504
Anal. Chem. 1984, 56,504-510
(23) Saeman, J. J.; Moore, W. E.; Mitchell, R. L.;Mlllett, M. A. Tappl 1854, 37, 336. (24) Dawson, R.; Mopper, K. Anal. Biochem. 1878, 8 4 , 166-190. (25) Hodge, J. E. J . Agric. Food Chem. 1853, I , 926-943. (26) Dutton. 0.G. S. Advan. Carbohyd. Chem. Biochem. 1873, 30, 38-40. (27) Percival, E.; McDowell, R. H. “Chemistry and Enzymology of Marine Algal Polysaccharides”; Academic Press: New York and London, 1967; pp 1-10, 64-96, 190-194. (28) Cowle, G. L.; Hedges, J. I., submltted for publication in Geochlm.
.
Cosmochlm Acta.
RECEIVED for review July 25, 1983. Accepted November 7 , 1983. This research was funded by National Science Foundation Grants OCE-8023970 and OCE-8219294. This is Contribution No. 1353 from the School of Oceanography, University of Washington, Seattle, WA.
Synthesis and Characterization of Polymeric C,*Stationary Phases for Liquid Chromatography Lane C. Sander* and Stephen A. Wise
Organic Analytical Research Division, Center for Analytical Chemistry, National Bureau of Standards, Washington, D.C. 20234
The synthesls of monomerlc, polymeric, and “ollgomerlc” C,( alkyl phases Is described for a serles of wide pore (300 A) siilca substrates. Chrornatographlc propertles of the phases are compared by use of polycyclic aromatic hydrocarbon (PAH) probes. A three-component test mlxture was used to evaluate the relatlve polymerlc nature of a glven phase. On the basis of the elutlon order of the components of thls mixture, the phase type could be classlfled rapldly and the selectlvlty toward more complex PAH mlxtures could be predlcted. Selectlvlty was observed to be related to surface coverage values while absolute retention was found to be more closely related to the total carbon contalned wlthln the column. Although In past work Intentional polymerlzatlon has usually been avoided In the preparations of alkyl-bonded phases, the unique selectlvlty of polymerlc phases makes them an excellent complement to monomeric phases.
To a large extent, the chromatographic properties of a reversed-phase sorbent are dependent on the conditions of the bonded phase synthesis. Understandably, the preparation and study of alkyl bonded phase materials have received considerable attention in the literature (1-10). A large portion of this work has been concerned with monomeric bonded phases. Monomeric phases result from the reaction of monofunctional silane reagents with silanol sites at the silica surface. Since only one bond is formed per silane molecule, monomeric surface modification results. Di- and trifunctional silane reagents may also be used to produce monomeric phases, if precautions are taken to exclude water. When water is not excluded, silane hydrolysis and polymerization are possible, forming a polymeric bonded phase. Although much research has been performed on monomeric phases, relatively little effort has been expended in the study of polymeric phases. The reluctance of workers to accept polymeric phases is probably the result of difficulties reported on early pellicular materials. Problems have included low column efficiency due to mass transfer limitations (11-13), poor peak shape (14),and difficulties with reproducibility from one synthesis to the next (15). Recently, Verzele and Mussche (16) synthesized a series of CISpolymeric phases on totally porous silica. They concluded that most polymeric phases
with surface coverages less than 3.5 kmol/m2 are actually monomeric in nature, but differences in polarity between the phases do exist. Other kinds of polymeric phases have also been produced for use in size exclusion chromatography (17) and ligand exchange chromatography (18). From a theoretical point of view, polymeric phases represent a more complex system than monomeric phases. Monomeric bonded phases are often described in picturesque terms such as “bristles”, “brushes”, or “molecular fur”. Polymeric phases are difficult to visualize because little is known about the extent of cross-linking and the degree of polymerization of the alkyl chains. The effect of polymerization on retention processes is not well understood, although selectivity differences among monomeric and polymeric alkyl sorbents have been reported (19-21). The aim of this study was to investigate the preparation of polymeric bonded phases and to examine some of the unique properties of these sorbents. A series of CISmonomeric and polymeric phases have been produced, as well as an intermediate class of phases which we designate as “oligomeric phases”. In addition, a simple empirical LC test is described for determining the relative polymeric nature of a CISbonded phase.
EXPERIMENTAL SECTION Reagents. Silane reagents were purchased from Petrarch Systems, Inc. (Bristol, PA), and were used without further purification. Chromatographic grade solvents were used in all syntheses, wash procedures, and LC separations. The following silica materials were used in preparation of the bonded phases: 8-pm Zorbax 300 silica (E.I. du Pont de Nemours and Co., Wilmington, DE), 5-pm Hypersil WP-300 (Shandon Southern Instruments, Sewickley,PA), 10-pmVydac TP silica (Separations Group, Hesperia, CA), 10-pm LiChrospher 300 (MCB, Gillstown, NJ), and 10-pm Protosil300 silica (Whatman Chemical Separations Inc., Clifton, NJ) (see Table I). A 16-component PAH mixture, SRM 1647 (National Bureau of Standards, Washington, DC), was used to evaluate the columns. Phenanthro[3,4-c]phenanthrene and benzo[a]pyrene were obtained from Aldrich Chemical Co. (Milwaukee, WI) and 1,2:3,4:5,6:7&tetrabenzonaphthalene (dibenzo[g,p]chrysene)was obtained from Rutgers (Caatrop-Rauxel,Federal Republic of Germany). Carbon analyses were performed at the Center for Analytical Chemistry using a LECO CS-244 elemental analyzer. Carbon determinations were made on both bonded and unbonded silica substrates and corrections were made for the carbon content of the unbonded silica.
This artlcle not subject to U S . Copyrlght. Published 1964 by the Arnerlcan Chemlcal Society
ANALYTICAL CHEMISTRY, VOL. 56,
1
-Si-OH
(-Si-OH synthesis
-Si-OH
'
s:
8:
Si-0- i-R
synthesis
I
I 0-iI-R fi-O-S,i-RlI tl
0Si.R O-Si-R OSI-R 0-Si-R 0-Si-R 0-Si-R 0-Si-R
Table I. Physical Properties of Silica Substrates
rSi-OH
- S i - 0 - i-R
-Si-OH
rnultisteps
O-)[-R
1
NO. 3, MARCH 1984 505
\
I
' E:
0- i-R
Flgure 1. Oligomeric phase synthesis scheme. Anhydrous slhnization and hydrolysis reactions are sequentlally alternated so that the bonded phase is built up one unlt at a time. The process can be considered a controlled, stepwise polymerization. Three to four measurements were made for each phase; the average standard deviation of these carbon measurements was 0.09%. Bonded Phase Preparation. Prior to aJl silanizationreactions, each silica sample was dried under vacuum at 150 "C for at least 1 h to remove physically adsorbed water. This treatment was important especially for the polymeric syntheses, where phase loadings were ultimately controlled by the quantity of water present in the reaction mixture. Water content of the silica substrates, as received from the manufacturer, ranged from trace levels to a surprisingly high value of 20%. Each of the monomeric, polymeric, and oligomeric reactions was carried out in a constant volume (100 mL) of carbon tetrachloride. The reaction mixtures were refluxed for 4 h, after which the bonded silica was filtered, washed, and air-dried in a sintered glass funnel. To ensure the removal of both polar and nonpolar residues, a range of solvents was used in the wash procedure: carbon tetrachloride, tetrahydrofuran, methanol, 50% methanol/water, methanol, and finally tetrahydrofuran. At each step the silica was suspended in the wash solvent by swirling, rather than stirring, to avoid particle breakage. The silica was usually dry (free flowing) after vacuum filtering for 30 min. Samples for carbon analysis were dried at 150 "C under vacuum. Monomeric Phase Synthesis. Monomeric C18 phases were prepared by exhaustive silanization with dimethyloctadecylchlorosilane. Approximately 3 g of silica was added to a solution of 100 mL of carbon tetrachloride containing a 10-fold excess of silane (based on two bonded groups/nm2). The slurry was refluxed for 4 h, filtered, washed, and dried as described above. Polymeric Phase Syntheses. Polymeric C18 phases were prepared by silanization with octadecyltrichlorosilane in the presence of water. Phase loadings were controlled by altering the quantity of water added. In a typical synthesis,10 mL of the silane was dissolved in 100 mL of CC4. The solution was heated and the dry silica added. Finally, the measured quantity of water was added, and the slurry was refluxed for 4 h. After refluxing, the silica was filtered, washed, and dried in the usual manner. Oligomeric Phase Synthesis. A third type of CISphase was produced by using a controlled, sequential reaction scheme (see Figure 1). This process consist9 of a series of linked reaction steps; each step was essentially a monomeric synthesis. The reaction consisted of four steps: anhydrous reaction with trichlorosilane, filtering and washing, hydrolysis, and drying. This sequence produces a monomeric bonded phase with additional silanol sites for further reaction. The phase is then built up one silane unit at a time by repeating the process. Five phases were produced, using a totalof nine reaction steps. A 17.5-g sample of silica was dried under vacuum at 150 "C for
silica silica A silica B silica C silica D silica E
specific surface area: m2/g 45 60 90
shape spherical spherical spherical spherical irregular
pore %carbon vol: (unbonded mL/g silica) 0.42 0.6 0.66 2.0
0.06 0.10
0.41 0.18 1.52 0.11 Mana Nominal values reported by the manufacturer. ufacturer's surface area value is 250 m2/g;the value listed was determined by BET measurement. 2006 250
'
1h. To the hot silica was added 100 mL of CC14and 15 mL of octadecyltrichlorosilane,and the resulting slurry was refluxed for 4 h. The bonded silica was filtered and immediately washed. Next, the silica was transferred to a flask and again refluxed, this time in 50% tetrahydrofuran/water for 1h. Following this hydrolysis reaction, the silica was filtered, rinsed with THF, and allowed to air-dry. Finally, the silica was dried under vacuum at 150 OC for 1 h. This complete process was repeated a total of nine times. Three-gram samples of silica were removed at steps 1,3,5,7,and 9 for column preparation. Additional 0.3-g samples were removed at each reaction step for carbon analysis. Each of the 3-g silica samples was endcapped by refluxing for 2 h in a mixture of 25 mL of CC14 and 25 mL of hexamethyldisilazane (HMDS). Column Preparation. All columns were prepared by using a stirred slurry reservoir operated at 10000 psi (silica D packed at 3000 psi; silica E at 7000 psi). Acetone was used as the slurry solvent, and columns were packed upward. Phases prepared on silicas B, C, and D were packed into 25 cm X 4.6 mm columns; all other phases were packed in 12.5 cm X 4.6 mm columns. Column efficiencies were typically 40 000 plates/m. Chromatography. All separations were performed with a liquid chromatograph consisting of two reciprocating piston pumps, solvent programer, 20-pL loop injector, and UV detector operated at 254 nm. The separation procedure for the 16-component polycyclic aromatic hydrocarbon mixture (SRM 1647)was performed by using a linear gradient: 40-100% acetonitrile/water over a 30-min period at 2 mL/min. The three-component test mixture was chromatographed isocratically at 85% acetonitrile/ water.
RESULTS AND DISCUSSION Five different base silicas were used in the preparation of the bonded phase sorbents. The physical properties of these substrates vary considerably as shown in Table I; however, all five silicas have 300-Apore diameters. A detailed study haa shown that selectivity differences between monomeric and polymeric phases are most pronounced on wide pore silicas and that polymeric CI8 phases produced on narrow pore substrates exhibit chromatographic properties similar to monomeric ClS phases (22). In addition, several recent studies have compared the separations of PAH on various commercial CIS bonded materials and found that only one material, namely, a polymeric C18 phase on a wide pore silica, was successful in resolving all the compounds of interest (20,21, 23, 24). Pearson et al. reported that wide pore silicas have advantages for the separation of large peptides (25). For these reasons, wide pore silica substrates were used exclusively in this study to evaluate the properties of different stationary phase types. As shown in Table I, the specific surface areas for the various silica substrates range from a low of 45 m2/g to a high of about 250 m2/g. These differences are a result of differences in pore volumes of the silicas (0.42-2.0 mL/g). Of the substrates used, all but one were spherical in shape. Three types of phases were synthesized: monomeric, polymeric, and a third intermediate phase termed "oligomeric". Physical and chromatographic properties of the various phases
506
ANALYTICAL CHEMISTRY, VOL. 58, NO. 3, MARCH 1984
Table 11. Physical and Chromatographic Properties of C,, Bonded Phases column
silica
phase type
1
A C C C C C C
monomeric monomeric oligomeric oligomeric oligomeric oligomeric oligomeric polymeric polymeric polymeric polymeric polymeric polymeric polymeric polymeric polymeric polymeric polymeric endcapped
2 3a 3c 3e
3g
3i
4 5 6 7 8 9
C A
D
5
C
10
B E
11
12a 12b a
C C C
C C
mL of water
carbon
coverage, pmol/m*
a(TBN/ PhPh)
a(TBN/ BaP)
a(PhPh/ BaP)
2.23 3.02 3.29 3.92 3.98 4.01 4.11 4.35 5.13 5.89 6.41 5.34 5.00 5.13 4.73 5.31 5.21 5.21a
1.69 1.69 1.71 1.73 1.72 1.72 1.69 1.64 1.48 1.36 1.43 1.60 1.48 1.48 1.49 1.42 1.43 1.47
2.12 2.00 1.77 1.58 1.47 1.46 1.31
0.2 0.5 1.0 5.0 0.5 0.5 0.5 0.5 0.5 0.5 0.5
2.10 5.42 5.82 6.82 6.92 6.97 7.12 7.49 8.66 9.74 10.47 4.81 16.25 8.66 5.61 19.95 8.77 8.93
1.25 1.18 1.03 0.91 0.86 0.85 0.78 0.68 0.47 0.36 0.37 0.55 0.48 0.47 0.44 0.46 0.43 0.46
%
1.11
0.69 0.49 0.53 0.88 0.71 0.69 0.66 0.65 0.62 0.68
Surface converage value based on unendcapped phase (column 12a).
are summdrized in Table 11. The monomeric syntheses were performed in a conventional manner by using an excess of dimethyloctadecylchlorosilanein carbon tetrachloride. The polymeric syntheses carried out were different than those previomly published (12). In the usual method for polymeric synthesis, water is added directly to the dry silica and the container is sealed. The mixture is allowed to equilibrate over a period of days and then is introduced into a solution containing the silane reagent. This slurry is then refluxed for an extended period. Verzele and Mussche (16)modified this procedure by fist adding the dry silica to dry xylene, and then equilibrating the slurry with a measured quantity of water. The silane reagent was introduced as the final step before silanization. In the current work, the water which was added to initiate polymerization was added directly to a slurry containing the silica, silane reagent, and carbon tetrachloride. Syntheses performed in this manner resulted in phases with reproducible surface coverages. The oligomeric phase synthesis can be thought of as a controlled sequential polymerization (see Figure 1). The maximum number of monomeric Cla unib that can be linked together is equal to the number of reaction steps. Each step is essentially a monomeric synthesis. At the end of each modification step, unreacted chloro groups on bonded silane molecules are hydrolyzed to create new silanols for further modification. Thus, two types of silanols can be distinguished surface silanols from the silica and silane silanols from the bonded silane molecules. Because each reaction step is performed under anhydrous, limiting conditions, only surface silanols are modified in the first reaction step. The C1a chains added in subsequent steps are bonded to silane silanols. It should be noted that cross-linkingreactions are not necessarily excluded in the oligomeric phases. Reaction of the trichlorosilane with silane silanols on adjacent chains is possible. However, since this reaction occurs at the silica surface, as opposed to in solution, cross-linking is probably less favored for oligomeric phases than for polymeric phases, where polymerization may occur in solution. Each of the synthesized bonded phases was characterized by carbon analysis. Background corrections were made for the bonded phases by subtracting the percent carbon determinations of the unbonded silica substrates (see Table I) from the bonded phase values. The carbon loadings listed in Table I1 reflect this correction. In the course of this work, a number of other unmodified silica substrates not listed in Table I were
analyzed for carbon. In the majority of these samples, carbon values were less than 0.2%. Because the percentage of carbon is small in the unmodified silica samples, background corrections are probably unnecessary for bonded phases prepared on high surface area supports. On low surface area supports or on supports with low coverage phases, background corrections may be more important. The calculation of surface coverage (N) values for monomeric phases (2) is straightforward
N (pmol/m2) =
lOT, 1 1200n, - P,(M - 1)
s
(1)
where P, is the percent carbon of the bonded phase, n, is the number of carbons in the bonded silane molecule, M is the molecular weight of the bonded silane molecule, and S is the specific surface area of the unbonded silica in m2/g. For polymeric phases the calculation of surface coverage values using eq 1 requires that an assumption be made about the molecular weight of the bonded phase groups. Since the degree of polymerization is unknown for polymeric phases, surface coverage values in this work were calculated by using CH3(CH2)17Si(OH)20-(mol w t 331) as the representative bonded unit comprising polymeric phases. The oxygen atoms in this molecule are introduced as a result of silane hydrolysis during polymerization or the subsequent wash steps and as such are considered part of the bonded phase rather than the silica surface. Polymeric phases can be envisioned in two ways. The first is a bonded surface with a uniform distribution of alkyl chains. In such a model, it is easy to show that even for very heavily loaded phases, the average degree of polymerization is low. For example, on a silica matrix with a specific surface area of 250 m2/g, a polymeric CIaphase with a 26% carbon loading corresponds to a surface coverage of about 8 wmol/m2. If a value of 4 pmol/m2 is taken as an upper limit for monomeric bonded phases (3), then the average degree of linear polymerization is only two. A second model can be envisioned in which the degree of polymerization is higher in some regions than others. Taken to extremes, such a phase might exist as isolated clumps of C18 polymer. Because the nature of the polymeric surface is unknown, surface coverage values should only be used as bulk measurementsto compare phase loadings for different surface area silica. The use of surface coverage values to calculate interchain distances is probably not jus-
7i PERCENT CARBON 4
1[ n
CARSON
67, 54-
321-
0, 0
0
1
2
3
4
5
8
7
8
I
1
1
2
1
3
I
4
I
5
9
R E A C T I O N STEP NUMBER
Figure 2. Carbon loading plotted as a function of reaction step number for the oligomeric phase synthesis. Circles represent carbon loading after endcapping with hexamethyldlsllazane.
tified for polymeric phases. For monomeric phases, converting surface coverages to interchain distances is justified only if a uniform distribution of alkyl chains (as opposed to isolated patches of chains) exists on the surface. Recently, Lochmiiller and co-workers (26) presented evidence supporting a “patch model” of c18 chain distribution for monomeric phases. Examination of the phases listed in Table I1 reveals that phase loadings are highest for the polymeric phases (4.35-6.41 pmol/m2), intermediate for the oligomeric phases (3.29-4.11 pmol/m2), and lowest for the monomeric phases (2.23-3.02 pmol/m2). The phase loading of the oligomeric phases increases with each reaction step. A plot of percent carbon vs. reaction step number for the oligomeric phases is shown in Figure 2. The open circle data points represent carbon loadings after endcapping with HMDS. Although further modification occurs with each reaction step, increases in loading are minimal after the first three steps. Actually, in all but the first step, the endcapping reactions add more carbon to the bonded phase than do the subsequent ClSreactions. This suggests that silanol groups are available to react with small silane molecules but modification is sterically hindered for the ClSreactions. For example, in step 7 (column 3g) the endcapping reaction resulted in an increase of carbon of 0.25%. If the same number of silanols were modified with octadecyltrichlorosilane, this increase would be about 1.5 % . Instead, the increase from step 7 to step 8 was 0.15%, so only about one out of every ten silanols accessable to HMDS was modified by ODs. The oligomeric phases are probably most like a monomeric phase with densely packed chains. After the first reaction step, all surface silanols accessible to the silane reagent molecules are modified. Hydrolysis produces silane silanols which may be available for further reaction. However, because these newly formed silane silanols are also very near the silica surface, modification is also sterically hindered. Any added alkyl chains would further increase the chain packing and decrease silane silanol accessibility in subsequent steps. With increasing distance from the surface, silane accessibility should increase. However, this increase may be insignificant for a small number of reaction sequences where alkyl chain length is still large compared to the distance from the surface of new silane silanols. This could account for the small but regular increases in carbon observed in columns 3c-3i (steps 3 to 9). As expected, the surface coverage values for the polymeric phases increased with the quantity of water added to the reaction mixture (see Table 11, columns 4-7). The amount of water added was in most cases substantially greater than that used in the study of Verzele and Mussche (16) in which the order of water addition was also different. Figure 3 is a plot of polymeric phase surface coverage as a function of the
volume of water added to the reaction mixture. The most pronounced increases in surface coverage occur for relatively small additions of water. At higher water concentrations, increases in surface coverage are correspondingly less, suggesting a limiting behavior. The effect of changing the order of addition of water in the reaction was examined for silica C. In the first case, water was added to a slurry of the silica and silane dissolved in carbon tetrachloride. In the second case, the same quantity of water was added to the dry silica in a sealed bottle and allowed to equilibrate over a period of 48 h. Synthesis was then carried out normally, except additional water was not added. Surface coverage values for the two bonded phases were nearly identical (4.73 vs. 4.77 pmol/m2). Thus, it appears that the order of addition of water in polymeric phase synthesis has little effect on phase loading. The reproducibility of the syntheses for the polymeric phases was excellent. Columns 5 and 8-11 were prepared under similar reaction conditions using the five different base silicas. Even though the sorbents were of substantially different surface areas, the resulting bonded phases had nearly identical surface coverages, Le., 5.13,5.34, 5.00,4.73, and 5.31 pmol/m2, respectively (relative standard deviation, 4.9%). Differences between these values are probably due to imprecision of the surface area values for the base silicas, since these surface area values are nominal values obtained from the manufacturers. The reproducibility of the polymeric phase synthesis was also examined for a single silica substrate. By use of a different lot number of silica C than that used for columns 5 and 12a, four separate polymeric phases were prepared under similar reaction conditions. Surface coverage values were calculated from carbon loading measurements using the nominal manufacturer’s surface area value of 90 m2/g. As expected, the phase coverage reproducibility was better than that observed for phases synthesized on different substrates. The four phases had surface coverages of 4.84, 4.79, 4.73, and 4.77 pmol/m2 (relative standard deviation 0.96%). A comparison of these values with the surface coverages of columns 5 and 12a, also prepared under similar reaction conditions, indicates the two lots of silica differ in specific surface area. Obviously, accurate measurements of accessible surface area are important where critical comparisons of surface coverage values are made. The retention behavior of the different phases was examined by using polycyclic aromatic hydrocarbons (PAH). PAH were selected as probe molecules since selectivity differences have been reported for these compounds on commercial monomeric and polymeric phases (19-22). Nonpolar compounds were studied exclusively in this work to avoid complications from solute-silanol contributions to retention. Rather dramatic changes in column selectivity are observed
508
ANALYTICAL CHEMISTRY, VOL. 56, NO. 3,MARCH 1984
PHASE TYPE COMPARISON
i
MONOMERIC
PhPh
BOP
POLYMERIC
6
I
1
I
0
5
I
I
I
I
I
.
I
10 15 80 95 30 35 40
,BN
flC +
-
0
5
l0 15
RETENTION TIME (minutes)
Figure 4. Separation of 16 polycyclic aromatic hydrocarbons (SRM 1647)on representative monomeric (Column 2), oligomeric (Column
3g), and polymeric phases (Column 5). Separation of the 16-component mixture was performed by using gradlent elution, 40-100% acetonitrile in water over 30 min at 2 mL/min. The three-component mixture was run isocratically at 85% acetonitrile/water. The elution order of benzo[a Ipyrene (BaP), phenanthro[3,4-~]phenanthrene (PhPh), and 1,2:3,4:5,6:7,8-tetrabenzonaphthalene(TBN) Is indicative of phase type. Component identlfication: (1) naphthalene, (2) acenaphthyiene, (3)indeno[ 1,2,3-cd]pyrene.(4) fluorene, (5) phenanthrene, (6) anthracene, (7) fluoranthene, (8)pyrene, (9) benz[a 1anthrancene, (10) chrysene, (1 1) benzo[b]fiuoranthene,(12) benzo[k]fluoranthene,(13)benzo[a]pyrene, (14) dibenz[a ,h]anthracene, (15) benzo[ghi]perylene, (16) indeno[ 1,2,3-cd]pyrene.
for the three classes of phases examined in this paper. Figure 4 illustrates typical separations of 16 priority pollutant PAH on representative monomeric (column 2), oligomeric (column 3g), and polymeric (column 5) columns. Similar separations were obtained for a large number of homemade and commercial monomeric C18phases not listed in Table 11. In every instance, under the chromatographic conditions employed, three sets of compounds were partially or completely unresolved on the monomeric phases: acenaphthalene and
fluorene, benz[a]anthracene and chrysene, and dibenz[a,h]anthracene, benzokhi]perylene, and indeno[1,2,3-cd]pyrene. The polymeric phases were found to be most selective toward PAH, and in several cases base line resolution of all 16 PAH was possible. In general, selectivity toward PAH increased with increasing surface coverage of the bonded phase. This trend has been observed by Wise and May (23) for a series of commercial polymeric cl8 phases. Columns 6 and 7 were an exception to this trend, in that the lower loaded phase of column 6 (5.89 pmol/m2) produced slightly better separations of the PAH probes than column 7 (6.41 pmol/m2). As might be expected, the oligomeric phases exhibited a selectivity toward PAH which was intermediate to the selectivity of the monomeric and polymeric phases. Regular changes in selectivity as a function of the reaction step were also observed. Column 3a, the first reaction step, behaved essentially as a monomeric phase (e.g., column 2). With the increasing number of cl8 units in the oligomeric phases, separations approached, but did not equal, those obtained from the polymeric phases. Column 3i, the ninth step oligomeric phase, behaved similarly to a low loaded polymeric phase (e.g., column 4). A simple empirical LC test was devised to gauge the relative monomeric or polymeric nature of a phase. The elution order of a three-componentmixture, phenanthr0[3,4-~]phenanthrene (PhPh), 1,2:3,4:5,6:7,8-tetrabenzonaphthalene(TBN), and benzo[a]pyrene (BaP) was strongly dependent on the type of phase and the surface coverage (25). Under mobile phase conditions of 85% acetonitrile/water, monomeric C18 phases on widely differing silica substrates not all listed in Table I1 produced the elution order BaP, PhPh, TBN. On the oligomeric series the three compounds eluted in the order PhPh, BaP, TBN; whereas on the moderately and heavily loaded polymeric phases they eluted in the order PhPh, TBN, BaP. Depending on the elution order of this mixture, new phases could be rapidly screened and column selectivity toward more complex PAH mixtures predicted (see Figure 4). The retention of benzo[a]pyrene changed markedly in relation to PhPh and TBN. The selectivity ratios for these three compounds are listed in Table 11. Since BaP consists of five condensed aromatic rings and both PhPh and TBN contain six condensed rings, it might be expected that BaP would elute before PhPh and TBN in all reversed-phase systems. However, this elution order was observed only for the monomeric phases (columns 1,2, and 3a). For the oligomeric and polymeric phases the retention of BaP relative to PhPh and TBN increased. The unusual retention behavior of these three compounds is probably related to the shape of the molecules. Due to steric hindrance of neighboring aromatic rings, both phenanthro[3,4-c]phenanthreneand 1,2:3,45,67,8-tetrabenzonaphthalene are nonplanar (27). PhPh is helically shaped, while TBN is shaped like a saddle. Benzo[a]pyrene,however, is completely planar. Apparently the planar shape of BaP permits an enhanced interaction of this molecule with the polymeric phases. On the basis of previously reported (19, 23, 25) selectivity differences for a number of PAH on monomeric and polymeric CI8columns, similar changes in elution order will be observed with other mixtures of planar and nonplanar PAH. Thus other test mixtures could be developed by using more widely available compounds (e.g., 9,lO-diphenylanthracene or mquinquephenyl). The effect of solute shape on retention is currently under investigation in our laboratory. Polymer phases of comparable surface coverage prepared on different silica substrates were examined by using the three-compenenttest mixture. Each of these phases exhibited remarkably similar selectivity toward PAH, regardless of the type of silica substrated used (Table 11). Selectivity factors
ANALYTICAL CHEMISTRY, VOL. 56, NO. 3, MARCH 1984
POLYMERIC BONDED PHASES
Table 111. Normalized Retention Values for C,, Bonded Phases k'BaPl
Column
k'Be
3.46 2 1.57 3a 2.11 2.90 3c 3e 2.97 3g 3.04 3i 3.24 4 4.04 5 6.46 6 7.03 7 7.51 8 5.60 9 6.57 5 6.46 10 4.93 11 13.07 12a 6.77 12b 6.27 1
%
c
1.65 0.29 0.36 0.43 0.43 0.44 0.46 0.54 0.75 0.72 0.72 1.16 0.40
0.75 0.88 0.66 0.17 0.70
k'PhPhl
509
ktTBNI
k'phph
%c
k'TBN
4.33 1.85 2.17 2.64 2.55 2.58 2.53 2.75 3.04 2.53 2.78 3.08 3.15 3.04 2.17 6.01 2.91 2.88
2.06 0.34 0.37 0.39 0.37 0.31 0.36 0.37 0.35 0.26 0.27 0.64 0.19 0.35 0.39 0.30 0.33 0.32
7.34 3.14 3.73 4.58 4.37 4.44 4.24 4.48 4.46 3.44 3.98 4.93 4.66 4.46 3.25 8.50 4.20 4.26
%
c
3.50 0.58 0.64 0.64 0.63 0.64 0.60 0.60 0.52 0.35 0.38 1.02 0.29 0.52 0.58 0.43 0.48 0.48
for TBN/PhPh, TBN/BaP, and PhPh/BaP generally varied less than 5% among this group of phases. Column 8, silica A, exhibited slightly higher CY values than expected for the measured phase loadings. Separations of SRM 1647 were also similar for columns 5 and 8-11. Differences, however, were noted in column back pressure among columns produced from the various silica substrates. Columns 9 and 11 (silicas D and E) exhibited higher than normal back pressures. These silicas have large pore volumes (2 and 1.52 mL/g) and may be more fragile than the other substrates (25). The resulting high column back pressures may be due to the formation of fines from particle breakage, or a result of the high carbon loadings of the phases. Although detailed Van Deempter type measurements were not performed on the columns, large differences in column efficiency were not observed between the monomeric and polymeric phases. This is an important observation, since polymer phases have traditionally been considered undesirable due to mass transfer limitations. It is informative to examine k' values normalized to the percent carbon loading of the phase as summarized in Table 111. For phenanthr0[3,4-~]phenanthrene, the normalized retention values are similar for most of the phases listed. Phases produced on substrate A exhibited significantly increased retention compared to phases of similar type and surface coverage on other substrates. However, column selectivity for phases produced on silica A was similar to phases of comparable surface coverage on other substrates. No explanation for this anomalous behavior is as yet apparent. The correlation of retention with carbon loading for the other phases is simply explained in terms of the phase ratio. Retention can be expressed as
k'= K4 where k' is the retention, K is the solute concentration in the stationary phase zone/solute concentration in the mobile phase zone, and 4 is the phase ratio, volume stationary phase zone/volume mobile phase zone. Perhaps just as important as the linear dependence of k' on 4 is the opposite effect-deviations from this relationship. Normalized k' values for benzo[a]pyrene exhibit some correlation to carbon loading, but the relation is not linear. Retention for BaP is greater for the polymeric phases than can be accounted for by a simple phase ratio effect. This suggests that changes in (normalized) retention between monomeric and polymeric phases reflect changes in the thermodynamic quantity K. In other words, the free energy of transfer of the solute between the mobile and stationary
ENDCAPPED
NOT ENDCAPPED
6
i
io
1'5 io RETENTION TIME
i5
io
Flgure 5. Comparlson of endcapped and nonendcapped polymeric phases for the separatlon of 16 PAH (SRM 1647).
phases is different for monomeric and polymeric phases, at least for certain solutes. The effect of endcapping the bonded phases (Le., reacting the bonded phase with hexamethyldisilazane or trimethylchlorosilane) has been investigated in a number of studies (2, 28, 29), but reported results have often been in conflict. Berendsen and de Galan (2) have suggested that for densely bonded monomeric phases, further reaction with trimethylsilane reagents is negligible. Others have demonstrated significant differences in chromatographic properties of endcapped and nonendcapped phases (29). The ability of polymeric CIS phases, endcapped and nonendcapped, to separate nonpolar solutes was studied. A 6-g sample of a polymeric phase was produced on silica C. Half of this silica was used to prepare column 12a, and the other half was endcapped
510
Anal. Chem. 1984, 56,510-517
with HMDS for 2 h. The resulting phase was packed into a column for comparison (column 12b). Only a small increase in carbon loading resulted from the endcappingprocedure (i.e., 0.16%). The effect of endcapping on the separation of the PAH mixture is shown in Figure 5. Differences in selectivity were negligible; however, retention was slightly longer on the unendcapped phase. No effort was made to examine the effects of endcapping on the separation of polar compounds, but more significant differences between the phases would be expected.
CONCLUSIONS The results of this study clearly indicate that significant differences exist in the chromatographic properties of CI8 bonded phases prepared in different ways. In general, column selectivity is directly related to bonded phase surface coverage values while absolute retention is more closely related to the amount of carbon contained within the column (the phase ratio). Selectivity varies in a continuous fashion with bonded phase surface coverage which suggests that differences between the monomeric, oligomeric, and polymeric phases are effectively a matter of degree rather than of some fundamental difference in phase type. Regardless of the origin of these differences, the unique selectivity exhibited by polymeric C18 phases toward polycyclic aromatic hydrocarbons makes them especially suited for the separation of PAH. Polymeric alkyl phases prepared on wide pore substrates represent a useful alternative to monomeric phases and deserve further consideration. ACKNOWLEDGMENT The authors thank Barry Diamondstone (Inorganic Analytical Research Division, NBS) for performing the numerous carbon analyses on the bonded phase sorbents and Richard Stout (E.I. de Pont de Nemours and Co.), Frederic Rabel (Whatman Chemical Separations, Inc.), and Debbie Johnson (Shandon Southern Instruments, Inc.) for providing experimental wide pore silicas for this study. LITERATURE CITED (1) Evans, M. 6.; Dale, A. D.; Llttle, C. J. Chromatographla 1880, 73,
5-10.
(2) Berendsen, G. E.; de Galan, L. J. Liq. Chromatogr. 1878, 7, 561-586. (3) Berendsen, 0. E.; Pikaart, K. A.; de Galan, L. J. Liq. Chromatogr. 1980, 3 , 1437-1464. (4) Engelhardt, H.; Ahr, G. ChfOmatOgf8phla 1881, 74, 227-233. (5) Colin, H.; Gulochon, G. J. Chromatogr. 1877, 747, 289-312. (6) Karch, K.; Sebestian, J.; Haiasz, I. J. Chromatogr. 1878, 722, 3-16. (7) Hennlon, M. C.; Picard, C.; Caude, M. J. Chromafogr. 1878, 766, 21-35. -. __ (8) Boksanyi, L.; Liardon, 0.;Kovats, E. Adv. Colloid Interphase Sci. 1978, 6, 95-137. (9) Unger, K. K.; Becker, N.; Roumeliotis, P. J. Chromatogr. 1878, 725, 115-127. (IO) Roumellotis, P.; Unger, K. K. J. Chromatogr. 1978, 749, 211-224. (11) Klrkland, J. J.; DeStefano, J. J. J. Chromatogr. Sci. 1970, 8 , 309-314. (12) Majors, R. E. J. Chromatogr. Sci. 1974, 72,767-778. (13) Kirkland, J. J. J. ChfOmatOgf. Scl. 1971, 9 , 206-214. (14) Knox, J. H.; Vasvari, 0. J. Chromatogr. 1873, 83, 181-194. (15) Novotny, M.; Bektesh, S. L.; Grohmann, K. L. J. Chromatogr. 1873, 83, 25-30. (16) Verzele, M.; Mussche, P. J. Chromatogr. 1083, 254, 117-122. (17) Herman, D. P.; Field, L. R.; Abbott, S . J. Chromatogr. Sci. 1881, 79, 470-476. (18) Gimpei, M.; Unger, K. Chromatographla 1083, 77, 200-204. (19) Wlse, S. A.; Bonnett, W. J.; Guenther, F. R.; May, W. E. J. Chromatogr. Sei. 1981, 19, 457-465. (20) Ogan, K.; Katz, E. J. Chromatogr. 1980, 788, 115-127. (21) Wise, S. A.; Bonnett. W. J.; May, W. E. “Polynuclear Aromatic Hydrocarbons: Chemistry and Biological Effects”; BjDrseth, A., Dennis, A. J., Eds.; Battelle Press: Columbus, OH, 1980; pp 791-806. (22) Sander, L. C.; Wise, S.A., National Bureau of Standards, unpublished work, 1983. (23) Wise, S. A.; May, W. E. Anal. Chem. 1083, 55, 1479-1485. (24) Amos, R. J. Chromatogr. 1081, 204, 469-478. (25) Pearson, J. D.; Lln, T. N.; Regnler, F. E. Anal. Blochem. 1882, 724, 217-230. (26) Lochmulier, C. H.; Colborn, A. S.;Hunnlcutt, M. L.; Harris, J. M. Anal. Chem. 1883, 55, 1344-1348. (27) Clar, E. “Polycyclic Aromatic Hydrocarbons”; Academic Press: London, 1964. (28) Lochmuller, C. H.; Marshall, D. B. Anal. Chim. Acta 1882, 142, 63-72. (29) Llttle, C. J.; Whatley, J. A.; Dale, A. D.; Evans, M. 8. J. Chromatogr. 1878, 777, 435-438.
RECEIVED for review September 30,1983. Accepted November 28, 1983. Certain commercial equipment, instruments, or materials are identified in this report to specify adequately the experimental procedure. Such identification does not imply recommendation or endorsement by the National Bureau of Standards, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.
Molecular Characterization and Profile Identifications of Vanadyl Compounds in Heavy Crude Petroleums by Liquid Chromatography/Graphite Furnace Atomic Absorption Spectrometry Richard H. Fish* and John J. Komlenic Lawrence Berkeley Laboratory, 70-110A, University of California, Berkeley, California 94720 Four heavy crude petroleums, Boscan, Cerro Negro, WIImlngton, and Prudhoe Bay, have been examined by hlghperformance llquld chromatography In comblnatlon wlth graphlte furnace atomlc absorption detectlon (HPLC/GFAA) to provide both a vanadlum flngerprlnt and a molecular weight categorlzatlon of the vanadyl porphyrln and non-porphyrln compounds present. We have also attempted to speclate the vanadyl porphyrln and non-porphyrln compounds In these heavy crude oils by comparlson of their slze excluslon and polar amlnocyano separated vanadlum hlstograms to authentic compounds. 0003-2700/84/0358-0510$01.50/0
One of the most interesting areas of petroleum research has been studies directed toward the isolation and identification of vanadyl compounds in heavy crude petroleums. In the majority of these studies, the focus has been on the vanadyl porphyrin compounds (1-6) and relatively little as yet is known about the vanadyl non-porphyrin compounds, even though they comprise from 50% to 80% of the total vanadium present (7-12). The problems involved in attempting to identify or speciate metalloorganic complexes of vanadyl ion (V02+) in crude petroleums are multidimensional. These complex matrices 0 1984 American Chemical Society