[Readers Insight] Mobile Phase Selection in Method Development: How to Optimize

In previous article, we studied the acidity/basicity through compound structure. Only by understanding the chemical structure of a compound can we design a scientifically sound method and decide, on the basis of structure, which mobile phase and column to select. Naturally, analyzing a compound’s acid–base properties is the simplest and most fundamental first step — and also the most important.

In this article, we will examine the relationship between mobile phase and molecular structure, from the perspective of the mobile phase.

In conventional reversed-phase chromatography, the mobile phase typically consists of a buffer salt and an organic modifier. Different buffer salts and organic modifiers can be combined in many ways, provided the two phases are mutually miscible.

If impurity pairs remain unresolved after trying methanol, acetonitrile, or their mixtures, try above less common organic modifiers. They can produce unexpectedly good results for structures that are difficult to separate by conventional reversed-phase, especially non-enantiomeric stereoisomers and positional isomers.

If we are to develop an HPLC–UV method, choose phosphate buffers where possible because of their low UV cutoff — they are friendly for low-wavelength detection and offer three buffering regions, providing greater selectivity. The main drawback is poorer miscibility with organic modifiers; avoid overly high organic percentages to prevent buffer precipitation and instrument clogging.

Ammonium acetate (CH₃COONH₄) and ammonium formate (HCOONH₄) are more miscible with organic solvents and are volatile, making them first choices for LC–MS. Their UV cutoff is ~210 nm, so they are less suitable for very low-wavelength UV work. Acetate also exhibits some ion-pairing character relative to phosphate and thus may provide different selectivity.

Trifluoroacetic acid (TFA), because of the strong electron-withdrawing effect of fluorine, has stronger ion-pairing behavior and often produces excellent peak shapes for basic compounds. Additives similar to TFA typically have a carboxylate head and a fluorinated hydrophobic tail, such as pentafluoropropionic acid (PFPrA), heptafluorobutanoic acid (PFBA/HFBA) and nonafluoropentanoic acid (PFPeA) — different chain lengths impart different retention and selectivity.

It should be noted, that TFA in the mass spectrometer suppresses negative-ion response, compromising negative-ion detection; only positive-ion signals are reliably observed.

Traditional ion-pair reagents were used for analytes not retained by conventional reversed-phase; however, because they irreversibly modify columns and require very long re-equilibration times, they have mostly fallen out of favor. They have been largely replaced by HILIC mode, mixed-mode, and newer chaotropic agents/counterions that do not modify the stationary phase, can improve peak shape for basic compounds, and have lower UV cutoffs (compared with TFA’s ≈210 nm).

Among chaotropic reagents, perchlorate shows better miscibility with organic solvents than hexafluorophosphate. However, since neither is volatile, they are not compatible with mass spectrometry.

Sulfate salts, on the other hand, are commonly used in peptide and protein analyses to modify peak shape. For example, the insulin related substance method in Chinese Pharmacopoeia uses mobile phase A composed of 0.2 mol/L sulfate buffer (weigh 28.4 g anhydrous sodium sulfate, dissolve in water, add 2.7 mL phosphoric acid, adjust pH to 2.3 with ethanolamine, and dilute to 1000 mL with water) and acetonitrile (82:18, v/v).

Chloride salts (e.g. sodium chloride, ammonium chloride) are no longer preferred because Cl⁻ corrodes stainless-steel plumbing.

It should be put into caution that, these chaotropic reagents cannot be used alone because they lack buffering capacity; they must be used in combination with buffer salts.

Buffer capacity is often overlooked in method development. The effective buffer capacity of a mobile phase refers to its ability to resist pH changes when solute is added. The following factors can increase effective buffer capacity:

• Reduce the difference between the buffer pKa and the mobile-phase pH.
  For example, the primary dissociation region of phosphate has a buffering range roughly between pH 2 and 3.5; if the target pH is 4, phosphate will be an ineffective buffer at pH 4.
• Increase the difference between the mobile-phase pH and the analyte pKa.
  From pH–k plots we know that mobile-phase pH near a compound’s pKa has the greatest effect on retention; therefore, mobile-phase pH is typically kept at least ±1.5 units away from the pKa. If the difference is large enough, buffer capacity becomes less critical.
• Increase buffer concentration.
  The 5% acetic acid mentioned in The Starting Point in Method Development likely originates from this reasoning.
• Reduce sample injection volume.
  Less sample diffusing into the mobile phase effectively increases buffer capacity.
• Adjust sample pH to match the mobile phase.
  A large pH difference between sample solution and mobile phase can overwhelm buffer capacity.

In summary, we recommend the following initial mobile phases:

• For acidic compounds, begin with 0.1% phosphoric acid. Note that for carboxylic acids, if the compounds contain phosphate esters, metal complexation may cause them to tail; if containing sulfo group, they may remain ionized since conventional acids cannot suppress sulfonic-acid ionization.
• For basic or amphoteric compounds: begin with 0.1% phosphoric acid plus 20 mM potassium hexafluorophosphate (KPF₆) or 20 mM sodium perchlorate (NaClO₄) if target compounds can meet the pH condition.

These mobile phase systems both have pH values around 2.1 and therefore conveniently avoid the pKa regions of most acidic and basic compounds, reducing the risk of retention-time drift and durability issues.

Impurity separation, on the other hand, can generally be achieved by screening different columns and organic modifiers. Yet, in situations where available column types and quantities are limited and target analytes, while eluting closely, differ in pKa, pH adjustment can be used to resolve them. However, when pH falls within ±1.5 units of a compound’s pKa, small pH fluctuations will cause large shifts in retention time; therefore, durability and robustness must be carefully evaluated.

Furthermore, in reversed-phase systems, sodium and potassium salts usually behave similarly, but potassium hexafluorophosphate (KPF₆) is substantially less expensive than sodium hexafluorophosphate (NaPF₆), and perchlorate salts are regulated materials in most countries and require appropriate qualification for purchase.