sensitivity. This is found to be the case for the compounds studied. Sensitivity either increases slightly or remains constant with the addition of methyl groups. There are two effects which cause a change in sensitivity upon alkylation of a PNA. With the increase in molecular weight, fewer moles would be present in a given weight of sample, thus indicating a lower sensitivity per gram. Concurrently, the addition of methyl groups would lower the ionization energy and thereby increase sensitivity. It appears that for addition of methyl groups, the decrease in ionization energy is slightly more important than the increase in molecular weight. In order to determine the relative weights of the parent PNA and its methyl homologues in the MS analysis mixture, a gas chromatographic technique was employed. The sensitivities in gas chromatography determination for the parent compound and its methyl derivatives were assumed to be ANALYTICAL CHEMISTRY, VOL. 49, NO. 8, JULY 1977
1261
identical. Other studies have found FID responses to be essentially identical for aromatics and alkyl aromatics (8). For example, naphthalene, methylnapthalene, and dimethylnaphthalene, dissolved in toluene to a known concentration, showed GC peak areas per gram to be identical within experimental error of 1-270. Therefore, quantitative GC was chosen to be the most accurate available technique to determine concentrations in LVMS analytical solutions. Previously published GC sensitivities using a flame ionization detector (9) reported a difference between PNAs, their methyl derivatives, and their hydro derivatives. Data in this study showed this not to be representative of compounds which are completely soluble in the solvent used for GC injection. The reported change in sensitivity with methylation included differences in behavior in the separation scheme used. Methylated PNAs were seemingly less sensitive than the parent PNA, but this decrease in sensitivity was due to differences in solubility rather than to a change in true detector response characteristics.
ACKNOWLEDGMENT The author thanks David J. Miller who performed some of the experimental work reported.
LITERATURE CITED H. R. Appell and I. Wender, Prepr., Div. Fuel Chem., Am. Chem. SOC., 12, (3),220 (1968). B. H. Johnson, and T. Aczel, Anal. Chem., 39,682 (1967). A. W. Peters, and J. G. Bendaraitis, Anal. Chem., 48, 968 (1976). T. Aczel, J. Q. Foster, and J. H. Karchmer, Prepr., Div. Fuel Chem., Am. Chem. Soc., 13, (l),8 (1969). J. L. Schuttz, R. A. Fiedel, and A. G. Sharkey, Jr.. U.S.Bur. Mines Rept. R I 7000 (1967). H. E. Lumpkin and T. Aczel, Anal. Chem., 36, 181 (1964). J. E. Schiller and D. H. Neal, “Preparation of Hydroaromatic Compounds for Mass Spectrometry Standards in Coal Conversion Analysis”, ERDA Rept. Invest., GFERC/RI-75/2 (1975). J. C. Sternber, W. S. Galloway, and D. L. T. Jones, “Gas Chromatography”, Academic Press, Inc., New York, 1962,Chap. 18. R . C. Lao, R. S. Thomas, H. Oja, and L. Dubois, Anal. Chem., 45,908
(1973).
JoseDh E. Schiller Grand Forks Energy Research Center Grand Forks, North Dakota 58202
RECEIVED for review February 10,1977. Accepted May 2,1977. Reference to specific manufacturers, their brands, or models is for identification only and does not represent endorsement by ERDA.
Analytical Possibilities of Phase Resolved Phosphorimetry Sir: The phase characteristics of several organic phosphors have been investigated as the basis of an analytical technique ( I ) . The sample is excited by a continuum source, the intensity of which is modulated, and the phase of the phosphorescence emission is compared with that of the exciting light. The phase shift (0) of the phosphorescence emission signal is a function of the frequency of modulation and the lifetime of the phosphorescence. By measuring the phase shift, phosphorescence lifetimes can be calculated from the basic equation, 0 = tan-’ ( w ~ T )where T is the phosphorescence lifetime and wJ is the angular frequency, usually the fundamental frequency, wl. The same theory applies to fluorescence, and a review by Birks and Munro (2) covers phase resolution via fluorimetry. Because the instrumentation and its operation are fairly simple in phase resolved phosphorimetry, efforts were made in this laboratory to extend the analytical capabilities of this technique with the hope of doing routine analysis of drugs. In the present work, minor modifications were carried out on the system used in the earlier work cited (1). EXPERIMENTAL Apparatus. The experimental setup used in this study was similar to that of Mousa and Winefordner ( I ) . The source of excitation was an Eimac 150-W lamp whose intensity was modulated mechanically rather than electronically as was previously done. The intensity of this lamp was far greater than the one used in prior work because of the use of an integral parabolic reflector. The lamp was powered by a Varian Illuminator Power Supply kept at 12 A. A block diagram of the instrumental system is shown in Figure 1 (Comments are described on the caption). Reagents. Benzophenone (Fisher Scientific Co., Fair Lawn, N.J.), 4-bromobiphenyl (Pfaltz and Bauer, Flushing, N.Y.), morphine, and codeine (Applied Science Labs., Inc., State College, Pa.), were all used as received. The solvent used in all cases was ethanol (U.S. Industrial Chemicals, Co., New York, N.Y.) which was made anhydrous using the method of Lund and Bjerrum ( 3 ) . Procedure. The operational procedure used was identical to the one described by Mousa and Winefordner (1). 1262
ANALYTICAL CHEMISTRY, VOL. 49, NO. 8 , JULY 1977
Table I. Phosphorescent Lifetimes from Phase Data Lifetimes, ms Phase Phasea (Ref. Molecule (this work) 1) Time resolvedb 5.9 I Benzophenone 6‘ ( 2 5 Hz) 5.3d ( 2 5 Hz) 4-Bromobiphenyl 17‘ (10 Hz) 14 17 32‘ (10 Hz) --39 Codeine a Values in parentheses are the optimum modulation frequencies in each case. The capillary cell for all lifetime Data were taken from measurements was not spun. Ref. 4-6; Glassy matrix. Snowy matrix.
RESULTS AND DISCUSSION Quantitative analysis and measurement of lifetimes of benzophenone and 4-bromobiphenyl in a clear-glassy matrix of pure ethanol were readily carried out with the experimental system described. The results obtained indicate good agreement with other published works as shown in Table I. It appears that phase resolved phosphorimetry would be useful for the analysis of these two compounds where the slope of the log-log analytical curve is close to unity and the range of linearity is fairly large ( lo3). The phosphorinietric limits of detection obtained were: 0.1 ppb (5 X lo-’’ M) for benzophenone; and 2 ppb (7 X lo-’ M) for 4-bromobiphenyl. The detection limit is defined as that concentration of analyte resulting in a signal-to-noise ratio of 3. The same procedure was carried out on benzophenone in a 10% ethanolfwater mixture which produced a snowy matrix to avoid the tedious handling of a clear-glassy matrix because formation of such a matrix sometimes proved to be a nuisance. The measured lifetime of benzophenone in this snowy-matrix (capillary cell not spinned) agreed quite well with that found in a clear glassy matrix (see Table I), but no quantitative analysis could be carried out. The inhomogeneity and opaqueness of the snowy N
