Type A vs Type B Silica: Why Basic Analytes Tail

Basic analytes can tail differently on nominally similar HPLC columns because their silica surfaces expose different secondary interaction sites. Older Type A silica carries metal impurities that make residual silanols more acidic than on high-purity Type B silica. Endcapping reduces accessible silanols but does not remove them. Compare a neutral marker, mobile-phase pH and column chemistry before blaming the system.^1,2

What distinguishes Type A silica from Type B silica?

Type A describes the older, less pure silica used in reversed-phase columns before the early 1990s. It contained significant concentrations of metals such as iron and aluminum, which made the residual silanols more acidic; those acidic silanols acted as cation-exchange sites and caused strong tailing of basic compounds.^1,2 Type B silica is much purer, with very low metal content and a far less acidic surface. It still carries residual silanols: a sufficiently basic compound at mobile-phase pH 6–8 will show some tailing even on the highest-purity columns.¹

The distinction matters when a legacy method meets a modern column with the same nominal bonded phase. The ligand name (C18) does not describe the silica underneath, so a column change can sharpen a basic peak while also changing retention and selectivity.^1,2

Table 1 compares the three surfaces. It describes mechanisms, not any manufacturer’s specification.

**Table 1.** Silica substrates and endcapping compared. Qualitative; the individual column’s documentation decides.^1,2
Surface	Metal impurities	Residual silanols	Basic analytes
Type A silica	Significant (iron, aluminum)	Acidic; act as cation-exchange sites	Strong tailing common
Type B silica	Very low	Far less acidic, but present	Tailing much reduced; a strong base can still tail at pH 6–8
Endcapped Type B silica	Very low	Some capped by small silyl groups; about half of the silanols remain unbonded	Further reduced access; not silanol-free

Figure 1 shows the three surfaces side by side, what each means for a basic analyte, and the checks to run when a legacy method moves to a modern column.

Three silica particles with green bonded chains. Type A shows filled red acidic silanols and amber metal impurities; Type B shows red-ring, less acidic silanols and no metal marks (very low metal content); endcapped Type B shows some sites capped by grey silyl groups and the rest as red rings, with no metal marks. Each lists metals, silanols and the effect on basic analytes. Below, four sequential steps for moving a legacy Type A method to a modern column. — **Figure 1. Type A, Type B and endcapped Type B silica.** Type A silica contains significant iron and aluminum, which make its residual silanols acidic; they act as cation-exchange sites and basic analytes commonly tail strongly. Type B silica has very low metal content and far less acidic silanols, but silanols remain, and a strongly basic compound can still tail at pH 6–8. Endcapping caps some silanols with small silyl groups, yet about half of the silanols remain unbonded. When a legacy Type A method moves to a modern column, record the original packing and conditions, compare a representative mixture under controlled conditions, measure retention, peak shape and critical-pair resolution, and re-establish method-specific system suitability. Schematic surfaces, not to scale; site counts are illustrative.

Diagnose this automatically

Peak Tailing Troubleshooter

Distinguish silanol effects, column overload, extra-column volume, and injection-solvent mismatch — systematically, in under ten questions.

Diagnose my peak shape

No account needed

Why do protonated basic analytes interact with residual silanols?

A residual silanol (Si–OH) can lose a proton to become a negatively charged silanolate (Si–O⁻). A protonated amine is attracted to that site far more strongly than to the bonded phase, and that electrostatic interaction causes peak tailing of amine-containing bases.² Where only part of the analyte band meets these stronger sites, the peak develops a prolonged trailing edge.^1,2

Not every basic analyte tails. Peak shape depends on the analyte’s charge, the accessible surface sites and the conditions. When one or a few peaks tail, the cause is most often chemical; when all peaks tail, the problem usually lies before the separation.³ A neutral marker run with the base turns this into a test: if the base tails while the neutral marker stays sharp, investigate the chemistry before taking the system apart.³

What does endcapping fix, and what does it leave behind?

Bonded C18 groups crowd the surface, so only about half of the silanols can react. Endcapping attaches small trimethylsilyl groups to some of the remaining silanols, yet approximately half of the silanols still remain unbonded.¹ Endcapped therefore does not mean silanol-free.

In practice, two endcapped C18 columns should not be presumed identical for a demanding basic analyte. Compare measured peak shape, retention and the critical separation under the actual method instead of relying on the endcapping label.^1,2

How does mobile-phase pH change the interaction?

At low pH, most silanols stay in the neutral Si–OH form, so fewer negatively charged sites are available to attract protonated bases. How far up the pH scale that holds depends on the silica: on Type A silica the fraction of Si–O⁻ increases significantly from around pH 3, whereas on Type B silica most surface silanols remain protonated and neutral at pH 6 or lower.² There is no single universal silanol pKa.

pH changes the analyte too. The same adjustment can alter the base’s ionization and retention while it changes the surface charge, so assess both peak shape and retention, and stay within the column’s documented pH range. Silica-based reversed-phase columns generally tolerate about pH 2–8.¹

The opposite lever is high pH. Well above the base’s pKa, the analyte is mostly neutral and the electrostatic attraction disappears. On a hybrid C18 column at pH 10, Davies and co-workers obtained good peak shape and improved loadability for moderately strong bases, while stronger bases, still charged, gave poor efficiency through interaction with increasingly ionized silanols.⁴ High pH is only an option on a column whose documented range covers it.

Figure 2 compares the three pH regimes qualitatively; it does not assign a universal pKa or calculate ionized fractions.

Three panels for low, intermediate and high pH. Each shows a silica surface with three silanols and a basic analyte above it: at low pH neutral Si–OH and a charged base with no attraction; at intermediate pH some charged Si–O⁻ sites, one attracting the charged base; at high pH on a pH-stable column charged silanols and a neutral base. Under each, a schematic chromatogram of a neutral marker and the base: sharp, base tailing, sharp. Below, a pH ruler from 1 to 12 marking the silica column range 2–8, Type A ionization from about pH 3, Type B silanols neutral to pH 6, and pH 10 on a hybrid C18. — **Figure 2. How mobile-phase pH changes the interaction.** At low pH most silanols are neutral (Si–OH), so a protonated base (BH⁺) meets little electrostatic retention and tails less. At intermediate pH more silanols ionize (Si–O⁻) while the base is still charged, and the base tails. Well above the base’s pKa, on a column rated for high pH, the base is mostly neutral (B): moderately strong bases give good peak shape, while stronger bases can still tail. On the ruler, silica-based columns tolerate about pH 2–8; on Type A silica the Si–O⁻ fraction rises significantly from about pH 3, whereas on Type B silica most silanols remain neutral at pH 6 or lower; at pH 10 on a hybrid C18, moderately strong bases are mostly neutral. Schematic traces, not data; the base keeps the same peak area in every panel. Peak positions are not retention predictions.

A competing amine additive is the legacy fix. Adding triethylamine, often around 25 mM, let it bind to silanols more strongly than most analytes and reduced tailing on Type A columns.¹ With Type B silica it is rarely needed, and Taylor and Stoll describe it as largely obsolete; they recommend understanding why a legacy method used it before keeping it.^1,2 In a manufacturer example on Type A silica, Phenomenex notes that a silanol suppressor generally leads to short column lifetime.⁵

How do I distinguish silanol tailing from a physical peak-shape problem?

Start with the pattern across compounds, not the appearance of one peak. Selective tailing of bases points to residual silanols; tailing, splitting or doubling of all peaks points to a problem before the separation, most often a partially blocked column inlet frit.³ Overload is a third possibility: its tail takes on a right-triangle shape and retention shortens as more sample is injected, and protonated bases can overload at low sample mass at low pH.^3,4

Table 2 turns these patterns into tests. The rows are alternatives; within a row, change one variable at a time and compare the same peak-shape metric before and after.

**Table 2.** Diagnostic decision table. Rows are alternatives; within a row, the steps are sequential.^1,2,3
Observed pattern	First discriminating test	Next test	Verification
Basic analyte tails; neutral marker stays sharp	Lower the mobile-phase pH within the column’s limits	Compare a competing amine additive, or a high-purity or pH-stable column	Reinject both markers; record retention and the same tailing metric
All peaks tail, split or double	Look before the separation: a partially blocked column inlet frit is the most common cause	If tails are right-triangle shaped and retention shortens, inject less sample	Repeat the injection after the correction; compare the same metric
Peak shape changes after a column substitution	Compare the legacy and replacement packing: silica type, endcapping, bonded phase	Compare critical-pair selectivity with a representative mixture	Re-establish method-specific system suitability

Figure 3 follows Table 2 row for row and ends each branch in its verification.

Workflow in two bands. Band A lists three controls: a neutral marker, one stated metric and one variable at a time. Band B shows three alternative branches from Table 2, each with a schematic: a base tailing next to a sharp neutral marker, all peaks tailing with one split peak, and a legacy column and its replacement, labelled. Each branch runs from a first discriminating test to a next test to its verification. — **Figure 3. Selective or universal tailing?** The figure follows Table 2 row for row. Before changing anything, run a neutral marker with the base, record one metric with its convention (5% tailing factor or 10% asymmetry factor) and change one variable at a time. If a basic analyte tails while the neutral marker stays sharp, lower the mobile-phase pH within the column’s limits, then compare a competing amine additive or a high-purity or pH-stable column, and reinject both markers. If all peaks tail, split or double, look before the separation, where a partially blocked column inlet frit is the most common cause; if tails are right-triangle shaped and retention shortens, inject less sample. If peak shape changes after a column substitution, compare the packings and the critical-pair selectivity, then re-establish method-specific system suitability. Schematic traces, not data.

For voids, extra-column volume, sample-solvent mismatch and detector settings, see why your HPLC peaks are tailing. Fronting, splitting and other distortions are covered in poor HPLC peak shapes.

Which peak-shape metric should I report?

Do not mix the 5%-height and 10%-height metrics. The pharmacopoeial metric measures the whole peak width at 5% of peak height and divides it by twice the front half-width; the harmonized USP ⟨621⟩ calls it the symmetry factor, A_S, also known as the asymmetry factor or tailing factor.^3,6 The asymmetry factor common outside pharmaceutical laboratories divides the back half-width by the front half-width at 10% of peak height.³

USP symmetry factor: A_S = W_0.05 / 2d (at 5% height)

10% asymmetry factor: b / f (at 10% height)

Here W_0.05 is the full peak width at 5% height, d is the distance from the perpendicular dropped from the peak maximum to the leading edge at that height (the front half-width), and f and b are the front and back half-widths at the stated height. For a symmetric peak both equal 1; as tailing grows the 10% value rises faster, so the two numbers are not interchangeable.³ Names do not settle it either: IUPAC calls the 10%-height back/front ratio the tailing factor, and defines its asymmetry factor as the full width divided by twice the front width at 10%.⁷ State the convention with every result.

Table 3 sets the three conventions side by side.

**Table 3.** Peak-shape metrics and measurement geometry (f = front half-width, b = back half-width at the stated height).^3,6,7
Metric	Height	Calculation	Note
USP symmetry factor, A_S (also called tailing factor)	5%	W_0.05 / 2d = (f + b) / 2f	Harmonized USP ⟨621⟩; 1.0 = symmetric, above 1.0 = tailing⁶
Asymmetry factor (common usage)	10%	b / f	Called the tailing factor by IUPAC⁷
IUPAC asymmetry factor, A_s	10%	(f + b) / 2f	Same form as the USP factor, different height⁷

The harmonized ⟨621⟩ adds a default peak-symmetry requirement: unless otherwise stated, the symmetry factor of the peak used for quantification in a test or assay is 0.8–1.8.⁶ USP has since announced a revision of that subsection, proposed to limit it to organic-impurity, related-substances and assay tests, with a targeted official date of 1 June 2026.⁸ Check the official text in force and the monograph before applying a limit. A validated method may set its own criterion, and outside compendial methods there is no universal limit. For validated or compendial methods, the applicable procedure and regulatory framework take precedence over the general rules of thumb given here.⁶

Why might a legacy Type A method transfer poorly?

A method developed on a more active silica may rely, deliberately or not, on the retention and selectivity that surface provides, and it may carry additives chosen for it.² Substituting a high-purity endcapped phase can reduce tailing yet change the relative retention of critical analytes. One sharper peak does not establish equivalence.^1,2

Record the original packing and conditions, then compare a representative mixture under controlled conditions: retention, peak shape, critical-pair resolution and method-specific system suitability. For how bonded-phase chemistry changes selectivity, see how stationary-phase chemistry changes selectivity.

Frequently asked questions

Does Type B silica eliminate basic-analyte tailing?

No. Type B silica has far fewer metal impurities and a less acidic surface, but residual silanols remain. A sufficiently basic compound at mobile-phase pH 6–8 can still tail on the highest-purity columns.¹

Is an endcapped C18 column always better for bases?

Not automatically. Endcapping caps some silanols, but about half remain unbonded, and columns differ in silica and bonding. Let the method’s critical separation and system suitability decide.¹

Can lowering the pH fix silanol tailing?

Often it helps: at low pH most silanols are neutral, so protonated bases are attracted less. It can also change retention and selectivity, so compare the whole chromatogram and stay within the column’s pH limits.^1,2

Can high pH fix tailing of bases?

For moderately strong bases it can, because the base becomes mostly neutral; stronger bases may still tail. Use only a column rated for high pH, since silica-based columns generally tolerate about pH 2–8.^1,4

Should I add triethylamine to the mobile phase?

Rarely, on modern Type B columns. It was the standard fix for Type A silica; today it is largely obsolete, and silanol suppressors can shorten column lifetime.^2,5

Can I use the same tailing limit on every method?

No. First check whether the method uses the 5%-height symmetry (tailing) factor or a 10%-height asymmetry factor, then apply the limit set by that validated procedure or monograph. For compendial tests and assays, the harmonized USP ⟨621⟩ sets a default of 0.8–1.8 unless the monograph states otherwise, and that subsection is under revision, so check the official text in force.^3,6,8

The takeaway

When a basic analyte tails, first ask whether the tailing is selective. Type A, Type B and endcapped Type B surfaces differ in how acidic and accessible their silanols are. Use a neutral marker, a controlled pH comparison and a deliberate column comparison to test the silanol hypothesis; if all peaks tail, look before the separation instead.^1,2,3

References

Dolan JW. What’s happening to my column? LCGC North America. 2006;24(9). https://www.chromatographyonline.com/view/whats-happening-my-column-0 — Type A and Type B silica, iron and aluminum, silanol acidity, endcapping (about half of the silanols unbonded), triethylamine, the pH 2–8 range of silica columns, tailing at pH 6–8 on high-purity columns.
Taylor T, Stoll DR. Methods from Mars? Coping with chromatographic legacies. LCGC Europe. 2019;32(5). https://www.chromatographyonline.com/view/methods-mars-coping-chromatographic-legacies-0 — metal content of Type A and Type B silica, Si–O⁻ formation versus pH on each, electrostatic retention of protonated bases, triethylamine and legacy methods.
Dolan JW. Troubleshooting basics, part IV: peak shape problems. LCGC North America. 2012;30(7). https://www.chromatographyonline.com/view/troubleshooting-basics-part-iv-peak-shape-problems — chemical versus all-peak tailing, blocked inlet frit, overload signs, 5% tailing factor versus 10% asymmetry factor.
Davies NH, Euerby MR, McCalley DV. Analysis of basic compounds by reversed-phase high-performance liquid chromatography using hybrid inorganic/organic phases at high pH. J Chromatogr A. 2008;1178(1–2):71–78. https://doi.org/10.1016/j.chroma.2007.11.040 — peak shape and mass overload of bases on a hybrid C18 from pH 2.7 to 12.
Phenomenex. Overcoming peak tailing of basic analytes: silica Type A stationary phases RP. https://www.phenomenex.com/resources/knowledge-center/hplc-knowledge-center/overcoming-peak-tailing-of-basic-analytes-in-silica-type-a-stationary-phases-in-rp (accessed 2026-10-09) — manufacturer example: low pH, triethylamine, short column lifetime.
United States Pharmacopeia. General Chapter ⟨621⟩ Chromatography, Stage 4 harmonized text (document M99380), official 1 December 2022. USP–NF. https://doi.usp.org/USPNF/USPNF_M99380_07_01.html — symmetry factor A_S = W_0.05/2d (also known as asymmetry or tailing factor); peak symmetry 0.8–1.8 for the quantified peak unless otherwise stated.
IUPAC. Compendium of Chemical Terminology (Gold Book): asymmetry factor, https://doi.org/10.1351/goldbook.09902; tailing factor, https://doi.org/10.1351/goldbook.09905; from Terminology of separation methods (IUPAC Recommendations 2017) — the IUPAC 10%-height definitions.
United States Pharmacopeia. ⟨621⟩ Chromatography: Notice of Intent to Revise, 27 September 2024, https://www.uspnf.com/notices/621-nitr-20240927; in-process revision (Pharmacopeial Forum), https://doi.usp.org/USPNF/USPNF_M99380_80201_01.html — System Sensitivity and Peak Symmetry subsections to be revised; Peak Symmetry proposed to apply only to organic-impurity, related-substances tests and assays; targeted official date 1 June 2026.

Type A vs Type B Silica and Endcapping: Why Basic Analytes Tail Differently