solvents, canbe calculated using an equation developed by Moore et al. [20], viz,                                            (52) where (y1) is the mole fraction of solvent (1), (y2) is the mole fraction of solvent (2), (h1) is the viscosity of solvent (1), and (h1) is the viscosity of solvent 2. By employing the viscosity values calculated for a range of temperatures and solvent mixtures in equation (50) the effect of temperature and solvent composition on the optimum mobile phase velocity could be calculated. A practical initial value for () was assumed to be 2.5 x 10-5 cm2s. This value was taken as the probable estimate from publishd data for similar compounds (21,22) for the diffusivity of

allowing a longer column to be used. The permeability increases as the square of the particle diameter but the variance per unit length only increases linearly with the particle diameter. Thus, doubling the particle diameter will allow a column four times the length to be used but the number of plates per unit length will be halved. Consequently, the column efficiency will be increased by a factor of two. It is also seen that the higher efficiencies will be obtained with mobile phases of low viscosity and for solutes of low diffusivity. Solvent viscosity and solute diffusivity tend to be inversely proportional to each other and so the sensitivity of the maximum obtainable efficiency to either solvent viscosity or solute diffusivity will generally not be large. The approximate length of a column that will provide the maximum column efficiency when operated at optimum velocity is given by, l = nHmin

nbsp;It is interesting to note from equation (60) that when a capillary column is run at its optimum velocity, the maximum efficiency attainable is directly proportional to the inlet pressure and the square of the radius and inversely proportional to the solvent viscosity and the diffusivity of the solute  in the mobile phase. This means that the maximum efficiency attainable from a capillary column increases with the column radius. Consequently, very high efficiencies will be obtained from relatively large  diameter, but very long, columns. Employing equation (60) it is possible to calculate the range of efficiencies attainable under operating conditions commonly used for packed columns. The conditions assumed are as follows,   

nbsp;                                        (51) where is the Diffusivity of the solute at temperature (Tx) and solvent composition (cx) is the Diffusivity of the solute at the reference temperature (To) and solvent composition (co) is the viscosity of the solvent at temperature (Tx) and solvent composition (cx) and is the viscosity of the solvent at the reference temperature (To) and solvent composition (co

Flow Controllers A constant pressure applied to a column does not ensure a constant flow of mobile phase though the chromatographic system, particularly if the column is being temperature programmed. Raising the temperature of a gas causes the viscosity to increase, and at a constant inlet pressure, the flow rate will fall. The reduction in flow rate will be related to the temperature program limits and to a certain extent on the temperature gradient. To obviate the flow rate change, mass controllers are used which ensure a constant mass of mobile passes through the column in unit time irrespective of the system temperature. A diagram of a mass flow controller is shown in figure 3.   Courtesy of Porter

nbsp; Thus,                             (44) It is seen that the column length varies inversely as the product of the solute diffusivity in the mobile phase and the mobile phase viscosity in much the same way as the column efficiency does when operating at the optimum velocity. As would be expected the column length is directly proportional to the inlet pressure but, less obviously is also proportional to the cube of the particle diameter. The analysis time for a solute mixture in which the last peak is eluted at a capacity ratio of k'F is given by,