Solvent Properties And Their Analytical Consequences

Relative Permittivity, Polarity Scales, And Protic Character

The relative permittivity of a liquid, still widely called its dielectric constant, expresses how far the medium attenuates the electric field between separated charges relative to vacuum. For a peptide — a polyelectrolyte carrying carboxylates, protonated amines, and a backbone of strongly dipolar amide units — this single number sets much of the energetic cost of holding charges apart. Representative tabulated values near 25 °C are water 78.4, dimethyl sulfoxide 46.7, N,N-dimethylformamide 36.7, acetonitrile 35.9, methanol 32.7, ethanol 24.5, acetone 20.7, 2-propanol 19.9, dichloromethane 8.9, tetrahydrofuran 7.5, and n-hexane 1.9. As permittivity falls, the Bjerrum length rises and counter-ions begin to associate with ionised side chains rather than remaining freely dissociated. A change in organic content therefore alters not only solvation but the effective charge borne by the molecule, which is the reason retention, electrophoretic mobility, and gas-phase charge state all move together when the organic fraction is varied.

Permittivity is nonetheless a bulk continuum property and a weak descriptor of what happens at molecular contact. Empirical polarity scales built from spectroscopic probes capture the missing specificity. Reichardt’s E_T(30) scale, derived from the solvatochromic shift of a pyridinium N-phenolate betaine dye, orders solvents by their combined dipolarity and hydrogen-bond-donating ability and is normalised as E_T^N between tetramethylsilane and water. The Kamlet–Taft treatment separates the effect into three terms: π* for dipolarity and polarizability, α for hydrogen-bond-donor acidity, and β for hydrogen-bond-acceptor basicity. These decompositions explain ordering that permittivity alone cannot: acetonitrile and methanol have similar dielectric constants but behave very differently toward solutes because methanol donates hydrogen bonds strongly and acetonitrile scarcely at all.

The protic–aprotic distinction follows directly from the α term. Protic solvents — water, the simple alcohols, formic and acetic acid, and the fluorinated alcohols — carry an O–H able to donate a hydrogen bond and therefore solvate anions effectively. Dipolar aprotic solvents such as acetonitrile, dimethyl sulfoxide, N,N-dimethylformamide, acetone, and tetrahydrofuran possess appreciable dipole moments and lone pairs but negligible donor acidity; they solvate cations through their donor atoms while leaving anions comparatively bare and correspondingly more reactive. For the peptide backbone, which presents amide N–H donors and carbonyl acceptors, the consequence is competition. In a strongly basic aprotic medium such as dimethyl sulfoxide (β near 0.76, α near zero), solvent molecules engage backbone N–H groups that would otherwise pair intramolecularly, and cooperative secondary structure is destabilised. Water, with α near 1.17 and β near 0.47, satisfies both donor and acceptor sites, so folding and self-association are driven chiefly by the hydrophobic effect. The fluorinated alcohols occupy the opposite corner: 2,2,2-trifluoroethanol and hexafluoroisopropanol are strong donors and very weak acceptors (α roughly 1.5 and 2.0 respectively, β near zero), so they leave backbone carbonyls free to accept intramolecular hydrogen bonds.

Miscibility, Mixed Solvents, And Phase Behaviour

Miscibility with water tracks cohesive energy density and hydrogen-bonding capacity. Methanol, ethanol, 2-propanol, acetonitrile, acetone, tetrahydrofuran, 1,4-dioxane, dimethyl sulfoxide, and N,N-dimethylformamide are miscible with water in all proportions at ordinary temperature; 1-butanol, ethyl acetate, diethyl ether, and dichloromethane are only partially miscible and form two phases over most compositions, each phase carrying a small and temperature-dependent quantity of the other component. Complete miscibility at 25 °C does not guarantee single-phase behaviour at all temperatures: acetonitrile–water mixtures of intermediate composition demix on cooling toward and below 0 °C, which is a recurring source of confusion when mobile phases or samples are held in chilled instrument compartments.

Mixing is not volumetrically additive. Alcohol–water systems show pronounced negative excess volume, so an eluent assembled by combining measured volumes of the two liquids does not have the same composition as one blended by a pump on a volumetric basis, and the two routes can give measurably different retention. Viscosity is likewise non-additive and passes through a maximum in mixed aqueous organics. Methanol–water reaches a viscosity near twice that of either pure component at intermediate composition, while acetonitrile–water peaks much lower and at a smaller organic fraction; this difference is the reason acetonitrile-based gradients run at lower back-pressure than methanol-based gradients on the same column. Electrolyte solubility falls steeply as the organic fraction rises, so buffer salts that are entirely soluble in water can approach their limit and precipitate at high organic content.

The strongly dipolar aprotic solvents are markedly hygroscopic. Dimethyl sulfoxide, N,N-dimethylformamide, and acetonitrile take up atmospheric water rapidly once a container is opened, and water contents of a few tenths of a percent can shift acid–base equilibria, retention, and reaction rates enough to be seen in routine data. Karl Fischer titration is the standard means of quantifying that uptake. Dimethyl sulfoxide has the further peculiarity of a melting point near 18.5 °C, so it solidifies in a cool room, and a high boiling point near 189 °C, which makes it difficult to remove by evaporation and troublesome as a background component in any method that concentrates the sample.

Acid–Base Behaviour And The Meaning Of pH In Mixed Solvents

pH is an operational quantity defined by the response of a glass electrode calibrated against standard buffers, and in a mixed solvent the number it yields depends on which convention was used. Three are in circulation and are conventionally distinguished by superscripts: calibration in aqueous buffers with measurement in the aqueous component before organic is added; calibration in aqueous buffers with measurement in the final mixture; and calibration in buffers made up in the same mixture as the sample. Only the last is thermodynamically referred to the mixed solvent as standard state. Correction terms for moving between these pH scales in acetonitrile–water have been published (reference 3). A pH value reported without the convention attached is therefore incomplete, and comparisons between laboratories that used different conventions are not meaningful at high organic content.

The underlying acid–base equilibria themselves shift with solvent composition, and they do not all shift the same way. Neutral acids — carboxylic acids, including the C-terminus and the aspartate and glutamate side chains — become weaker as acetonitrile is added, because dissociation creates a charge-separated pair that the lower-permittivity medium solvates poorly; the apparent pK_a can rise by more than a unit at intermediate composition and considerably more at high organic content. Cationic acids such as protonated lysine, arginine, and the N-terminal amine dissociate without net charge creation and shift far less, sometimes slightly downward. The result is that a peptide’s net charge, and hence its isoelectric behaviour, is not constant across a gradient even when the nominal mobile-phase pH is held fixed.

Buffer selection in analytical work is governed by three separate requirements that often conflict: adequate capacity within roughly one unit of the buffer pK_a, compatibility with the detector, and solubility across the full composition range used. Phosphate and borate are optically transparent well into the low ultraviolet and are well suited to absorbance detection, but they are non-volatile and cannot be admitted to a mass spectrometer source. Formate, acetate, ammonium bicarbonate, and ammonium formate or acetate are volatile and therefore compatible with electrospray, at the cost of narrower useful pH windows and greater ultraviolet absorbance. The zwitterionic buffers introduced by Good and co-workers were designed against an explicit set of criteria — pK_a in the near-neutral range, high water solubility, minimal metal binding, and chemical stability — and remain the reference point for buffer choice in aqueous biochemical measurement, though their non-volatility rules them out for direct mass-spectrometric work.

Solubility And Conformational State In Different Solvent Environments

Peptide solubility in aqueous media is set principally by net charge, by the fraction of hydrophobic residues, and by the propensity of the sequence to form intermolecular structure. Solubility passes through a minimum near the isoelectric point, where net charge and hence intermolecular electrostatic repulsion are smallest, and generally rises as the solution pH moves away from that point in either direction. Neutral salts modify this in a manner ordered by the Hofmeister series: kosmotropic anions such as sulfate and phosphate lower solubility at high ionic strength, while chaotropic species such as thiocyanate and perchlorate raise it. Sequences with high β-sheet propensity, particularly those with alternating hydrophobic patterns, can pass from apparent solubility into a gelled or fibrillar state over hours without any visible precipitate, which makes concentration determination and chromatographic yield unreliable in ways that are easy to misattribute to the instrument.

Organic cosolvents change conformational populations, not merely solubility, and the effect is well characterised spectroscopically. The fluorinated alcohols shift the equilibrium of many peptides toward helical structure, an effect attributed to weakened solvation of backbone carbonyls and to preferential accumulation of cosolvent around the peptide rather than to any direct templating (reference 5). Dimethyl sulfoxide moves in the opposite direction: with increasing dimethyl sulfoxide content, native secondary structure in α-helical and β-sheet proteins is progressively lost and, at intermediate content, intermolecular β-sheet and aggregation appear (reference 6). The immediate analytical consequence is that circular dichroism spectra, hydrodynamic radii, and even chromatographic peak shapes are properties of the peptide-plus-solvent system, and cannot be compared across solvent environments without saying so.

Solvent and pH also determine the rate at which a peptide changes covalently while it is in solution, which matters because such changes are frequently mistaken for pre-existing heterogeneity. Asparaginyl deamidation proceeds through a cyclic succinimide intermediate and yields a mixture of aspartyl and isoaspartyl products with partial racemisation; the rate depends strongly on the residue following asparagine and rises sharply above neutral pH (reference 7). Aspartyl-prolyl bonds are unusually labile to acid-catalysed cleavage. Methionine, tryptophan, and cysteine are oxidised by trace peroxides, which accumulate in ethers such as tetrahydrofuran and diethyl ether on exposure to air, and free thiols undergo disulfide exchange readily above neutral pH. None of these are instrument artefacts, but all of them are generated during the interval between a sample entering solution and the measurement being made, and the interval is part of the method.

Solvent Selectivity In Reversed-Phase Chromatography

In reversed-phase separations the organic modifier acts on retention through its ability to solvate the hydrophobic solute surface that would otherwise associate with the bonded stationary phase. Acetonitrile is the stronger eluent on a volume-fraction basis than methanol, so the same gradient slope expressed in percent organic produces different elution windows for the two. More important than strength is selectivity, which changes with the identity of the modifier rather than its amount. Snyder’s classification of solvents by their proton-donor, proton-acceptor, and dipolar interaction contributions places methanol, tetrahydrofuran, and acetonitrile in distinct groups, and this is the physical basis of the long-standing practice of substituting one modifier for another to alter peak order when adjusting gradient conditions alone fails to resolve a pair.

Detection imposes its own constraints. Peptide bonds absorb near 210–220 nm, where solvent transparency becomes limiting. Approximate ultraviolet cutoffs are 190 nm for acetonitrile, 205 nm for methanol and 2-propanol, 212 nm for tetrahydrofuran, 233 nm for dichloromethane, and around 268 nm for both dimethyl sulfoxide and N,N-dimethylformamide, with acetone opaque to roughly 330 nm. That ranking, together with its low viscosity, is why acetonitrile dominates peptide separations even where methanol would give better selectivity. Acid modifiers absorb as well: trifluoroacetic acid contributes appreciable absorbance in the low ultraviolet, so a gradient in which its concentration is not matched between the two mobile phases produces baseline drift that can be mistaken for column bleed. Trifluoroacetic acid also acts as a hydrophobic ion-pairing agent for protonated basic residues, which changes selectivity and peak shape independently of its role in setting pH.

The solvent in which a sample arrives at the column head is a variable in its own right. When that solvent is stronger than the starting mobile phase, the sample band is not focused at the inlet and elutes as a broadened, fronting, or split peak, an effect that grows with the ratio of sample-solvent strength to eluent strength and with the volume introduced. This is the most common reason a peak that is symmetrical under isocratic conditions distorts in gradient mode. Related solvent-side effects include dissolved-gas outgassing when a low-solubility mixture is formed at the mixing point, refractive-index disturbances at the solvent front in absorbance detection, and pressure changes traceable to the viscosity maximum of the mixed eluent rather than to any change in the column.

Solvent Effects In Electrospray Ionisation

Electrospray produces gas-phase ions by dispersing a conducting liquid into charged droplets that shrink by evaporation until Coulombic repulsion drives fission and, ultimately, ion release. Every step depends on bulk solvent properties. Surface tension sets the field required to form a stable Taylor cone: water is near 72 mN/m at 25 °C, methanol near 22, and acetonitrile near 29, which is why aqueous streams spray less stably than organic-rich ones and why organic makeup solvent is commonly seen introduced downstream of the column in aqueous-rich methods. Volatility governs the rate of droplet shrinkage, permittivity and conductivity govern charge density on the droplet, and viscosity affects the size distribution produced at the nozzle. Response in electrospray correlates with nonpolar surface area and with reversed-phase retention (reference 10) — a reminder that ionisation efficiency is a property of the analyte–solvent pair, not of the analyte alone.

Additives must be volatile, and among the volatile options they are not equivalent. Trifluoroacetic acid, valuable chromatographically, suppresses electrospray response substantially; the suppression is attributed to gas-phase ion pairing between the trifluoroacetate anion and protonated analyte and to the altered surface tension of the sprayed liquid, and signal can be restored by post-column addition of a weaker acid in 2-propanol (reference 11). Formic and acetic acid are correspondingly weaker ion-pairing agents and are the usual substitutes when mass-spectrometric response is the priority. Non-volatile buffers deposit in the source, foul the sampling orifice, and generate persistent sodium and potassium adducts that distribute analyte signal across several channels and lower apparent sensitivity.

Solvent composition also determines the observed charge-state distribution, which is not a fixed property of the molecule. Acidic, organic-rich conditions that unfold a peptide or protein expose more basic sites and shift the envelope to higher charge and lower mass-to-charge; near-neutral aqueous ammonium acetate preserves compact conformations and yields narrower, lower-charge envelopes. Small quantities of low-volatility additives such as m-nitrobenzyl alcohol or sulfolane shift the distribution to higher charge, an effect generally attributed to their persistence in the late-stage droplet and to their surface tension. Because charge-state distribution is used both for deconvolution and as a conformational readout, the solvent under which a spectrum was acquired is part of the result and belongs with it in any report.

Purity Grades, Trace Contaminants, And Blank Control

Solvent grades differ in which impurities are controlled, not simply in how pure the material is overall. Reagent or ACS grade certifies assay and a list of specified impurities but says nothing about optical background. HPLC and gradient grades add ultraviolet absorbance limits at stated wavelengths, a gradient baseline test that reveals late-eluting absorbing impurities concentrated at the column head during the aqueous portion of a run, and low evaporation residue. LC-MS grade adds constraints that matter only to a mass spectrometer: low trace metals, low extractables from the container and closure, filtration to a fine cut, and testing for background ion intensity in both polarities. Spectrophotometric grade optimises optical transparency without addressing mass-spectrometric background. Water is specified separately; ASTM D1193 Type I reagent water, at 18.2 MΩ·cm resistivity at 25 °C with controlled total organic carbon, is the usual analytical reference point, and resistivity alone does not certify low organic content.

The recurring contaminants are well catalogued. The interferences routinely observed in electrospray and MALDI have been catalogued (reference 12): phthalate plasticisers from tubing and container liners, polyethylene glycol and polysorbate series from detergents and labware, cyclic polysiloxanes from septa and silicone grease, fatty-acid amide slip agents such as erucamide from plastic surfaces, and alkali-metal adducts from glass. Tetrahydrofuran and other ethers are frequently supplied stabilised with butylated hydroxytoluene, which appears as a discrete peak in both ultraviolet and mass-spectrometric traces and which is absent from unstabilised lots — with the corresponding hazard that unstabilised ether accumulates peroxides. Because many of these species are introduced downstream of the bottle, solvent grade alone does not determine the background: filters, frits, transfer tubing, and sample containers all contribute.

Control of that background is empirical rather than documentary. A solvent blank carried through the full gradient and the full detection method, run at the same sensitivity as the samples, is the only reliable way to distinguish an analyte signal from a solvent-borne one, and its validity lapses when a solvent lot, a container type, or a piece of tubing changes. Grade selection follows from the demands of the detection method — a separation monitored at 214 nm and one monitored by high-resolution mass spectrometry at low femtomole levels have different and only partly overlapping requirements — rather than from a general presumption that a nominally higher grade is uniformly better. Where a background peak cannot be eliminated, characterising and documenting it is worth more to the integrity of the data than an unexplained subtraction.

References

1. Reichardt, C. “Solvatochromic Dyes as Solvent Polarity Indicators.” Chemical Reviews 1994, 94(8), 2319–2358. DOI: 10.1021/cr00032a005
2. Marcus, Y. “The properties of organic liquids that are relevant to their use as solvating solvents.” Chemical Society Reviews 1993, 22(6), 409–416. DOI: 10.1039/CS9932200409. Source for the tabulated Kamlet–Taft α, β, and π* values cited.
3. Espinosa, S.; Bosch, E.; Rosés, M. “Retention of Ionizable Compounds on HPLC. 5. pH Scales and the Retention of Acids and Bases with Acetonitrile–Water Mobile Phases.” Analytical Chemistry 2000, 72(21), 5193–5200.
4. Good, N. E.; Winget, G. D.; Winter, W.; Connolly, T. N.; Izawa, S.; Singh, R. M. M. “Hydrogen Ion Buffers for Biological Research.” Biochemistry 1966, 5(2), 467–477. DOI: 10.1021/bi00866a011
5. Buck, M. “Trifluoroethanol and colleagues: cosolvents come of age. Recent studies with peptides and proteins.” Quarterly Reviews of Biophysics 1998, 31(3), 297–355. DOI: 10.1017/S003358359800345X
6. Jackson, M.; Mantsch, H. H. “Beware of proteins in DMSO.” Biochimica et Biophysica Acta 1991, 1078(2), 231–235. DOI: 10.1016/0167-4838(91)90563-F
7. Geiger, T.; Clarke, S. “Deamidation, isomerization, and racemization at asparaginyl and aspartyl residues in peptides. Succinimide-linked reactions that contribute to protein degradation.” Journal of Biological Chemistry 1987, 262(2), 785–794. PMID: 3805008
8. Snyder, L. R. “Classification of the solvent properties of common liquids.” Journal of Chromatography 1974, 92(2), 223–230. DOI: 10.1016/S0021-9673(00)85732-5
9. Snyder, L. R.; Kirkland, J. J.; Dolan, J. W. Introduction to Modern Liquid Chromatography, 3rd ed.; Wiley: Hoboken, NJ, 2010. General reference for eluotropic strength, solvent-strength mismatch at the column inlet, and mobile-phase viscosity effects.
10. Cech, N. B.; Enke, C. G. “Practical implications of some recent studies in electrospray ionization fundamentals.” Mass Spectrometry Reviews 2001, 20(6), 362–387. DOI: 10.1002/mas.10008
11. Apffel, A.; Fischer, S.; Goldberg, G.; Goodley, P. C.; Kuhlmann, F. E. “Enhanced sensitivity for peptide mapping with electrospray liquid chromatography-mass spectrometry in the presence of signal suppression due to trifluoroacetic acid-containing mobile phases.” Journal of Chromatography A 1995, 712(1), 177–190. DOI: 10.1016/0021-9673(95)00175-M
12. Keller, B. O.; Sui, J.; Young, A. B.; Whittal, R. M. “Interferences and contaminants encountered in modern mass spectrometry.” Analytica Chimica Acta 2008, 627(1), 71–81. DOI: 10.1016/j.aca.2008.04.043