No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Free release static coating of glass capillary columns; Static coating at elevated pressure — Theoretical considerations and practice
Abstract
Static coating of glass capillary columns has hitherto solely been carried out by evaporation of the stationary phase solvents under vacuum conditions. However, since a solvent vapour pressure higher than the external pressure is the only requirement for the vapour to flow out of the column, evacuation should not be necessary. Several important factors in the static coating procedure, such as mass (solvent vapour) — and heat transfer, heat of vaporisation of the solvent and viscosity of the stationary phase solution at elevated temperature and outlet pressure are discussed, principally to rationalise an improved static coating procedure. The alternative, so-called free-release static coating procedure, was evaluated practically by coating several columns with OV-101 and SE-30. Coating speed was found to be rapid and relatively constant whereas coating efficiency was between 80 and 100%.
... A 58 m glass capillary WCOT column (I. D. 0.30 mm) was used with OV-1701 as stationary phase, which was prepared according to Xu and Vermeulen (1984). Temperatures of injection port, column and detector were 250 ~ 230 ~ and 300 ~ respectively. ...
... A Hewlett Packard 3390A automatic integrator was used and the column was a 21 m (I. D. 0.30 mm) glass capillary WCOT column, coated with OV-1701 (Xu and Vermeulen, 1984). Temperatures of injection port, column and detector were 250 ~ 220 ~ and 300 ~ respectively. ...
Z- and E-1,3-dichloropropene, mutagenic geometric isomers and major constituents of commercial soil fumigants, were found to be metabolized to mercapturic acid derivatives by the rat. Extremely small quantities of mixtures of the parent compounds were administered intraperitoneally to the rat and the isomeric urinary mercapturic acids were quantified in three ways. Gas chromatographic procedures with nitrogen selective, sulphur selective and mass spectrometric detection, using negative chemical ionization with single ion detection, were evaluated with respect to selectivity and sensitivity. Applying the former two techniques, urinary mercapturic acids could still be quantified following 5 micrograms doses of each of the dichloropropene isomers. With gas chromatography-negative chemical ionization mass spectrometry, only mercapturic acid metabolites arising from 25 micrograms doses and higher could be quantified because of interference from endogenous compounds. These results suggest that all three analytical methods can be used to determine exposure of men to soil fumigants containing low levels of 1,3-dichloropropene.
... A 58 m glass capillary WCOT column (I. D. 0.30 mm) was used with OV-1701 as stationary phase, which was prepared according to Xu and Vermeulen (1984). Temperatures of injection port, column and detector were 250 ~ 230 ~ and 300 ~ respectively. ...
... A Hewlett Packard 3390A automatic integrator was used and the column was a 21 m (I. D. 0.30 mm) glass capillary WCOT column, coated with OV-1701 (Xu and Vermeulen, 1984). Temperatures of injection port, column and detector were 250 ~ 220 ~ and 300 ~ respectively. ...
1,3-Dibromopropane (1,3-DBP) was administered i.p. in doses ranging from 5.6 to 54 mg to male Wistar rats. Four different mercapturic acids, viz. N-acetyl-S-3-bromopropyl-(MA I), N-acetyl-S-3-chloropropyl-(MA II), N-acetyl-S-2-carboxyethyl-(MA III) and N-acetyl-S-3-hydroxypropyl(-1-)cysteine (MA IV) were synthesized and identified as metabolites in urine by g.l.c.-mass spectrometry. 1,1,3,3-Tetradeutero-1,3-dibromopropane was used to study the mechanism of formation of the mercapturic acids in more detail. It was found that in the formation of MA IV a reactive episulphonium ion could be involved. Gas chromatographic quantification of the mercapturic acids (mercapturic acid assay) was correlated with a spectrophotometric thioether determination of the metabolites (thioether test). At doses up to 30 mg of 1,3-DBP, excretion of mercapturic acids was virtually complete in 24 h urine and amounted to about 19% of the dose (11.3% MA I, 4.9% MA II, 2.6% MA III and 0.2% MA IV). From excretion rate curves a half-time t1/2 was calculated as being about 4.5 h. A plateau in the dose-excretion curve was observed at 1,3-DBP doses higher than 40 mg, probably caused by glutathione depletion.
... The gas chromatograph was equipped with a splitless injector and nitrogen phosphorous detector. A SE-54 capillary column was prepared according to Xu and Vermeulen [29]. The operation temperatures of gas chromatography were 200" (injector) and 220" (detector). ...
1,1-Dichloro-2,2-difluoroethylene (DCDFE) produced marked nephrotoxicity in rats upon an i.p. dose of 150 mumole/kg. At doses higher than 375 mumole/kg, DCDFE also produced hepatotoxicity. Aminooxyacetic acid, an inhibitor of cysteine conjugate beta-lyase, appeared to be slightly nephrotoxic in Wistar rats. Nevertheless it exerted an inhibitory effect on the nephrotoxicity of DCDFE. The N-acetylcysteine conjugate of DCDFE was identified as a major urinary metabolite of DCDFE. When administered as such, this conjugate appeared to be a potent nephrotoxin, without any effect on the liver, indicating that glutathione conjugation of DCDFE is most likely a bioactivation step for nephrotoxicity. The appearance of traces of chlorodifluoroacetic acid in urine of rats treated with higher doses of DCDFE indicates the existence of an oxidative pathway of metabolism of DCDFE, probably involving epoxidation by hepatic mixed-function oxidases. It is speculated that the latter route might account for the hepatotoxicity at higher doses of DCDFE. The nephro- and hepatotoxicity of DCDFE, therefore, most likely are the result of two different mechanisms of bioactivation.
Suspension Method
Preparation of Adsorbent Layer by Synthesis from Materials of the Inner Capillary WallsPreparation of ALOT Columns by Sorbent Synthesis inside the Capillary ColumnFormation of Adsorption Layer during Glass Capillary DrawingConclusion
References
A wall-coated capillary column provides a relatively simple macroscopic method for measuring intracrystalline diffusion in small (micron-sized) zeolite crystals. We report here a procedure for the preparation of such columns.
Narrow-bore, (5–15 μm), long (up to 6 m), efficient and stable open tubular columns with four different polyorganosiloxane coatings, were prepared by the static coating and the precipitation coating methods. The polysiloxanes were substituted with phenyl- (OV-17-vi and PS-264), methyl- (PS-255), and octadecyl- (PMSC18) groups. They were studied in their regular cross-linked state as well as after swelling with n-heptane. Anthracene derivatives were used as model compounds. Column properties such as retention, selectivity and efficiency were investigated in reversed-phase liquid chromatography. GC was used for complementary studies of 50 μm I.D. columns. The highest retention was obtained on the columns coated with OV-17-vi and on the swelled stationary phases. A considerable resolution improvement was obtained in the columns, after swelling with n-heptane owing to the increase in retention. More than 106 theoretical plates were achieved in 70 min at a k′ of 0.46 in a 6 m ×6 μm I.D. column.
The outstanding properties of fused-silica capillary columns coated with graphitized carbon black impregnated with liquid phases are described together with the details of their preparation, HETP vs. curves are reported, together with some separations of industrial and environmental interest. Fast and more effective analyses may be achieved with such columns.
Dichlorofluoromethane, trichlorofluoromethane and tetramethylsilane were used in the static coating of small diameter capillary columns (5 to 50 μm I.D.) in order to obtain highly efficient columns for gas and supercritical fluid chromatography. Capillary columns of 5-, 10-, 25-, and 50-μm I.D. were coated with stationary phase films of SE-33, SE-54, OV-215, 50% n-octyl, 45% phenoxypolyethyl ether, 50% liquid crystal, 25% biphenyl, 50% pentafluorophenyl and 50% cyanopropyl polysiloxane stationary phases. Resultant evaluations of these columns in gas chromatography gave approximately 9000, 66 000, 45 000, and 19 000 plates m−1, respectively, for the different internal diameters. Important parameters which affect coating efficiency are identified and discussed in detail.
Fused-silica capillary columns coated with a layer of non-porous, homogeneous graphitized carbon black, modified with a variety of stationary phases, were prepared. A series of parameters were determined. The selectivity and efficiency of the columns were much improved. These columns were used in the analysis of complex pyrolysis products and the separation of isomers.
The kinetic and thermodynamic changes induced in capillary columns with the internal area modified by a thin layer of graphitized carbon blacks are discussed and were tested experimentally. The selectivity due to the mechanism of gas—liquid—solid chromatography is greatly increased and the efficiency in terms of HETP is much improved at high linear gas velocities. This results in the possibility of obtaining faster analyses without the usual increase in operating temperature with respect to the gas—liquid technique. Examples of applications in practical analysis are given.
A new, simple procedure to deactivate fused silica capillaries without hydrothermal treatment is proposed. Based on high temperature reactions of octamethylcyclotetrasiloxane (D4) in the presence of oxygen, this procedure results, upon coating with dimethylsilicone (SE-30), in columns which give almost symmetric peaks for as little as 150 pg of critical compounds, such as decylamine, thereby demonstrating very high inertness. A mechanism of the deactivation procedure, which, in contrast to previously developed procedures, is based on thermal reactions under aerobic conditions, is proposed.
A method for preparation of capillary columns and traps with a very thick film (up to 100 μm) of stationary phase is described. The principle of this method is based on an immediate fixation of a film of prepolymer, formed during dynamic coating. Thus, the development of film irregularities, such as are caused by Rayleigh instability is avoided. Fixation of the film is conveniently accomplished by heat-accelerated crosslinking as was demonstrated in this work, where a commercially available silicone prepolymer (Sylgard 184, Dow Corning) was employed.
The low phase ratio columns which thus can be prepared are interesting both in chromatography with dense mobile phases and as enrichment devices. Examples of the latter application are shown, where trace organic components from air and water were concentrated.
The coating speed upon static coating of glass capillary columans was evaluated in terms of inner diameter and length of the column, viscosity and pressure of solvent vapor, etc. From the equation obtained it can be shown that a smaller diameter of a microbore column restricts solvent vapor transfer to the orifice of the column drastically. To compensate for this restriction, a higher pressure at the meniscus is needed. As an alternative to using a higher coating temperature, application of more volatile solvents such as n-butane and isobutane is proposed. Several glass capillary columns (130 μm i.d.) were coated with SE-54 dissolved inpentane-acetone mixed with n-butane or isobutane and the column performances were evaluated. Selection of these solvents permitted free release static coating of ca. 100 μm columns at lower or even ambient temperatures and they were equally suitable as commonly applied solvents (e.g. pentane) to coat highly efficient columns.
In an earlier report the authors recommended pressurizing as a valuable 'repair' technique in case of trouble with breakthrough. The authors now wish to complete this information by recommending application of compressed gas at 4-5 bar for 10-15 min to every freshly filled and closed column. This simple precaution practically eliminates breakthrough effects during pumping.
The technique of spontaneous coating of capillary columns employing liquefied butane and ethylene chloride as solvents of the stationary phase is described. These solvents and their mixtures are suitable for apolar and medium-polar stationary phases. The dependence of coating time on coating temperature and on the composition of the mixture of solvents was observed. It was found that the coating efficiency decreases with increasing coating speed.
The cause of an observed acceleration in the free release static coating process of a capillary when more polar solvents are added to alkanes as stationary phase solvents has been studied. By using a computer program, 9 of the relevant physical parameers were calculated in relation to the composition of the solvent hixtures. Curves of these parameters vs composition of the mixures were drawn. From the results it can be deduced that the change of these parameters with the increase of polar solvents in the solution, is probably insufficient to explain the drastic increase of the free release static coating speed. A physical model is proposed to calculate the depth of the meniscus around which sufficient heat can be transferred into the evaporating surface per unit time, to evaporate the solvent at a rate that the coating speed indicates. It is demonstrated that when pentane-acetone mixtures are used to coat a capillary, the depth of the meniscus indeed is greater than when pure pentane is used. The proposed model may lead to a better understanding of phenomena involved in the static coating process.
Open tubular columns of 25 and 50 μm i.d. have been coated with up to 2 μm thick layers of SE-54 (lowest β-value = 6.25). Filling of the columns and evaporation of the solvent were studied as a function of column length and time. Since the viscosity of the stationary phase solution increases exponentially with concentration, the filling process becomes time consuming and high filling pressures are required. The evaporation or coating speed is shown to be controlled by the resistance to mass transfer of the solvent vapor through the coated sections of the column. For 25 and 50 μm i.d. columns, experimental evaporation times deviate little from theoretical predictions. Almost no influence of the stationary phase on the coating speed was observed. The transition from mass transfer limited coating speed to heat transfer limited coating speed is illustrated using literature data.
This paper describes a high pressure device for filling small diameter capillaries with stationary phase solutions is described. A liquid is forced into the capillary column with the help of high pressure syringe whose needle (provided with a side opening) is tightened in a PTFE seal. The device allows use of liquefied gas as solvent. A detailed procedure is given for filling the capillary with stationary phase solution. The performance of the device was evaluated by filling 12 m × 15 μm i. d. glass capillary with 6.5 % (w/v) SE-54.
The preparation of wide bore (320 μm) and narrow bore (50 μm) fused silica capillary columns is described for immobilized cyanopropyl substituted silicones containing 60 and 88% substitution. The effect of high temperature deactivation with cyanopropylcyclosiloxanes was studied with a special test mixture. Curing was achieved with dicumyl peroxide or azo-tert-butane.
The columns were evaluated and compared in terms of efficiency, activity, polarity, and temperature stability. Different coating methods were compared for the narrow bore columns. The activity of the 60% cyanopropyl columns that had been immobilized with dicumyl peroxide was significantly larger than for azo-tert-butane immobilized columns. The polarity of polar columns appeared to depend greatly on column temperature and is completely different for wide and narrow bore columns.
Static coating of phenyl and biphenyl polysiloxane stationary phases on 50 μm i.d. open tubular capillaries was studied. The influence of coating solvents and coating temperatures on the viscosities and surface tensions of the polymer stationary phases and their coating solutions was determined. A measure of the Rayleigh instability paralled the observed coating efficiencies.
Three aspects with respect to the selection of solvents for static coating of capillary columns, i.e. coating speed, occurrence of bumping and solubility of stationary phases are discussed. Hypotheses are proposed in an attempt to explain the observed facts that mixed solvents result in much higher coating speeds than those obtained from pure solvents, that a proper choice of solvents together with a good and uniform deactivitation of the column inner wall is needed to prevent bumping. Stationary phase solubility vs. solvent composition is also briefly discussed.
A very fast method for continous monitoring of BTEX compounds in different matrices is described. The use of a new graphite-lined
(GLOT) capillary column, allows the complete separation of BTEX at isothermal conditions and very short analysis time. The
same column is also able to separate polar compounds and other VOCs.
By in situ photopolymerization of acrylates, thick polyacrylate films can be immobilized in 10-μm fused-silica capillaries. Up to 1.9-μm thick films can be produced in 0.3–1.2 m long capillaries, resulting in phase ratios > 1.0 and leading to columns with high mass loadability. Owing to the high mass loadability, application of “on-column” UV detection is possible. The effect of incorporating alkyl acrylates in the film on retention behaviour and efficiency was extensively investigated. The success rate of the preparation of silicone—ethylhexyl acrylate films is almost 100%. The polyacrylate films are stable towards basic solutions up to pH 12. The column efficiency is demonstrated with separations of anthracene derivatives. A fast separation of methyl-substituted benzenes with on-column UV detection is shown.
A vacuum assisted dynamic solvent evaporation interface for coupling of two-dimensional normal phase/reverse phase liquid chromatography was developed and evaluated. A normal-phase liquid chromatographic (NPLC) column of a 250mmx4.6mm I.D. 5microm CN phase was used as the first dimension, and a reversed-phase liquid chromatographic (RPLC) column of 250mmx4.6mm I.D. 5micromC(18) phase was used as the second dimension. The eluent from the first dimension flowed into a fraction loop, and the solvent in the eluent was dynamically evaporated and removed by vacuum as it was entering the fraction loop of the interface. The non-evaporable analytes was retained and enriched in about 5-25microL solution within the loop. Up to 1mL/min of mobile phase from the first dimension can be evaporated and removed dynamically by the interface. The mobile phase from the second dimension then entered the loop, and dissolved the concentrated analytes retained inside the loop, and carried them onto the second dimension column for further separation. The operation conditions of the two dimensions were independent from each other, and both dimensions were operated at their optimal chromatographic conditions. We evaluated the interface by controlling the loop temperature in a water bath at normal temperature, and investigated the sample losses by using standard samples with different boiling points. It was found that the sample loss due to evaporation in the interface was negligible for non-volatile samples or for components with boiling point above 340 degrees C. The interface realizes fast solvent removal of mL volume of fraction and concentration of the fraction into tenth of microL volume, and injection of the concentrated fraction on the secondary column. The chromatographic performance of the two-dimensional LC system was enhanced without compromise of separation efficiency and selectivity on each dimension.
An approach of immobilizing mobile crystalline material (MCM)-41 mesoporous silica nanoparticles on the inner wall of an open-tubular (OT) capillary as the support for coating chiral selector of cellulose tris(3,5-dimethylphenyl-carbamate) (CDMPC) was carried out. By taking advantage of the improved phase ratio of OT capillary with the immobilization of MCM-41 mesoporous material, the cellulose derivative of CDMPC as the chiral selector was simply coated on the MCM-41 nanoparticle layer via the hydrogen-bonding interaction, and the enantioseparation was successively carried out. Eight pairs of acidic, neutral and basic enantiomers were resolved in capillary electrochromatography or capillary liquid chromatography mode. The concentration of CDMPC for coating was systematically investigated to obtain the optimized chromatographic properties on enantioseparation by controlling the supposed film thickness of CDMPC on MCM-41 nanoparticle layer. Comparing with a bare fused silica capillary column coated with CDMPC under the same coating procedure as MCM-41-modified capillary did, the MCM-41-modified capillary column offered much higher enantioselectivity. This result indicated the significance of using the mesoporous nanoparticles as the electrochromatographic support to enhance the phase ratio of OT capillary column in capillary electrochromatography and capillary liquid chromatography. For investigating the effect of experimental conditions on the enantioseparation with this prepared OT capillary column, the content of organic modifier acetonitrile in the mobile phase was thus extensively evaluated to achieve a better chiral separation.
Recently, Se-substituted selenocysteine conjugates were proposed as potential prodrugs to target biologically active selenol compounds to tissues containing high activities of cysteine conjugate beta-lyases, such as the kidneys. However, several selenium compounds are known to be relatively toxic compounds. In the present study, the cytotoxicity of 14 selenocysteine Se-conjugates was determined in freshly isolated rat renal proximal tubular cells (RPTC). The results of this study show that four selenocysteine Se-conjugates with alkyl substituents (methyl, ethyl, n-propyl, and n-butyl) did not cause significant cytotoxicity to RPTC up to concentrations of 500 microM after 90 min of incubation. Also, no effect was observed on mitochondrial functioning as indicated by the unaffected mitochondrial membrane potential (delta psi). Se-(i-Propyl)-selenocysteine, however, appeared to be a cytotoxic compound, causing time- and dose-dependent cytotoxicity, and caused a decrease of delta psi in remaining viable cells. Aminooxyacetic acid (AOAA) provided significant protection against cell death of Se-(i-propyl)-selenocysteine, pointing to involvement of cysteine conjugate beta-lyase. AOAA, however, did not prevent the decrease of delta psi. Differentially substituted Se-(phenyl)-L-selenocysteine and Se-(benzyl)-L-selenocysteine conjugates appeared to be cytotoxic to RPTC at a concentration of 200 microM, as indicated by increased cell death and a decreased delta psi in remaining viable cells. Within the Se-benzyl-series, Se-(4-methoxybenzyl)-L-selenocysteine was the most toxic conjugate, whereas Se-(4-chlorophenyl)-L-selenocysteine was the most toxic conjugate of the Se-phenyl compounds. The selenocysteine Se-conjugates with nonsubstituted phenyl and benzyl substituents were nontoxic at 200 microM, but caused significant cell death at a concentration of 500 microM. Preincubation with AOAA, an inhibitor of cysteine conjugate beta-lyase, provided only partial protection against the cytotoxicity of Se-(phenyl)-L-selenocysteine (500 microM) and Se-(4-methoxybenzyl)-L-selenocysteine (200 microM). AOAA did not protect against cytotoxicity of the other conjugates, suggesting direct effects of these compounds or involvement of alternative routes of bioactivation. This study demonstrates that cytotoxicity of selenocysteine Se-conjugates is strongly dependent on the nature of the Se-bound substituent. The nontoxic Se-(alkyl)-Se-conjugates may be promising candidates for further evaluation for chemopreventive activities.
An orthogonal two-dimensional liquid chromatographic (2D-LC) system was developed by using a vacuum-evaporation loop-type valve interface. Normal-phase liquid chromatography (NPLC) with a bonded CN phase column was used as the first dimension, and reversed-phase liquid chromatography (RPLC) with a C(18) column was used as the second dimension. All the solvents in the loop of the interface were evaporated at 90 degrees C under vacuum conditions, leaving the analytes on the inner wall of the loop. The mobile phase of the second dimension dissolved the analytes in the loop and injected them onto the secondary column, allowing an on-line solvent exchange of a selected fraction from the first dimension to the second dimension. The chromatographic resolution of analytes on the two dimensions was maintained at their optimal condition. Sample loss due to evaporation in the interface was observed that depended on the boiling point of the compound. Separation of sixteen polycyclic aromatic hydrocarbon mixtures and a traditional Chinese medicine Angelica dahurica was demonstrated.
OV-17 stationary phase has been modified by heating at 340°C and 380°C in a stream of oxygen. Many polar compounds produce peaks with improved symmetry on the modified phase and there is little or no change in the symmetry of other less polar compounds. The improvements in peak symmetry are accompanied by small changes in retention index and there is a reduction in the polarity of the stationary phase as measured by the McReynolds value.
Methods for the reproducible synthesis of curable cyanosilicone stationary phases have been developed. Different paths of synthesis has been compared, and the influence of these on the phase properties examined. A series of phases having 33, 50, 60, 75% cyano substitution, respectively, was prepared, in a manner so as to avoid the formation of amides. The effects of the use of methyltolyl, bis(tolyl) and methylvinyl blocks and bis(dimethyl)silylarylene units on cross-linking of cyano-containing organosiloxanes were examined. These phases were characterized by various chromatographic, spectroscopic and thermal analytical methods. Columns coated with the phases were tested with respect to efficiency, activity and thermal stability. The usefulness of the phases for selective separations of fatty acid methyl esters is demonstrated both with standards and actual samples of different origins.
A technique is described for obtaining immobilized coatings of methylphenylsilicones in fused-silica capillaries. The columns are statically coated with a very viscous methylphenylsilicone polymer which is terminated with silanol functions. Gummification in the column is carried out by heat-curing. An immobilized and solvent-resistant methylphenylsilicone film is then obtained without the need for addition of peroxides.
A critical comparison of commercial with laboratory-mad columns has revealed important new facts. Whereas commercial columns are now preferentially made of fused silica, there are strong arguments for retaining glass for laboratory- mad capillary columns. In parallel with the intense development of commercial columns there has been a considerable increase in interest in making individual columns. A primary reason may be that the study and optimization of new analytical applications require rapid and flexible column tailoring, which is not available from commercial manufacturers.The laboratory preparation on inert columns has recently been facilitated. Simpler and safer preparation techniques have become available, which permit non- specialized laboratories to include column preparation in their regular analytical activities. To support such laboratories, a detailed and comprehensive description of recent column preparation techniques is presented.
Non-extractable organic layers have been obtained by reacting typical organosilicon cyclics (siloxane or silazane ring) with supports common to gas and liquid chromatography. The resulting phases performed well in both techniques.
The phenomena created by ban broadening in space are eliminated by the use of a retention gap in the column inlet, i.e., by keeping ca. 50–80 cm of the capillary column free of stationay phase. The extraction of the stationary phase frm the retention gap zone is easy for polar coating. For non-polar silicone stationary phases, especially if immobilized, other methods have to be applied: (a) the phase may be burnt, (b) deactivated, uncoated pre-column is connected to the column inlet or (c) the column is prepared with an inlet section free of stationay phase. Several methods for the preparation of a joint between a pre-column and the separating column aer discussed. Glued joints are considered to be the most promising. It is proposed that commercial columns should be offered with an uncoated inlet section 3 m in length
High-temperature silylation (HTS) used for the deactivation of capillary columns was studied on the silica commonly used in HPLC to gain a better insight into this process. LiChrosorb Si 100 was silylated with disilazanes at different temperatures and the materials obtained were compared in terms of organic and silanol group surface concentration, IR reflection spectra, HPLC behavior, and pore distribution parameters. Applying diphenyltetramethyldisilazane, the volatile reaction products were monitored during the HTS process. With increasing silylation temperature up to 400°C the silanol surface concentration is reduced to a very small level independent of the organic group concentration which exhibits a broad maximum depending upon substituents and temperature. Up to 350°C triorganosiloxy groups prevail as bonded organic groups. It could be proved that HTS is accompanied by pore alterations of the silica matrix. Arguments proposed by different authors in explaining HTS effects are discussed.
Small diameter capillary columns have been statically coated with SE-54 at elevated coating temperatures successfully. The method should prove equally successful for the preparation of nonextractable, highly efficient columns coated with other stationary phases in other solvents. This should facilitate the use of small diameter capillary columns in both gas and supercritical fluid chromatography.
Persilylated glass capillary columns are the most universal GC columns available. They are relatively easy to prepare, but there is nevertheless a great need for helpful practical instructions. In the recent literature, methods and conditions for the preparation of persilylated columns have been recommended, but these might confuse the individual column maker, since they seem to originate from limited observations rather than from comprehensive investigation. They will hardly help to make column preparation safer. This paper attempts a summarized view of the present experience with persilylation and provides basic information for column makers.
The effect of deactivating a fused silica surface by silylation with 1,1,3,3-tetraphenyl-1,3-dimethylilazane (TPDMDS), triphenylsilylamine (TPSA), and octamethylcyclotetrasiloxane (D4) and by polydimethylsiloxane degradation (PSD) is studied. Rehydrated, dried, and deactivated Cab-O-Sil M5 samples are used as model materials for 29Si CP-MAS NMR analysis.
At about 350 °C, TPDMDS yelds mainly diphenylmethylsiloxysilane, dimethyldisiloxysilane, and triphenylsiloxysilane groups. TPSA yields phenyltrisiloxysilane, diphenyldisiloxysilane, and triphenylsiloxysilane groups. At 400°C, the products formed initially are eventually replaced by methyltrisiloxysilane or phenyltrisiloxysilane groups, while a substantial number of silanol groups still remains. The possible consequences for wettability are discussed.
D4 reacts with Cab-O-Sil even at 200°C, but a large number of silanol groups remains. This number decreases gradually at higher temperatures and becomes negligible above 400°C. The formation of methyltrisiloxysilane groups, which starts at 425°C, is predominant at 490°C.
A deactivation procedure is described based on a published method using hexamethyldisilazane in the gas phase. In addition to unusually high inertia and thermostability, the method produces truly neutral columns which allow simultaneous analysis of moderately strong free acids and bases. The silylated columns show their full potential only with on-column injection. Preliminary experimental directions are given; more elaborate directions will become available after extended optimization work.
The activity of fused silica capillary tubing has been shown to vary considerably, which severely influences the reproducibility of column deactivation and inertness. Methods for producing a reproducible, high activity surface with maximum hydroxylation by various hydrothermal treatments were evaluated. Columns pretreated with 20% nitric acid at 200°C for 10 hours, deactivated with D4, coated with OV-73, and crosslinked with azo-tert-butane were found to possess excellent inertness for sensitive analytes such as alcohols, amines, and alkaloid drugs.
Nonpolar capillary columns of high quality are now commercially available. Columns separating according to selective mechanisms are presently being developed, and some are already available. Such columns may make separation of new types of samples feasible. In addition, the reliability of semi-identification by retention index may be improved when separation is performed on columns of different polarities. This can easily be executed by mounting two columns in parallel in one injector.
After a general introduction on the advantages and special requirements of gas chromatography in glass capillaries the methods of preparing and testing thin film capillaries are described to an extent not convered in the authors' previous paper. By means of a series of chromatograms of typical artificial and genuine samples it is shown that the use of glass capillaries of variable length and with various stationary liquids does not present technical or methodical difficulties. With regard to temperature stability there is no difference in performance between packed and glass capillary columns. Moreover the preparation of glass capillary columns is easy, their separation efficiency is much better and the tailing behaviour is often superior in comparison with packed columns.
The concentration of surface hydroxyl groups on three types of capillary columns was determined by exchanging the hydroxyl protons with tritium. The tritium was quantified by combustion to tritiated water followed by scintillation counting. The number of hydroxyl groups on a leached and presumably hydroxylated Pyrex glass capillary column was found to be about 2.8 groups per square nanometer. This value was slightly less than the 4.6 groups per square nanometer that is generally accepted for a fully hydroxylated porous silica. Dehydroxylation of the same glass by heating at 600C left only 0.4 groups per square nanometer while an untreated fused silica had only 0.2 groups per square nanometer.
A new method of making alkylpolysiloxane (OV 1, OV 101, OV 17, Dexsil 400, SE 52, SE 54) columns with high stability at temperatures above 300C of both the stationary liquid itself (low bleeding) and of the surface deactivation (no tailing of polar solutes with extended use at temperatures beyond 250C) is described. The best results were achieved on dealkalinized alkali glass but also on borosilicate surfaces which have been additionally HF treated before coating in both cases. The procedure can also be applied to leached alkali and borosilicate surfaces successfully. The best deactivations of the glass surface are attained in a two step procedure: firstly by treatment with gaseous HF/N2 mixture and secondly by simple heating of the coated and closed column at temperatures between 350 and 450C for several hours. The tailing behaviour of such columns for polar solutes is considerably improved by this procedure as shown by test chromatograms. The procedure can also be applied to leached alkali and borosilicate surfaces successfully.
The method of support surface deactivation by PSD (alkylpolysiloxane degradation) at temperature between 300 and 450C previously described was used to deactivate both fused silica and alkali glass surfaces of capillary columns. The latter surfaces had to be pretreated before deactivation with aqueous HCl leaching or by the dealkalisation method using flowing HCl gas at 450C and subsequent rinsing with water for alkali removal. Excellent alkylpolysiloxane columns with regard to tailing and irreversible adsorption of highly polar solutes have been obtained on both fused silica and the pretreated alkali glass. Fused silica does not require pretreatment before deactivation by the PSD-method, however. Good polyethyleneglycol (Carbowax 20 M) columns can also be obtained by coating the two types of surfaces when no deactivation is necessary. Deactivation by the PSD method cannot be applied in this case because polar stationary liquids do not adhere to alkylpolysiloxane deactivated surfaces. Sample capacity problems arising when separating highly polar solutes with non-polar stationary phases have also been investigated.
Both a fully hydroxylated silica (HiSil 233) and a partially hydroxylated silica (Cab-o-Sil M-5) were treated with organosilanes and a silazane. Water adsorption isotherms were then monitored on the resulting products. Reflectance IR spectroscopic measurements were used to follow the reactions of the surface hydroxyls with the silanes and to examine the subsequent water adsorption. The main objective was to ascertain the water susceptibilities of the different functional groups and the availability to water of any residual hydroxyls.The silazane and silanes employed were: hexamethyldisilazane, vinyl-trichlorosilane, and the alkoxysilanes: γ-glycidoxypropyltrimethoxysilane, 3-mercaptopropyltrimethyoxysilane, and amino-propyltriethoxysilane. The latter three were reacted from the liquid phase, first hydrolyzing to the silanetriols, with additional complications from condensation reactions. The vinyl compound and the disilazane were reacted from the vapor phase.Sharp differences between the results for the fully hydroxylated and the partially (25%) hydroxylated silica were found. The silanetriols upon heat treatment formed a multilayer polymer network on the latter and only a two-dimensional network on the former apparently due to hydrogen bonding to residual hydroxyls. The reacted silanes formed umbrellas over reacted hydroxyls on the former which were nevertheless available to water molecules. With the latter, however, a much more hydrophobic surface was produced. A number of details were also provided by the reflectance IR measurements.
Handbook of Chemistry and Physics” 52. Ed. The Chemical Rubber Co
- E H Ratcliffe
- E. H. Ratcliffe
The Proceedings of 6th Int
- Markides
- R C Kong
- M L Lee
- R. C. Kong