Gas, liquid and other chromatographic methods

A method name does not establish compatibility

Gas chromatography uses a gaseous mobile phase; appropriate analytes must be transferable and sufficiently stable under the particular conditions. Liquid chromatography uses a liquid mobile phase and suitable stationary interactions. Solubility, chemical form, stability, matrix, phase compatibility and detector response all matter. Neither name guarantees a separation or a quantitative result.

A hypothetical constituent X is stipulated to be insufficiently volatile and to transform under a proposed direct-GC method, while remaining soluble and stable in a given LC medium with a calibrated detector response. The proposed direct-GC option is incompatible with the stated X conditions. LC is a candidate requiring selectivity and procedure validation, not an automatically successful solution.

A different hypothetical Y is volatile and stable in a proposed GC method but gives no response in the proposed ultraviolet channel. Those separation prerequisites do not repair the missing detector response. Derivatization may improve suitability in some methods, but it changes the measured entity and introduces conversion and recovery questions. No actual substances or operating programs are assigned here.

The phase interaction supplies selectivity

Adsorption emphasizes interaction with a surface; partition describes distribution between phases. Ion-exchange retention depends on ionic interactions and the relevant chemical forms and competing ions. Size-exclusion chromatography emphasizes access to pore volume under conditions that minimize other interactions. Real stationary phases can combine mechanisms.

In LC, an isocratic method maintains the stipulated mobile-phase composition, while a gradient changes composition over time. Such changes can alter retention and response, so a constant-k equilibrium model should not be transplanted unchanged into a gradient. In GC, temperature and gas-pressure conventions affect time, volume and velocity comparisons.

A retention index compares an analyte with a specified reference series under a defined convention. For an isothermal Kováts index, logarithms of adjusted retention times are used between bracketing normal alkanes. Temperature-programmed linear indices use a different relation. Reference normalization improves a comparison’s structure but does not make it unique identity evidence or independent of method conditions.

A plate-height minimum depends on the chosen model

A simple packed-column van Deemter model is H=A+B/u+Cu, where u is the explicitly chosen linear velocity and H is plate height. The terms represent different contributions to band broadening within that model. A, B and C are not universal constants: analyte, phase, temperature and geometry influence them.

Assign hypothetical A=0.0200 mm, B=0.0600 mm² s⁻¹ and C=0.00500 s. With u in mm s⁻¹, all three terms have mm units. The model minimum occurs at u*=√(B/C)=3.46410 mm s⁻¹, where H*=A+2√(BC)=0.0546410 mm. The low-velocity B/u term and high-velocity Cu term are equal at this point.

For an assigned L=150 mm, N=L/H*=2745.19. If this toy treats u as the appropriate mobile passage velocity and k=3.00, t_M=L/u*=43.3013 s and t_R=(1+k)t_M=173.205 s. The calculation does not optimize a real column, pressure, gradient, compressible flow or extra-column contribution. Faster motion alone can increase the Cu contribution.

Volume flow and linear velocity have different meanings

A volume flow rate F divided by the full column cross-sectional area gives a superficial velocity u_s=F/A_cross. If only an assigned fraction ε of that area is available to interparticle flow, an interstitial definition gives u_i=F/(εA_cross). These are different bases for the same volume flow. A transport equation must state which one its parameters use.

In a pure unit and geometry exercise, stipulate incompressible F=0.600 mL min⁻¹, diameter 1.00 mm and interparticle porosity ε=0.400. Flow is 10.0 mm³ s⁻¹ and cross-sectional area is π/4=0.785398 mm². Superficial velocity is 12.7324 mm s⁻¹; interstitial velocity is 31.8310 mm s⁻¹.

Total accessible mobile volume in porous particles can require another volume definition. A GC volume flow also needs the relevant temperature and pressure conventions. These numerical inputs supply geometry, not evidence that a particular column or pressure is usable. They are separate from the previous plate-height example.

A size-exclusion model tracks accessible volume

For an ideal size-exclusion volume model, V_e=V_0+K_access V_i. Here V_e is retention volume, V_0 is outside-pore mobile volume and V_i is the fully accessible pore volume. The coefficient K_access lies between zero and one in this restricted model. Excluded solutes approach V_0; fully accessing solutes approach V_0+V_i.

Assign V_0=11.0 mL and total fully accessible mobile volume V_t=31.0 mL, so V_i=20.0 mL. For observed hypothetical V_e=23.0 mL, K_access=(23−11)/20=0.600. The formula describes pore-volume access rather than directly providing a molecular mass. Size, shape, calibration chemistry and non-size interactions affect interpretation.

A stipulated comparison at V_e=9.00 mL would give K_access=−0.100. That is outside the assigned pure access model; it is not a negative molecular size. Additional delay conventions, exclusion anomalies, interactions or a wrong volume reference would need investigation rather than silently accepting the model.

An average retention factor uses mobile-form fractions

Suppose symbolic HA and A⁻ interconvert rapidly in the mobile phase. In an ideal model with fixed form-specific k_HA and k_A, stationary inventory is the sum k_HA n_HA,m+k_A n_A,m. Dividing by total mobile inventory gives apparent k=f_HA,m k_HA+f_A,m k_A. The weights are mobile-phase fractions, not fractions of the whole mobile-plus-stationary inventory.

Assign ideal pK_a=6.00 with activity coefficients one and standard concentration 1 mol L⁻¹, k_HA=4.00 and k_A=0.500. At pH 5, 6 and 7, the neutral mobile fractions are 10/11, 1/2 and 1/11. Apparent k values are 3.68182, 2.25000 and 0.818182. This example deliberately holds solvent, temperature, ionic environment, stationary phase and form-specific k values fixed.

Real pH changes can also change surface charge and other interactions. The decreasing trend in this invented case is not universal, and it is not a GC prediction. Method choice requires the actual forms and mechanism, not a broad rule based only on a method label or a solute charge.

Worked examples

Try it yourself

1. In the same H model, evaluate H at u=2.00 mm s⁻¹ and calculate N for L=150 mm.

   Show the answer and reasoning

   H=0.0600 mm; N=2500.

   The three contributions are 0.0200, 0.0300 and 0.0100 mm. Divide 150 mm by their sum.

2. Keep V_0=11.0 mL and V_t=31.0 mL but assign V_e=21.0 mL. Find K_access.

   Show the answer and reasoning

   0.500.

   The solute accesses 10.0 mL of the assigned 20.0 mL pore volume in this model.

3. Hypothetical X transforms under a proposed direct-GC method but is stable and soluble in a stated LC medium with calibrated response. What conclusion follows?

   Show the answer and reasoning

   The proposed direct-GC option is incompatible; LC is a candidate requiring selectivity and full-procedure validation.

   Solubility and response prerequisites do not yet show that LC separates X from relevant matrix constituents or measures the intended quantity reliably.

4. Why is an apparent k calculated from the mobile-form fractions rather than simply averaging fractions of the whole phase inventory?

   Show the answer and reasoning

   k is stationary amount divided by mobile amount.

   Each stationary-form amount is its form-specific k times its mobile-form amount. Summing those amounts and dividing by total mobile amount gives mobile-fraction weights.

What to carry forward

• Match analyte, medium, stationary interactions and detector response.
• Temperature, composition and pressure conventions affect retention comparisons.
• A transport minimum belongs to specified constants and a velocity definition.
• Accessible-volume coefficients require a compatible size-exclusion model.
• Rapid-exchange apparent retention uses mobile-form inventory weights.