Solvents and Caveats for LC/MS

Solvents are typically chosen based on a compound of interest’s solubility and compatibility with various ionization techniques used in LC/MS. Volatility and the solvent’s ability to donate a proton are important in ESI and other atmospheric ionization techniques.

Protic primary solvents like methanol and mixtures with water, such as 1:1 methanol/water or 1:1 acetonitrile/H₂O, are used (although the water/methanol mixture increases viscosity well beyond either water or menthol as a neat solvent because of a resulting exothermic reaction). Water’s relatively low vapor pressure can be detrimental to sensitivity when employed at 100%. Better sensitivity results when surface tension is decreased through addition of a volatile organic solvent. Surfactants with higher proton affinity, though they increase ion liberation from nebulized droplets, can also reduce sensitivity.

Aprotic co-solvents like 10% DMSO in water and isopropyl alcohol improve solubility for some compounds. Formic acid is often added at low levels (0.1%) to facilitate ionization by ensuring the analyte is more basic than the solvent. Even in small amounts, however, some acids, like TFA, though necessary for otherwise insoluble compounds, can limit sensitivity.

In the ESI ionization mode, buffers and salts (Na⁺, K⁺, and phosphate) cause a reduction in the vapor pressure and consequently a reduced signal. The increased surface tension of the droplets, and resultant reduction of volatility, can be remedied by using relatively more volatile buffers like ammonium acetate, formed by a weak acid-base pair.

Solvent considerations

• Solvent in the gas phase limits ionization by ESI to molecules more basic than the solvent. The exception is photoionization (which is not acid/base ionization) but nonetheless mediated by solvent.
• Removing solvent and water vapor from the ionization region increases types of compounds that can be ionized at atmospheric pressure.
• Reducing liquid volume relative to the sample or analyte of interest contained in the liquid improves ESI performance (i.e., lower flow rates).
• Useful Solvents
• Water
• Acetonitrile
• Methanol
• Ethanol
• Propanol
• Isopropanol
• Acceptable additives
• Acetic acid
• Formic acid
• Ammonium hydroxide
• Ammonium formate (salt concentration = 10 mM or less)
• Ammonium acetate (salt concentration = 10 mM or less)
• Nonvolatile salts (phosphate, borate, citrate, etc.)
• Can deposit in source and plug capillaries thus requiring more cleaning and maintenance operations
• Modern source designs can handle nonvolatiles better than older designs
• Surface-active agents (surfactants/detergents) suppress electrospray ionization
• Inorganic acids are corrosive
• Trifluoroacetic acid (TFA)
• To some extent suppresses positive-ion electrospray at levels exceeding 0.01%.
• Greatly suppressed negative-ion electrospray.
• Triethylamine (TEA)
• High PA (232 Kcal/mole) yields an intense [M+H]+ ion at m/z 102
• Suppresses positive ion electrospray of less basic compounds.
• Tetrahydrofuran (THF)
• 100% THF is highly flammable, so APCI and most interface techniques use nitrogen as the nebulizer gas. (Using air creates an explosion hazard).
• Reacts with PEEK^® tubing.

Ion Suppression and choice of Chemistries

Ion suppression is one of the more visible issues confronting spectrometrists using ESI as the ionization technique. The U.S. Food and Drug Administration’s (FDA) publication, Guidance for Industry on Bioanalytical Method Validation (Federal. Register, 66, 100, 28526) in 2001 indicates the need for such consideration to ensure the quality of analysis is not compromised. The article notes several experimental protocols for evaluating the presence of ion suppression. One compares the multiple-reaction-monitoring (MRM) response (peak areas or peak heights) of an analyte in a spiked, post-extraction sample to that of the analyte injected directly into the neat mobile phase. A low analyte signal in the matrix compared to the pure solvent indicates the presence of interfering entities.

A publication by C. Mallet et al. describes where in the chromatogram matrix effects on the analyte (and internal standard) are present. The experimenters use a continuous flow of a standard solution containing the analyte of interest and its internal standard added to the column effluent. After injecting a blank sample extract into the LC system, a drop in the constant baseline indicates suppression in ionization of the analyte due to the presence of interfering material.

Column chemistries

One enabling change in technology is the advent of hybrid column chemistries and highly selective particles of less than two micrometers’ diameter. The hybrid chemistries rely less on mobile phase modifiers that can cause ion suppression, and the increased selectivity of particles.

Ultra high pressure LC vs. traditional HPLC

The recent commercialization of work by Professor J. Jorgenson (University of North Carolina), often generically referred to as UHPLC (ultra high pressure liquid chromatography), brings a potential to increase the information derived from typical LCMS analyses. Commercialized by Waters Corporation as UPLC^®, or ultra performance liquid chromatography, the increased peak capacity relative to HPLC makes possible defining chemical entities that would otherwise have co-eluted under the broader peaks of HPLC. Concentrating peaks into bands of (typically) two-second widths or less raises the potential for increased sensitivity by favoring the mass spectrometer’s response to improvements in signal-to-noise ratio.

The UPLC concept alters familiar parameters established in traditional separations practice like flow rates, particle sizes—even our appreciation of van Deemter curves. As operating pressure increases from ~2000 psi to as high as 20,000 psi, particle diameters less than 2 µm approach the theoretical limit described in 1969 by John Knox in his "Knox equation". Once the attendant problems of increased mechanical stress and exaggerated thermal effects have been addressed the improvements in MS performance come as a somewhat counterintuitive consequence of theory.