The quantitative impact of the mesopore size on the mass transfer mechanism of the new 1.9 µm fully porous Titan-C18 particles II--analysis of biomolecules.

The kinetic performances of 3.0 × 100 mm columns packed with 1.9 µm Titan-C18 particles with average mesopore sizes of 80 Å and 120 Å were investigated quantitatively for the analysis of biomolecules. Large mesopores are expected to speed up the rate of diffusivity of high-molecular-weight compounds across the stationary phase and to generate higher plate counts at high velocities. The mass transfer mechanism of bradykinin acetate salt (1060 Da) and insulin (5733 Da) was determined over a range of flow rates from 0.025 to 1.0 mL/min. The pore diffusivities of these two biomolecules were accurately measured from the peak parking method. Even though the gain in column efficiency was not found significant for small molecules such as valerophenone (162 Da), enlarging the average pore size from 80 to 120 Å induces a measurable diminution of the reduced plate height, h, of bradykinin (from 17 to 11 or -35% at a reduced velocity of 50) and a significant reduction for insulin (from 43 to 12 or -72% at a reduced velocity of 90). Remarkably, while the increase of the column efficiency for bradykinin is consistent with a faster diffusivity of bradykinin across the 120 Å Titan-C18 particles, the higher column efficiencies measured for insulin are mostly due to a faster absorption kinetics into the 120 Å than that into the 80 Å Titan-C18 particles. This result is supported by the fact that the effective pore diffusivity of insulin is even slightly smaller across the 120 Å than that across the 80 Å 1.9µm Titan-C18 particles.

Hydrophilic interaction chromatography: A promising alternative to reversed-phase liquid chromatography systems for the purification of small protonated bases.

The overloaded band profiles of the protonated species of propranolol and amitriptyline were recorded under acidic conditions on four classes of stationary phases including a conventional silica/organic hybrid material in reversed-phase liquid chromatography mode (BEH-C18), an electrostatic repulsion reversed-phase liquid chromatography C18 column (BEH-C18+), a poly(styrene-divinylbenzene) monolithic column, and a hydrophilic interaction chromatography stationary phase (underivatized BEH). The same amounts of protonated bases per unit volume of stationary phase were injected in each column (16, 47, and 141 µg/cm(3)). The performance of the propranolol/amitriptyline purification was assessed on the basis of the asymmetry of the recorded band profiles and on the selectivity factor achieved. The results show that the separation performed under reversed-phase liquid chromatography like conditions (with BEH-C18, BEH-C18+, and polymer monolith materials) provide the largest selectivity factors due to the difference in the hydrophobic character of the two compounds. However, they also provide the most distorted overloaded band profiles due to a too small loading capacity. Remarkably, symmetric band profiles were observed with the hydrophilic interaction chromatography column. The larger loading capacity of the hydrophilic interaction chromatography column is due to the accumulation of the protonated bases into the diffuse water layer formed at the surface of the polar adsorbent. This work encourages purifying ionizable compounds on hydrophilic interaction chromatography columns rather than on reversed-phase liquid chromatography columns.

The quantitative impact of the mesopore size on the mass transfer mechanism of the new 1.9µm fully porous Titan-C18 particles. I: analysis of small molecules.

Previous data have shown that could deliver a minimum reduced plate height as small as 1.7. Additionally, the reduction of the mesopore size after C18 derivatization and the subsequent restriction for sample diffusivity across the Titan-C18 particles were found responsible for the unusually small value of the experimental optimum reduced velocity (5 versus 10 for conventional particles) and for the large values of the average reduced solid-liquid mass transfer resistance coefficients (0.032 versus 0.016) measured for a series of seven n-alkanophenones. The improvements in column efficiency made by increasing the average mesopore size of the Titan silica from 80 to 120Å are investigated from a quantitative viewpoint based on the accurate measurements of the reduced coefficients (longitudinal diffusion, trans-particle mass transfer resistance, and eddy diffusion) and of the intra-particle diffusivity, pore, and surface diffusion for the same series of n-alkanophenone compounds. The experimental results reveal an increase (from 0% to 30%) of the longitudinal diffusion coefficients for the same sample concentration distribution (from 0.25 to 4) between the particle volume and the external volume of the column, a 40% increase of the intra-particle diffusivity for the same sample distribution (from 1 to 7) between the particle skeleton volume and the bulk phase, and a 15-30% decrease of the solid-liquid mass transfer coefficient for the n-alkanophenone compounds. Pore and surface diffusion are increased by 60% and 20%, respectively. The eddy dispersion term and the maximum column efficiency (295000plates/m) remain virtually unchanged. The rate of increase of the total plate height with increasing the chromatographic speed is reduced by 20% and it is mostly controlled (75% and 70% for 80 and 120Å pore size) by the flow rate dependence of the eddy dispersion term.

The relative importance of the adsorption and partitioning mechanisms in hydrophilic interaction liquid chromatography.

We propose an original model of effective diffusion along packed beds of mesoporous particles for HILIC developed by combining Torquatos model for heterogeneous beds (external eluent+particles), Landauers model for porous particles (solid skeleton+internal eluent), and the time-averaged model for the internal eluent (bulk phase+diffuse water (W) layer+rigid W layer). The new model allows to determine the analyte concentration in rigid and diffuse W layer from the experimentally determined retention factor and intra-particle diffusivity and thus to distinguish the retentive contributions from adsorption and partitioning. We apply the model to investigate the separation of toluene (TO, as a non-retained compound), nortriptyline (NT), cytosine (CYT), and niacin (NA) on an organic ethyl/inorganic silica hybrid adsorbent. Elution conditions are varied through the choice of a third solvent (W, ethanol, tetrahydrofuran (THF), acetonitrile (ACN), or n-hexane) in a mobile phase (MP) of ACN/aqueous acetate buffer (pH 5)/third solvent (90/5/5, v/v/v). Whereas NA and CYT retention factors increase monotonously from W to n-hexane as third solvent, NT retention reaches its maximum with polar aprotic third solvents. The involved equilibrium constants for adsorption and partitioning, however, do not follow the same trends as the overall retention factors. NT retention is dominated by partitioning and NA retention by adsorption, while CYT retention is controlled by adsorption rather than partitioning. Our results reveal that the relative importance of adsorption and partitioning mechanisms depends in a complex way from analyte properties and experimental parameters and cannot be predicted generally.

Separation of peptides and intact proteins by electrostatic repulsion reversed phase liquid chromatography.

A new brand of BEH-C18 hybrid particles chemically bonded to a leash carrying an amine group permits the implementation of electrostatic repulsive interactions chromatography. Using columns packed with this material, the influence of the concentration of positive charges bonded to the BEH-C18 surface on the overloaded band profiles of a few positively charged peptides and proteins was investigated in the gradient elution mode. Three columns packed with endcapped BEH-C18 particles bonded with three different surface-charge densities (LOW, MEDIUM and HIGH) were used and compared with those provided by a column packed with non-doped, endcapped BEH-C18 particles. The surface concentrations of fixed charges in the LOW, MEDIUM and HIGH columns were estimated at 0.029, 0.050, and 0.064µmol/m(2), for example, about two orders of magnitude smaller than the surface density of bonded C18 chains (2.1µmol/m(2)). Three different mobile phase additives (0.1% v/v of trifluoro-acetic, phosphoric, and formic acid) were used to optimize the purification levels of proteins under different loading conditions. The weak ion-pairing ions (formate and phosphate) generate smaller retention but broader, more fronting band profiles than those eluted with a stronger ion-pairing ion (trifluoroactate). This effect is worse in the presence of fixed charges at the surface of the BEH-C18 particles. This was explained by an enhanced anti-Langmuirian adsorption behavior of the charged proteins in the presence of fixed surface charges. As the protein concentration increases in the bulk, so does the internal ionic strength, the electrostatic repulsive interactions weaken, and retention increases. Band fronting is mostly eliminated by replacing weak ion-pairing acids with TFA with which the adsorption isotherm remains weakly langmuirian. Faster but still complete gradient separation of insulin and myoglobin were achieved with the HIGH column than with the reference neutral column, despite a measurable loss in selectivity.

Effects of the surface concentration of fixed charges in C18-bonded stationary phases on the adsorption process and on the preparative chromatography of small ionizable compounds.

The effects of the surface concentration of positive charges attached to the surface of research BEH-C18 hybrid particles on the overloaded band profiles and the adsorption isotherms of a neutral (caffeine) and a positively charged (nortryptilinium hydrochloride) compounds were measured and investigated. The inverse method (IM) of chromatography was used to determine the isotherm parameters. Three columns were packed with endcapped BEH-C18 particles doped with three different charge densities on their surfaces (LOW, MEDIUM and HIGH). Two other columns packed with unbonded, non-endcapped, and endcapped BEH-C18 particles served as standard reference materials. Minor disturbance method (MDM) experiments were conducted with acetonitrile/water mixtures in order to assess qualitatively the surface densities of the fixed positive charges. A more quantitative approach based on the solution of the linearized Poisson-Boltzmann equation and the decrease of the experimental Henry constant was also applied. The results show that the surface concentrations of the fixed charges in the LOW, MEDIUM and HIGH columns were 0.029, 0.050, and 0.064µmol/m2, e.g., close to two orders of magnitude smaller than the surface density of bonded C18 chains (2.1µmol/m2). The adsorption isotherm of the ionizable compound nortryptilinium onto the BEH-C18 columns is consistent with a two-sites adsorption model. The density of the high energy sites correlates directly to the total amount of the fixed charges and isolated silanols amidst the C18-bonded chains. The amount of low energy sites reflects the specific surface area of the adsorbent. The binding constants on the high- and low-energy adsorption sites are respectively ten and two times lower on the HIGH column than on the reference endcapped column. The active sites are closer to the adsorbent surface than the weak adsorption sites. Finally, a higher production rate of ionizable compounds can be achieved in preparative chromatography with the charge doped than with the reference RPLC-C18 columns. For the same analysis time, larger recovery yields are observed despite a slight loss in selectivity due to non-selective repulsive electrostatic interactions.

Estimations of temperature deviations in chromatographic columns using isenthalpic plots. I. Theory for isocratic systems.

We propose to use constant enthalpy or isenthalpic diagrams as a tool to estimate the extent of the temperature variations caused by the mobile phase pressure drop along a chromatographic column, e.g. of its cooling in supercritical fluid and its heating in ultra-performance liquid chromatography. Temperature strongly affects chromatographic phenomena. Any of its variations inside the column, whether intended or not, can lead to significant changes in separation performance. Although instruments use column ovens in order to keep constant the column temperature, operating conditions leading to a high pressure drop may cause significant variations of the column temperature, both in the axial and the radial directions, from the set value. Different ways of measuring these temperature variations are available but they are too inconvenient to be employed in many practical situations. In contrast, the thermodynamic plot-based method that we describe here can easily be used with only a ruler and a pencil. They should be helpful in developing methods or in analyzing results in analytical laboratories. Although the most effective application area for this approach should be SFC (supercritical fluid chromatography), it can be applied to any chromatographic conditions in which temperature variations take place along the column due to the pressure drop, e.g. in ultra-high pressure liquid chromatography (UHPLC). The method proposed here is applicable to isocractic conditions only.

Characterization and kinetic performance of 2.1 × 100 mm production columns packed with new 1.6 µm superficially porous particles.

The overall kinetic performance of three production columns (2.1 mm × 100 mm format) packed with 1.6 µm superficially porous CORTECS-C18 + particles was assessed on a low-dispersive I-class ACQUITY instrument. The values of their minimum intrinsic reduced plate heights (h(min) = 1.42, 1.57, and 1.75) were measured at room temperature (295 K) for a small molecule (naphthalene) with an acetonitrile/water eluent mixture (75:25, v/v). These narrow-bore columns provide an average intrinsic efficiency of 395,000 plates per meter. The gradient separation of 14 small molecules shows that these columns have a peak capacity about 25% larger than similar ones packed with fully porous BEH-C18 particles (1.7 µm) or shorter (50 mm) columns packed with smaller core-shell particles (1.3 µm) operated under very high pressure (>1000 bar) for steep gradient elution (analysis time 80 s). In contrast, because their permeabilities are lower than those of columns packed with larger core-shell particles, their peak capacities are 25% smaller than those of narrow-bore columns packed with standard 2.7 µm core-shell particles.

Mass transport of small retained molecules in polymer-based monolithic columns.

The mass transport properties of a non-retained (thiourea) and three retained low molecular weight compounds (acetophenone, valerophenone, and octanophenone) along a 4.6mm×45mm PROSWIFT™ RP-1S monolithic column made of rigid cross-linked poly(styrene-divinylbenzene) copolymer was investigated in depth. Accurate protocols (peak parking experiments, measurement of the first and second central moments of peak profiles by numerical integration) combined with the use of validated models of effective diffusion along monolithic structures were applied for the determination of the longitudinal diffusion, the eddy dispersion, and the skeleton-eluent mass transfer resistances due to the finite analyte diffusivity across the polymer skeleton and to the slow absorption kinetics into the polymer volume. Experimental results show by increasing order of importance evidence that the resolution performance of this short and wide polymer-based monolithic HPLC column is limited by the slow analyte diffusivity across the polymer skeleton (smaller than one tenth of the bulk diffusion coefficient for k'>1), its large eddy dispersion HETP (Heddy≃100µm), and the slow rate of absorption (≃10Hz only) in the polymer volume for retained analytes. The column performance could be improved by preparing a more homogeneous material with a rigid internal mesoporous structure. This would provide a column bed having a larger specific surface area, allowing faster analyte diffusion across the mesoporous skeleton, a smaller eddy dispersion HETP, and a faster absorption kinetics in the polymeric monolith than those observed for the currently available materials.

Particle size distribution and column efficiency. An ongoing debate revived with 1.9µm Titan-C18 particles.

The mass transfer mechanism in four prototype columns (2.1 and 3.0×50mm, 2.1 and 3.0×100mm) packed with 1.9µm fully porous Titan-C18 particles was investigated by using two previously reported home-made protocols. The first one was used to measure the eddy dispersion HETP of these new columns, the second one to estimate their intrinsic (corrected for HPLC system contribution) HETPs. Titan particles are fully porous particles with a narrow particle size distribution (RSD of 9.2%). The mean Sauter diameter (dSauter=2.04µm) was determined from Coulter counter measurements on the raw silica material (before C18 derivatization) and in the absence of a dispersant agent (Triton X-100) in a 2% NaCl electrolyte solution. The results show that these RPLC Titan columns have intrinsic minimum reduced HETPs ranging from 1.7 to 1.9 and generate up to 290,000 plates per meter. The 3.0mm i.d. columns are more efficient than the 2.1mm i.d. ones and short columns are preferred to minimize efficiency losses due to frictional heating at high speeds. This work also revealed that (1) the lowest h values of the Titan columns are observed at low reduced velocities (νopt=5); (2) this is due to the unusually small diffusivity of analytes across the porous Titan-C18 particles; and (3) the Titan columns are not packed more uniformly than conventional columns packed with fully porous particles. Earlier and recent findings showing that the PSD has no direct physical impact on eddy dispersion and column efficiency are confirmed by these results.

Calculated and experimental chromatograms for distorted gradients and non-linear solvation strength retention models.

Computer calculations of gradient chromatograms were performed by taking into account the adsorption behavior of the strong eluent in RPLC and the true Henry constant of the analytes. This improves the accuracy of classical gradient calculations, which all assume no affinity of the eluent modifier for the stationary phase and that the linear solvation strength model (LSSM) applies. The excess adsorption isotherm of acetonitrile with respect to water was measured by the minor disturbance method onto a Symmetry-C18 RPLC adsorbent. The variations of the Henry constants of a nine compound mixture with the volume fraction of acetonitrile in the aqueous mobile phase were measured. The equilibrium dispersive model of chromatography combined with orthogonal collocation on finite elements was used to calculate chromatograms of the sample mixture for four gradient times decreasing from 25 to 1 min. The results predict a loss of resolution for the less retained analytes when the gradient times becomes smaller than 4 min. They also predict that this behavior can be eliminated when applying a quadratic gradient profile rather than a classical linear gradient. The predictions were validated by the agreement between the calculated and experimental chromatograms.

The rationale for the optimum efficiency of columns packed with new 1.9µm fully porous Titan-C18 particles-a detailed investigation of the intra-particle diffusivity.

In a previous report, it was reported that columns packed with fully porous 1.9µm Titan-C18 particles provided a minimum reduced plate height as small as 1.7 for the most retained compound (n-octanophenone) under RPLC conditions. These particles are characterized by a relatively narrow size distribution with a relative standard deviation (RSD) of only 10%. A column packed with classical 5µm Symmetry-C18 particles, used as a reference RPLC column, generated a minimum reduced plate height of 2.1 for the same retained compound. This work demonstrates that this was due to an unusually low intra-particle diffusivity across these particles, which leads to a small longitudinal diffusion coefficient along the column. The demonstration is based on the combination of accurate measurements of the height equivalent to a theoretical plate (HETP), inverse size exclusion chromatography (ISEC), peak parking (PP), and minor disturbance method (MDM) experiments. The experimental results show that the reduced eddy dispersion HETP term (A=0.8 for a reduced velocity of 5), the internal particle porosity (ϵp=0.35), and the enrichment of acetonitrile in the pore volume (75% acetonitrile in the bulk, 85% inside the mesoporous volume) are identical on both the Titan-C18 and Symmetry-C18 columns. The difference between the internal structures of these two brands of RPLC-C18 fully porous particles lies in the values of the internal obstruction factor γp, which is 0.42 for the Symmetry-C18 but only 0.26 for the Titan-C18 particles. This is in part related to the diffusion hindrance due to the small average pore size of the Titan-C18 particles, around 59Å versus 77Å for Symmetry-C18 particles. A simple model of constriction along diffusion paths having the shape of a truncated cone suggests that the width of the pore size distribution (RSD of 30% and 20% for Titan-C18 and Symmetry-C18 particles) is mostly responsible for the difference in their obstruction factors.

Effect of the pressure on pre-column sample dispersion theory, experiments, and practical consequences.

The effect of the pressure on the dispersion of a low molecular weight compound along 0.508 and 1.016 mm i.d. × 50 cm long open circular tubes was investigated theoretically and experimentally. The theoretical predictions were based on the early models of dispersion derived by Aris and Taylor (1953) and on the approximate model of Alizadeh for the time moments (1980). Experimentally, the system pressure was increased at constant flow rate (0.15 mL/min) from less than 20 to nearly 1,000 bar by using a series of capillary tubes (25 µm i.d. PEEKSIL tubes) of increasing flow resistances placed upstream the detection cell of a commercial very high pressure liquid chromatograph (vHPLC) but downstream the 50 cm long tube. Theoretical and experimental results agree that the peak variance increases linearly with increasing pressure in the tube volume. The relative increase of the peak variance is 7% above that measured at low pressure (<20 bar) for each 100 bar increment in the tube volume. This result confirms that accurate measurements of the column efficiency corrected for extra-column contribution cannot be made by replacing the column with a zero dead volume union connector, because the pressures applied in the pre-column volume are significantly different in these two cases. This work shows also that increasing the pressure in the pre-column volume by increasing the flow rate affects the apparent column efficiency that is measured, independently of the direct effect of the flow rate. For a 2.1 × 50mm column packed with 1.3 µm core-shell particles run with a classic Acquity system, the associated relative decreases of the column efficiency are expected to be -30%, -20%, and -5% for retention factors of 1, 3, and 10, respectively. The column HETP is no longer independent of its length.

Separations by gradient elution: why are steep gradient profiles distorted and what is their impact on resolution in reversed-phase liquid chromatography.

The formation of a concentration shock layer of either methanol or acetonitrile used as the modifier in steep RPLC aqueous/organic gradients was predictable from their excess adsorption isotherm and was directly observed from their elution profiles recorded in UV absorption. A discontinuity of the concentration profile of the organic modifier arises and grows along the column. Its effects on the peak shapes and the resolution levels achieved in steep gradient chromatography were measured in the recorded chromatograms of a sample mixture containing 14 different analytes uniformly distributed across the gradient retention window. The results showed severe peak shape deformation for some compounds and a significant degradation of the baseline resolution of weakly (when using acetonitrile) and moderately (when using methanol) retained analytes. Solutions to limit this loss of resolution may lie in changing the initial gradient conditions or in adjusting the curvature of the gradient profile at the column inlet.

The adsorption of naproxen enantiomers on the chiral stationary phase (R,R)-whelk-O1 under supercritical fluid conditions.

The adsorption isotherms of the Naproxen enantiomers eluted with a mixture of carbon dioxide and methanol on a (R,R)-Whelk-O1 column were measured using frontal analysis and the inverse method. Sharp breakthrough fronts provided using a modified design of the instrument allowed an accurate determination of these isotherms. The adsorption model that best accounts for the data was derived from the analysis of the adsorption isotherm data. The affinity energy distribution of the adsorption of the two enantiomers were calculated to determine the heterogeneity of the adsorbent surface for each compound. The adsorption model chosen based on the adsorbent heterogeneity was fitted to the adsorption isotherm data points and the best fitted isotherm parameters were derived from this fit. The inverse method was used as a second method to derive the parameters of the same adsorption models for the overloaded elution bands of (S)- and (R)-naproxen. The isotherm parameters derived from these independent numerical calculations are discussed and compared in the case of single component adsorption and of competitive adsorption as well.

Possible resolution gain in enantioseparations afforded by core-shell particle technology.

Whether columns packed with core-shell particles may outperform those packed with fully porous particles for chiral separations is controversial. The potential advantages of such columns are investigated from a theoretical viewpoint. The height equivalent to a theoretical plate (HETP) associated to the slow adsorption-desorption process typical of chiral chromatography was derived from the Laplace transform of the general rate model of chromatography for core-shell particles. The relationship between the resolution factor and the core-to-particle diameter ratio is predicted at constant selectivity. The calculations are based on a complete set of actual kinetic parameters (longitudinal diffusion, eddy dispersion, intra-particle diffusivity, and adsorption-desorption constant) measured for a reference column packed with Lux Cellulose-1 fully porous particles. If we compare columns packed with the best procedure available in either case, the results demonstrate that those packed with core-shell particles may outperform to a degree those packed with fully porous particles. The minimum reduced HETP could decrease from 2.0 to 1.7. The maximum relative gain in resolution is about 10%, which is not negligible for critical enantioselective-separations. This gain is observed only if the packing uniformity of the core-shell particles is achieved.

Band broadening along gradient reversed phase columns: a potential gain in resolution factor.

The advantages of using columns packed with stationary phases having a composition gradient so that retention factors increase toward the outlet and eluting them with an isocratic mobile phase may provide increased peak resolution in liquid chromatography. This approach is discussed from a theoretical viewpoint. The retention factor is assumed to increase linearly along the column. The peak width can be predicted under three different conditions: (1) the possible band compression is neglected (the Giddings model); (2) band compression is taken into account (the Poppe model); and (3) both band compression and extra-column effects are considered in the calculations. The impact of a stationary phase gradient on the resolution of a pair of compound difficult to separate (constant selectivity α=1.05) is illustrated for a 3mm×100mm silica monolithic column of the second generation (porosity ϵt=0.85, plate height H=6µm) operated with a new generation of liquid chromatograph (pre- and post-column volume variances of 10 and 5µL(2), respectively). The results show that the resolution factor RS becomes optimum for a specific positive value of the retention gradient along the column. This optimum depends on the retention factor at the column inlet.

Comparison of volume and concentration overloadings in preparative enantio-separations by supercritical fluid chromatography.

The adsorption isotherms of both naproxen enantiomers were measured on a (R,R)-Whelk-O1 chiral adsorbent from a mobile phase made of liquid carbon dioxide and methanol, using frontal analysis. These isotherms were used to model the competitive adsorption behavior of a racemic mixture of naproxen enantiomers. Their overloaded elution band profiles were calculated numerically using these isotherm data. The injected volume and the injected concentration of the samples of racemic mixture were carried out simultaneously, using two objective functions. The maximization of the production rate and of the product of the recovery yield and the production rate were performed. Based on the optimum points obtained, the two objective functions were compared for several recovery yield constraints in a range of injected volumes and concentrations. The consequences of choosing a separation design focused on the degree of band overlapping are discussed based on the results of these optimizations.

Volume based vs. time based chromatograms: reproducibility of data for gradient separations under high and low pressure conditions.

A critical aspect in fast gradient separations carried out under constant pressure, in the very high pressure liquid chromatography (VHPLC) mode is that time-based chromatograms may not yield highly reproducible separations. A proposed solution to improve the reproducibility of these separations involves plotting the chromatograms as functions of the volume eluted vs. UV absorbance instead of time vs. UV. To study the consequences of using the volume-based rather than the time-based chromatograms, separations were first performed under low pressures that do not generate significant amounts of heat and for which the variations of the eluent density along the columns are negligible. Secondly, they were performed under very high pressures that do generate heat and measurable variations of the local retention factor and eluent density along the column. Comparison of the results provides estimates of the improvements obtained when volume based chromatograms are used in gradient analyses. Using a column packed with fully porous particles, four different types of methods and several sets for each method were used to perform the gradient elution runs: two sets of constant flow rate operations, four sets of constant pressure operations, two sets of constant pressure operations with programmed flow rate, and one set using the constant heat loss approach. The differences between time-based and volume-based chromatograms are demonstrated by using eight replicates of early, middle, and last eluting peaks. The results show that volume-based chromatograms improve the retention time reproducibility of the four constant pressure methods by a factor of 3.7 on average. If the column is not thermally conditioned prior to performing a long series of separations, flow controlled methods (constant flow rate, programmed constant pressure, and constant wall heat approaches) are more precise. If one gradient run is used to bring the column to a relatively stable temperature, constant pressure separations have a factor of 3 times better reproducibility of retention times with respect to constant flow rate gradient separations.