Usage Notes

HPLC ColumnsUsage Notes

From goods-in inspection to the decision to retire a column, covering general questions on HPLC and UPLC columns, end fittings, guard columns, fitting screws and tubing. With 20 cross-sections, exploded views and perspective drawings, focusing on fitting specifications, understanding 1/16″ and 1/8″ tubing, where PEEK and stainless steel each apply, and locating leaks.

Part One · Overview and Hardware

HPLC Column Usage Notes

Between delivery and retirement, a column passes through four routine stages — goods-in inspection, installation, routine operation and shutdown storage — plus two non-routine states: leaks and performance problems. Most problems in use come not from the medium itself but from the connections outside the column: a mismatched fitting standard, the wrong insertion depth, the wrong tubing bore, a reused ferrule. This page sets these out in sequence and gives cross-sections and exploded views for the fittings and tubing.

← Swipe to see the full diagram →

Fig. 1 · Column lifecycle swimlane diagram Time runs along the horizontal axis and responsibility down the vertical; red nodes mark where problems concentrate Goods-in inspection Installation Routine operation Shutdown storage Deviation Handling Delivery First run Batch running Long-term use Retirement Appearance and end plugs Check against the QC report Common: storage solvent evaporated / flow arrow not confirmed / fitting standard does not match the system Select fittings Cut tubing and fit the ferrule Install in the direction of flow Equilibrate the system Common: dead volume from mismatched insertion depth / leaks from a reused ferrule / crushed PEEK threads from over-tightening Pressure monitoring pH / temperature limits Sample preparation Guard column replacement System suitability Common: particulates accumulating at the head / buffer salt precipitating / pH and temperature limits exceeded / guard column overdue for replacement Flush out salt Change to storage solvent Cap both ends Common: reversed-phase column stored long term in pure water / buffer salt not fully flushed out and crystallizing Leaks Abnormal pressure Peak shape degrading Retention drifting Regeneration / backflushing Retirement and replacement Red boxes within a lane are the problem sets covered on this page; grey boxes are actions.

How to read it:read down and the problems group by responsibility; read across and the focus differs at each stage of the same column's life. Confirmation at goods-in is cheap while investigation during operation is expensive — confirming the fitting standard and flow arrow on receipt avoids most later rework on connections.

Extra-column factors account for a large shareA considerable proportion of peak shape and retention problems come from the connecting tubing and fittings rather than from a failed medium. Confirm extra-column factors first.
A ferrule cannot be undoneOnce tightened, a stainless steel ferrule deforms and bites into the tubing; moved to a fitting of different insertion depth, it usually will not seal again.
The size names are easily confused1/16″ and 1/8″ refer to the outside diameter and say nothing about the bore. The bore determines extra-column volume; the outside diameter determines which fitting to use.
Each material has its limitsPEEK allows tool-free work but has limited pressure and solvent resistance; stainless steel takes high pressure but is sensitive to halides and strong acids.

1 · Column body and end fitting anatomy

An analytical HPLC column consists of the column tube, an end fitting at each end, frits, seals and the packed bed. Understanding how these parts relate is the basis for locating a leak or a source of dead volume.

← Swipe to see the full diagram →

Fig. 2 · Longitudinal section of an analytical column A 4.6 × 250 mm stainless steel column, sectioned lengthwise, flow left to right Packed bed (bonded silica / polymer substrate) Mobile phase inlet To the detector Inlet fitting An inlet frit that is Outlet frit Outlet fitting Column tube (316L stainless steel / PEEK-lined) Frit pore size Usually 1/3 to 1/2 of the particle size — a 2 µm frit for 5 µm medium The most common place for pressure to start rising The inlet frit retains particulates, and the pressure differential appears before efficiency falls What happens if it is installed backwards The bed is looser at the outlet end, and reverse flow drives the bed surface into a void

Key point:the two ends of the column tube are not symmetrical. During packing the material is driven in from the inlet, so the bed near the outlet is less dense. The arrow on the column marks the direction in which the bed was consolidated; flowing with the arrow keeps the bed under compression, while reverse flow applies force in the direction that loosens it.

1.1 Function and failure mode of each part

PartFunctionCommon failureExternally observable sign
Column tubeA pressure vessel that constrains the bedInternal corrosion, halide pittingMetal ion interference on the baseline, worse peak tailing
An inlet frit that isStops particulates and distributes the flowBlockage, local foulingColumn pressure rises progressively, with efficiency changing later
Outlet fritRetains the packingDamage, packing passing throughRising detector noise, blockage downstream
End fittingProvides the sealing face and the flow path transitionScratched sealing face, damaged threadsWeeping at the fitting; still leaks after tightening
Packed bedProvides retention and selectivityBed collapse, voids, chemical degradationPeak splitting, doublets, retention shifting earlier

1.2 Exploded view of the end fitting

The order of parts at the end fitting determines whether the connection seals and whether dead volume is left. Below is the exploded arrangement of a common compression fitting.

← Swipe to see the full diagram →

Fig. 3 · Exploded view of a column end fitting Assembly order left to right; the section below shows it fully seated ① Tubing 1/16″ or 1/8″ outside diameter ② Nut 10-32 UNF or another standard Provides the axial clamping force ③ Ferrule Bites into the tubing wall under compression Deforms irreversibly ④ Fitting body Contains the sealing cone and the flow path ⑤ Frit ⑥ Column tube and packed bed Section with everything seated Where the tube end meets the sealing cone Any gap here becomes dead volume The tube end must be square and seated fully The ferrule seals only; it does not locate the tube

Key point:A compression fitting seals through the ferrule and cone together, while the tube is located by being seated against the bottom of the fitting. If the tube is not pushed fully home before tightening, the ferrule bites at the wrong position, and no amount of retightening afterwards will close that gap.

Once tightened, a ferrule deforms plastically onto the tubing. Moving the same tubing with its ferrule to a fitting of different insertion depth normally means a new ferrule and a fresh cut, not reusing what is there.

2 · Differences between HPLC and UPLC columns

The separation mechanism is the same; the differences lie in particle size, pressure rating and sensitivity to extra-column volume. Fitting a UPLC column to an HPLC system often means its efficiency advantage never appears, because system extra-column broadening is too large.

← Swipe to see the full diagram →

Fig. 4 · Scale comparison of HPLC and UPLC columns Top: column geometry to scale. Bottom: particle size and relative sensitivity to extra-column volume HPLC column 4.6 × 250 mm, 5 µm Column pressure usually < 40 MPa Flow rate 0.8–1.5 mL/min UPLC column 2.1 × 100 mm, 1.7 µm Column pressure can exceed 100 MPa Flow rate 0.2–0.6 mL/min Tolerance for extra-column volume (relative, schematic) HPLC: the column volume is large, so it is relatively insensitive to extra-column volume UPLC The column volume is an order of magnitude smaller, so the same fitting gap takes a much larger share of the peak width HPLC peak Ideal UPLC peak UPLC with extra-column broadening The same extra-column volume affects a narrow peak far more than a broad one.

Key point:The volume of a UPLC column is of the order of a tenth that of a conventional HPLC column. If the bore and length of the connecting tubing are carried over from an HPLC configuration, extra-column broadening takes up a large share of the peak width budget and the measured efficiency falls short of what the column can do.

DimensionHPLC columnUPLC / UHPLC columnOperational consequence
Particle size3–5 µm1.5–2.0 µmHalving the particle size raises column pressure roughly fourfold
Usual bore3.0 / 4.6 mm1.0 / 2.1 mmThe bore sets the linear velocity and the upper limit on injection volume
Pressure ratingGenerally within 40 MPaCan exceed 100 MPaFittings and tubing must be rated to match
Inlet frit pore sizeAbout 2 µmAbout 0.2–0.5 µmThe finer the pore, the more sensitive to particulates
Connecting tubing bore0.17 mm (0.007″) is usual0.10 mm (0.004″) or finerCarrying over wider tubing causes marked broadening
Sample preparation0.45 µm membrane0.22 µm membraneInadequate filtration accelerates frit blockage

The particle size and pressure relationship above follows the usual consequence of Darcy's law (column pressure varies inversely with the square of particle size); the actual figures vary with column length, flow rate and mobile phase viscosity.

3 · Guard column systems

A guard column takes on particulates and strongly adsorbed components in a low-cost replaceable unit, slowing degradation of the analytical column's inlet frit and the front of the bed. The three common forms trade dead volume against replacement cost.

← Swipe to see the full diagram →

Fig. 5 · How the three forms of guard column are assembled Flow left to right; the dashed box is the replaceable unit A · In-line frit filter Analytical column The smallest addition to dead volume Intercepts particulates only, with no chemical retention B · Guard cartridge with a reusable holder Analytical column The cartridge is cheap, which is economical when replacement is frequent The holder and cartridge must come from the same supply system C · One-piece guard column Replaced as a whole Analytical column Two fittings, so a relatively larger addition to dead volume Simple to fit, and suits the method development stage The guard column packing should have the same phase chemistry as the analytical column, or it introduces an additional difference in selectivity.

Which to choose:for a relatively clean matrix where efficiency comes first, choose A; for large daily sample numbers with frequent replacement, choose B; where the method is still being adjusted and needs flexible assembly, choose C. All three add extra-column volume, so under UPLC conditions A or a low dead volume version of B is preferable.

Replacement CriteriaUse the rise in column pressure as the main indicator. Replacing at about 20% above the initial value allows intervention before efficiency falls.
Match the phase chemistryThe guard column should be C18 or the same class of bonded phase as the analytical column, so no different retention behaviour is introduced ahead of it.
Not a substitute for sample preparationA guard column takes on residual particulates. Filtration and centrifugation must still be done before injection.
Record when it is changedReplacing a guard column shifts retention slightly; recording when it was done helps distinguish method drift from a hardware change.
Part Two · Fittings, Threads and Tubing

What 1/16″ and 1/8″ actually mean

What are commonly called "1/16 tubing" and "1/8 tubing" in an LC system refer to the tubing'soutside diameter, which determines the fitting and ferrule to use; theInternal Diameteris stated separately and determines the internal volume and extra-column broadening. The two are independent, and several bores are available at the same outside diameter. Confusing them is a fairly common source of error in ordering and connecting.

← Swipe to see the full diagram →

Fig. 6 · Outside diameter versus bore (perspective and end views) Left: three bores in 1/16″ outside diameter tubing; right: 1/8″ outside diameter; drawn enlarged and schematic Outside diameter 1/16″ = 1.59 mm Length determines internal volume, together with the bore Common bores at 1/16″ outside diameter (end section) 0.10 mm(0.004″) UHPLC pre-column connection 0.17 mm(0.007″) Conventional HPLC pre-column 0.50 mm(0.020″) Waste and low-pressure sections Outside diameter 1/8″ = 3.18 mm Common bores 1.0 / 1.6 / 2.0 mm 1/8″ outside diameter, 1.6 mm bore Used for preparative LC, pump inlets, solvent delivery and high-flow waste lines Not used for pre-column connections on analytical columns The end sections are drawn enlarged and are not to a common scale; follow the supplier's stated dimensions for actual selection.

Key point:The outside diameter determines whether it fits; the bore determines how well it runs. Both must be given when ordering — for example "1/16″ OD × 0.17 mm ID PEEK tubing". Saying only "1/16 tubing" is not enough to specify what is needed.

4.1 Orders of magnitude for internal volume

The volume of tubing before and after the column adds directly to the peak width budget. Below are the volumes per 10 cm of length at common bores, for estimating the volume cost of a connection scheme.

Internal DiameterImperialVolume per 10 cmTypical useNotes
0.064 mm0.0025″About 0.3 µLUHPLC pre-columnBlocks easily; needs 0.22 µm filtration
0.10 mm0.004″About 0.8 µLUPLC pre-columnThe usual configuration for 2.1 mm bore columns
0.13 mm0.005″About 1.3 µLUPLC / narrow-bore HPLCBalances pressure drop against broadening
0.17 mm0.007″About 2.3 µLConventional HPLC pre-columnThe usual configuration for 4.6 mm bore columns
0.25 mm0.010″About 4.9 µLColumn outlet to detectorA low-pressure section; the broadening effect must still be assessed
0.50 mm0.020″About 19.6 µLWaste, pump to injection valveNot for the critical flow path before or after the column
1.6 mm1/16″ boreAbout 201 µLPreparative LC, solvent deliveryUsually with 1/8″ outside diameter tubing

Volumes are calculated as V = π(d/2)²L from the nominal bore, without allowing for the internal volume of the fittings. Measured values vary with tubing tolerance.

How to estimate: keeping the combined volume of tubing before and after the column within a suitable proportion of the column volume reduces broadening. For a 2.1 × 50 mm column (column volume about 173 µL at a porosity of 0.65), 30 cm of 0.17 mm bore tubing before the column alone is about 6.8 µL; switching to 0.10 mm bore brings it down to about 2.4 µL.

5 · Thread standards and fitting geometry

Thread standards for LC fittings are not universal. Even for the same 1/16″ tubing, the column end, injection valve end and detector end may use different threads and different cone depths. With the wrong standard, either the thread will not engage, or it engages but the sealing faces do not meet and the joint weeps.

← Swipe to see the full diagram →

Fig. 7 · Sections through common thread standards Schematic sections, enlarged; the thread profiles are indicative and not a machining reference 10-32 UNF The mainstream in analytical LC, with 1/16″ tubing 32 threads per inch Major diameter 4.83 mm 1/4-28 UNF Low-pressure and preparative, with 1/16″ or 1/8″ tubing 28 threads per inch Major diameter 6.35 mm 6-40 UNF A miniaturized fitting used on some UHPLC systems 40 threads per inch Major diameter 3.50 mm M6 × 1 (metric) Used on some European and Chinese preparative systems Pitch 1.0 mm Major diameter 6.00 mm Two pairs that look alike but are not interchangeable 10-32 UNF and M5 × 0.8: the major diameters are close, so the first few threads can be forced, after which the internal thread is damaged and the sealing cone still will not meet. 1/4-28 UNF and M6 × 1: they look and feel similar, and mixing them shows up as persistent weeping even when fully tightened. If the first two threads feel tight on entry, stop and check the standard.

Suggested practice:keep a fitting list for each instrument, recording the thread standard and cone depth at the injection valve outlet, before the column, after the column and at the detector inlet. Check the list before changing column brand, then decide whether an adapter is needed.

5.1 How a ferrule seals

As it is tightened, a ferrule is squeezed by the cone, contracts radially and bites into the tubing wall to form a sealing band. This deformation is plastic and does not spring back when the fitting is removed.

← Swipe to see the full diagram →

Fig. 8 · Section through a ferrule before and after tightening Top: not tightened. Bottom: fully tightened Before tightening There is a gap between the ferrule cone and the fitting cone The tube is not seated, leaving a cavity ahead of it After tightening The ferrule contracts radially and bites into the tubing wall to form a sealing band The tube is seated against the bottom of the fitting, and the flow path is continuous with no cavity The deformation is irreversible. Once removed, the ferrule's axial position on the tubing is fixed, which determines the insertion depth available thereafter.

Order of operations:push the tubing to the bottom of the fitting and hold axial pressure, then screw the nut in finger-tight, and finally tighten by the specified number of turns or torque. Reverse the order and the ferrule bites with the tube unseated, creating the cavity shown in the upper half of the figure.

Ferrule materialReusableIndicative pressure ratingSuited toCaution
Stainless steelNo (the position is fixed)HighHigh-pressure pre-column, UHPLCChanging fitting normally means a new ferrule and a fresh cut
One-piece PEEKA limited number of timesMedium (within about 34 MPa)Conventional HPLC, biological samplesSealing force falls after repeated assembly
Two-piece PEEKYesMediumWhere frequent assembly is neededThe orientation must not be reversed
Finger-tight, no toolsYesMedium to lowMethod development, temporary connectionsOver-tightening crushes the thread

The indicative pressure ratings are the usual published ranges; follow the technical data for the product in use.

6 · Insertion depth differences and dead volume

The distance from the sealing cone to the bottom of the flow path — the insertion depth, or stub length — differs between brands. Moving a length of tubing with a ferrule already swaged onto it to a fitting of a different standard gives one of two outcomes: the tube does not reach the bottom, or it goes in too far.

← Swipe to see the full diagram →

Fig. 9 · Sections through three insertion depth conditions Flow left to right; the red area is the dead volume created A · Insertion depth matched The tube end is flush with the flow path, with no stagnant region Peaks are symmetrical and the plate number reaches the stated value B · Tube not seated (ferrule too far back) Stagnant cavity The flow path expands abruptly, creating eddies and stagnation Shows up as peak tailing, a lower plate number and poorer gradient reproducibility C · Tube inserted too far (crushing the sealing face) The ferrule cannot meet the cone and the seal fails Shows up as persistent weeping even when fully tightened, or damage to the fitting's sealing face

How to tell:disconnect and look at the tube end. If the length protruding beyond the ferrule does not match the fitting's stated insertion depth, it is condition B or C. Before connecting across brands, use a universal adjustable fitting, or cut the tubing afresh and swage a new ferrule to the target fitting.

← Swipe to see the full diagram →

Fig. 10 · How dead volume affects the chromatographic peak Left: flow field. Right: the corresponding peak Continuous flow path Symmetrical peak A stagnant cavity is present Some solute is held back and rejoins late Tailing peak The effect is proportional to column volume The same 2 µL stagnant cavity: 4.6 × 250 mm column (column volume about 2.7 mL) A small proportion; the change in peak shape is hard to notice 2.1 × 50 mm column (column volume about 0.17 mL) The proportion rises by about an order of magnitude and the tailing is visible Column volume estimated at a porosity of 0.65

Order of investigation:when tailing appears, first replace the column with a zero dead volume union and inject, to see the peak shape of the system alone. If the peak is normal with the union but tails with the column fitted, the problem is at the column or its fittings; if it still tails with the union, the problem is in the system flow path.

7 · Choosing between PEEK and stainless steel tubing

The two materials have different limits in pressure, solvent resistance, biocompatibility and ease of use, and are usually selected by flow path section and sample type rather than standardized across the whole system.

← Swipe to see the full diagram →

Fig. 11 · Where PEEK and stainless steel tubing each apply Top: appearance and end section. Bottom: relative performance across six dimensions PEEK tubing Can be cut by hand and sealed with a finger-tight fitting No metal ion leaching, suiting biological samples 316L stainless steel tubing Requires a proper cutter, with a demanding requirement for a square end High pressure rating, suiting UHPLC pre-column Maximum pressure PEEK about 34 MPa (falling with larger bore and higher temperature) Stainless steel above 100 MPa Solvent range Most common solvents are fine; tetrahydrofuran, dichloromethane, dimethyl sulfoxide and concentrated nitric acid are not Organic solvents are generally fine; halide ions and strong acids readily cause pitting Biocompatibility No metal surface, so adsorption of proteins and phosphorylated compounds is lower The metal surface can interact with chelating analytes Ease of use Cut by hand, fitted finger-tight, and readily readjusted Requires a tubing cutter and torque control The bar lengths are relative and not to a linear scale; the pressure figures vary with bore, wall thickness and working temperature, so follow the product documentation.

Common configuration:stainless steel or reinforced PEEK (PEEKsil, fused-silica lined) for the high-pressure section before the column; PEEK from the column outlet to the detector; for biological samples and phosphorylated compounds, PEEK throughout or tubing with an inert surface treatment; with halide-containing mobile phases, avoid leaving them standing in stainless steel tubing.

PEEK swells in tetrahydrofuran, dichloromethane and dimethyl sulfoxide, showing up as a changed bore and weeping at the fittings, and degrades in concentrated nitric or sulfuric acid. Methods using these solvents call for stainless steel or reinforced tubing.

7.1 Effect of the cut end

← Swipe to see the full diagram →

Fig. 12 · Comparison of tubing end quality Squareness and burrs directly affect the seal and the flow field An acceptable end Square to the axis, no burrs Meets the cone all the way round An angled end One side contacts first, so the seal is uneven Shows up as intermittent weeping An end with burrs The burr projects into the flow path and disturbs it It may break off and enter the column head Remedy Use a proper cutter and deburr stainless steel tubing; cut PEEK squarely in one pass with a sharp blade, avoiding repeated sawing that crushes the end. Inspect the end visually after cutting, with a magnifier if necessary.

Key point:End quality problems often do not show at low pressure, appearing as intermittent weeping only as pressure rises or during a gradient, where they are easily mistaken for a fitting or column problem. Checking the end before changing the fitting saves wasted investigation.

Part Three · Goods-in, Installation, Operation and Storage

8 · Goods-in inspection

Confirmation at goods-in is cheap, while the cost of an omission is magnified later. There is a set of items to confirm before unpacking a new column and another afterwards. Checking the fitting standard and flow arrow at this point avoids rework at installation.

← Swipe to see the full diagram →

Fig. 13 · Inspection points for a new column How the markings on the column relate to the accompanying documents HPLCONE C18 5 µm 4.6 × 250 mm S/N 2026-0731-08 Storage solvent: acetonitrile / water (65:35) Flow direction ① End plugs ② Grade / dimensions / lot number ③ Flow arrow ④ Fitting thread standard ⑤ Condition of the tube and legibility of the markings What to read from the accompanying QC report · Theoretical plate number N and the test conditions · Asymmetry factor As or tailing factor T · Probe compound and mobile phase used · Test flow rate and column temperature · Reference column pressure · Lot number and date of manufacture The first run must reproduce the conditions listed in the report, or the data will not be comparable Order of confirmation before and after unpacking Outer packaging intact End plugs in place Label checked against the order Fitting standard checked First run and re-measurement

Key point:End plugs coming loose is a fairly common problem in transit. With a plug missing, the storage solvent evaporates and the bed dries locally, so a void can form when liquid is passed through again. If a plug is found missing on arrival, or there are obvious signs of liquid having leaked from the tube, contact the supplier before running the column.

CheckAcceptable conditionAbnormal signRemedy
Outer packagingNo crushing, no liquid marksPackaging deformed, liner dampPhotograph it and contact the supplier
End plugsBoth in place and tightLoose or missingInform the supplier before running; do not simply pass liquid through
Column tube appearanceNo dents or scratchesTube deformedDo not run; go through the returns process
Label informationMatches the orderPhase, particle size or dimensions do not matchCheck against the order and hold off using it
Fitting standardMatches the existing systemDifferent thread or insertion depthPrepare the matching fitting or an adapter
QC reportLot number matches the columnLot number does not match, or pages missingRequest the report for that lot
First re-measurementN and As close to the reported valuesClearly differentInvestigate extra-column factors in the system first, then judge the column
A discrepancy on first re-measurement should not be attributed to the column straight away. Test conditions, system extra-column volume, detector time constant and data acquisition rate all affect the measured plate number. Use a known good column on the same system as a reference.

9 · Installation and connection

← Swipe to see the full diagram →

Fig. 14 · Installation steps and risk points A six-step procedure, with the common errors at each step noted below ① Check the flow direction The arrow points towards the detector ② Select the fitting Standard and insertion depth ③ Cut and deburr A square end ④ Seat, then tighten Hold axial pressure ⑤ Leak-test at low flow Starting at 0.2 mL/min ⑥ Equilibrate to a stable baseline 10–20 column volumes Reverse installation applies force in the direction that loosens the bed Reusing an old ferrule on a fitting of a different standard An angled cut or a burr causing intermittent weeping Tightening before seating creates dead volume before the column Going straight to a high flow rate Bed Injecting before equilibration is complete Retention drifts over the first few injections Judging the tightening force Stainless steel ferrule: once finger-tight, tighten by the supplier's stated number of turns (commonly 1/2 to 3/4). After the first swage the ferrule position is fixed, so a reinstallation needs only finger-tight plus about 1/8 turn. PEEK finger-tight fittings: tighten with finger pressure only. If it still weeps, remove it and check the end and insertion depth rather than applying more force — more force will crush the thread and ruin the fitting. Signs of over-tightening: the thread suddenly feels loose on entry, it still weeps when fully tightened, or the thread is deformed or the ferrule cracked on removal. Replace the fitting and ferrule if any of these appear.

How to leak-test:after connecting, start the flow at 0.2 mL/min and wipe around the fitting with lint-free paper, looking for a damp mark. Raise gradually to the working flow rate, repeating the check as the pressure rises. No leak at low flow but a leak at high flow usually points to the end quality or an under-tightened ferrule.

10 · Everyday operating limits

Column life is governed mainly by chemical conditions and particulate load. The four limits below apply to most bonded silica columns; follow the technical data for the product in use for exact figures.

LimitConventional bonded silicaBase-stable bonded phasePolymer substrateWhat happens outside the limit
pH range2.0–8.01.5–12.01–14Hydrolysis of the bonded phase at low pH; dissolution of the silica skeleton at high pH
Maximum temperatureAbout 60 °CAbout 80 °CAbout 80 °CHigh temperature accelerates hydrolysis and markedly shortens life
Maximum pressureAs rated for the columnAs rated for the columnUsually lower than silica columnsBed consolidation or frit deformation
Fully aqueous conditionsC18 chains may collapse at high aqueous contentAs leftUnaffectedRetention falls sharply and re-equilibration is needed to recover

pH and temperature act together more strongly than either alone. Operating near the pH limit at an elevated temperature shortens life by more than the sum of the two effects individually (inferred on this page; no unified quantitative model was found).

10.1 Using and switching buffers

  • Filter through 0.45 µm after preparation (0.22 µm for UPLC), to keep insoluble matter out of the system.
  • Salt concentration and organic proportion must be considered together for solubility. Phosphate solubility falls as the acetonitrile proportion rises, and it readily precipitates at the mixing point.
  • Before switching from a salt-containing mobile phase to a high organic content, flush for about 10 min with 5%–10% organic in water, then raise the organic proportion in steps.
  • Flush out the buffer salt before shutdown storage, to avoid crystallization blocking the frit or corroding stainless steel flow paths.

10.2 Sample preparation

Filtration grade0.45 µm for HPLC, 0.22 µm for UPLC. The membrane material must be compatible with the sample solvent.
Matching solvent strengthWhere the injection solvent is stronger than the starting mobile phase, early peaks broaden or split. Dissolve the sample in the starting mobile phase.
Matrix removalFor matrices such as serum and fermentation broth, use solid phase extraction or protein precipitation first; a guard column alone is not enough.
Injection volumeToo large a volume overloads the column head and broadens peaks. The acceptable injection volume on a narrow-bore column scales with the square of the bore.

11 · Storage and care

← Swipe to see the full diagram →

Fig. 15 · Shutdown storage procedure Branching by phase type; dashed lines are the additional steps where buffer salt has been used End of run Reduce to a low flow rate If buffer salt has been used 5–10% organic in water Flush for about 10 min to remove salt Raise to 50% organic A transition, avoiding an abrupt change Reversed-phase columns (C18 / C8 / phenyl) Neat acetonitrile or methanol HILIC columns 90% acetonitrile / water SEC / GPC columns The storage solution specified for the medium Ion exchange columns Dilute salt solution with a preservative Cap both ends Room temperature, away from light Storage practices to avoid Storing a reversed-phase column in pure water for long periods (microbial growth and C18 chain collapse); sealing it with buffer salt still present; storing it where the solution inside could freeze.

Duration:for a shutdown of a few days, cap the column under the organic conditions of the last run; beyond a week, change to the storage solvent shown above. No periodic flushing is needed during long-term storage.

Phase typeShort term (days)Long term (over a week)Avoid
C18 / C8 / phenylThe last mobile phase used (salt removed)Neat acetonitrile or methanolProlonged storage in pure water
Hydrophilic C18 / polar-embeddedThe last mobile phase usedAcetonitrile / water (65:35) or per the instructionsStrongly alkaline solutions
HILICThe last mobile phase used90% acetonitrile / waterHigh aqueous content
Normal-phase silicaHexane / isopropanolHexaneSolvents containing water
Ion exchangeThe last buffer usedDilute salt solution with a preservativeAqueous solutions without preservative
SEC / GPCThe last mobile phase usedAs specified for the mediumChanging solvent systems without a transition

11.1 Regeneration and backflushing

Where the pressure has risen and does not recover after replacing the guard column, regeneration can be tried. Backflushing (reverse flow) can push out particulates retained by the inlet frit, but it applies force to the bed in the direction that loosens it.

  • Disconnect the column from the detector before backflushing and route the effluent to waste, so particulates do not reach the flow cell.
  • Use less than half the normal flow rate, for of the order of 10 min.
  • Re-equilibrate in the forward direction afterwards and re-measure plate number and asymmetry factor to confirm the bed is unaffected.
  • Some suppliers do not recommend backflushing their columns. Check the technical data for the product in use before proceeding.
Strong-solvent flush sequence (reversed-phase columns): water → methanol → isopropanol → dichloromethane → isopropanol → methanol → water, 10–20 column volumes each. Confirm tubing compatibility for the dichloromethane steps (PEEK tubing is not suitable).
Part Four · Leaks, Troubleshooting and Quick Reference

12 · Leaks

A leak gives fairly direct feedback but has a wide range of causes. The same damp fitting may come from an under-tightened ferrule, an angled end, a mismatched insertion depth or a scratched sealing face — or from an upstream component with the liquid running along the tubing to that point. Working through the points one by one is more effective than repeated tightening.

← Swipe to see the full diagram →

Fig. 16 · Where leaks occur in the system Flow order left to right; the dots mark common leak points Pump Plunger seal Injection valve Sample loop fitting Guard Column Analytical column Detector Flow cell 1 2 3 4 5 6 7 8 Points and what to check first 1 · Pump outlet fitting — a high-pressure section; check the ferrule and end 2 · Injection valve inlet — often a different standard from the pump end 3 · Injection valve outlet — an aged rotor seal also shows up here 4 · Guard column inlet — assembled often, so the ferrule fatigues 5 · Analytical column inlet — higher pressure, so leaks are more likely 6 · Analytical column outlet — low pressure, usually an end or insertion depth problem 7 · Detector inlet — over-tightening readily cracks the flow cell fitting 8 · Pump head plunger seal — shows as weeping below the pump head, not a fitting problem How to locate it Liquid runs down the tubing under gravity, so the visible point may be below the source. Wipe every fitting dry, then raise the pressure from upstream to downstream, watching for the first point where a damp mark reappears.

Suggested order:wipe dry and record the starting state, then run at the working flow rate for 5–10 min and re-check each point. If point 5 is damp while point 4 is dry, the problem is at the analytical column inlet; if both 4 and 5 are damp, suspect a leak upstream running down the tubing.

← Swipe to see the full diagram →

Fig. 17 · Three forms of leak trace and what they mean The form of the trace indicates how long the leak has run and what the mobile phase contains Damp mark (droplet) Leaking now, with an aqueous or slowly evaporating mobile phase. Reappears soon after wiping. White crystals A slow leak of a buffered mobile phase over time, leaving salt after the solvent evaporates. The leak has run for some while. Dried mark It has leaked at some point but may not be leaking now. Wipe clean and observe again to confirm. Additional action where crystals are found Where salt crystals are seen, besides dealing with the leak point, check for salt accumulation there and downstream and assess the pitting risk to stainless steel parts. Clean by wetting with water and wiping, rather than scraping dry and scratching the sealing face.

Key point:A leak of a purely organic mobile phase evaporates quickly and may leave no obvious trace, showing only as low pressure or an unstable baseline. Such cases need a pressure test and a flow rate check to identify.

SymptomPossible causeOrder of action
Leaks when finger-tight, stops when tightened furtherUnder-tightenedTighten by the specified number of turns; if it takes clearly more than usual force to stop, check the end
Weeps persistently even when fully tightenedWrong standard / wrong insertion depth / scratched sealing faceRemove and check the thread and insertion depth; inspect the fitting cone
No leak at low flow, leaks as pressure risesAngled end / under-tightened ferruleRecut and deburr the tubing, fit a new ferrule and reassemble
intermittent weepingUneven end / material deformation from temperature variationCheck the end; watch for thermal deformation of PEEK parts inside the column oven
Thread strips after repeated tighteningDamage from over-tighteningReplace the fitting and ferrule; do not reuse them
Fittings dry but pressure lowLeak upstream / pump seal failureDisconnect the column and check the flow rate, locating it section by section

13 · Pressure and peak shape troubleshooting

← Swipe to see the full diagram →

Fig. 18 · Column pressure decision tree First distinguish rising, falling and fluctuating, then follow the branch Column pressure departs from the baseline Rising Falling Fluctuating Replace the guard cartridge first → does the pressure recover? Yes: sample preparation is inadequate No: continue Remove the column and run the system alone → is the pressure normal? Yes: blockage at the column head No: blockage in the system Action: forward flush with a strong solvent → if still high, backflush → if still high, replace the inlet frit or retire the column Check every fitting and the pump head for leaks Yes: locate it using Fig. 16 No: continue Are there bubbles in the pump, and are the check valves working Has the bed collapsed (accompanied by peak splitting) Action: purge, service the check valves; a collapsed bed is usually irreversible, so arrange replacement Does the fluctuation match the pump stroke frequency Yes: the pump or the damper No: continue Is the mobile phase degassed and the mixing stable Is the column temperature stable, and is there intermittent weeping Action: in-line degassing, check the check valves and the column oven; re-check the fitting seals

Establishing the baseline:after the first equilibration of a new column, record the column pressure under standard conditions (fixed flow rate, temperature and mobile phase composition) as that column's baseline. Judgements thereafter are based on the change relative to the baseline rather than on the pressure reading itself — tubing configuration differs between systems by several MPa.

← Swipe to see the full diagram →

Fig. 19 · Peak shape atlas and where to look first Six typical peak shapes; the dashed line is the symmetrical reference Symmetrical peak (reference) As about 0.9–1.2 Tailing Dead volume before the column / silanol interaction / contamination at the head Fronting Column head overload / injection solvent too strong Splitting / doublet A void in the bed / partial frit blockage (when every peak splits) General broadening Excessive extra-column volume / loss of efficiency Ghost peak Ghost / carryover peak Carryover from the previous injection / mobile phase impurity / contaminated needle A general way to separate column from system Every peak (including unretained components) affected to the same degree → points to an extra-column problem or to the bed as a whole. Only some peaks affected, in a way related to compound properties (basic compounds tailing, for example) → points to chemical interaction inside the column.

How to use this:Look at the peak shape of the unretained component (the dead time marker) first. That peak undergoes no retention, so its broadening and tailing come mainly from extra-column volume and flow path geometry, which makes it useful for separating column from system.

14 · One-page reference

14.1 Fittings and tubing

ItemPoint
What 1/16″ meansTubing outside diameter 1.59 mm, unrelated to the bore; the bore must be given as well when ordering
What 1/8″ meansTubing outside diameter 3.18 mm, mostly for preparative work, pump inlets and waste
Choosing the pre-column bore0.17 mm for a 4.6 mm column; 0.10 mm or finer for a 2.1 mm column
Common threads10-32 UNF (the analytical mainstream), 1/4-28 UNF (low pressure / preparative), 6-40 (miniature), M6×1 (some European systems)
Easily confused standards10-32 UNF with M5×0.8; 1/4-28 UNF with M6×1
Reusing a ferruleOnce tightened, a stainless steel ferrule's position is fixed; changing to a fitting of a different standard means a new ferrule and a fresh cut
Assembly orderSeat the tubing → hold axial pressure → finger-tight → tighten by the specified number of turns
Solvents to avoid with PEEKTetrahydrofuran, dichloromethane, dimethyl sulfoxide, concentrated nitric acid, concentrated sulfuric acid
Points to watch with stainless steelHalide ions and strong acids readily cause pitting; chelating analytes may interact with the metal surface

14.2 Symptoms and what to check first

SymptomCheck firstThenFinally
Column pressure risingReplace the guard cartridgeRemove the column and measure system pressureStrong-solvent flush / backflush
Column pressure fallingLeaks at the fittingsBubbles in the pump and the check valvesBed collapse (with peak splitting)
Column pressure fluctuatingWhether it matches the pump stroke frequencyMobile phase degassingColumn temperature stability
Peak tailingInsertion depth at the pre-column fittingContamination at the column headMobile phase pH and silanol interaction
Peak frontingInjection solvent strengthInjection volumeColumn head overload
Peak splittingWhether every peak splitsPartial blockage of the inlet fritA void in the bed
General broadeningPeak shape of the dead time markerConnecting tubing bore and lengthLoss of column efficiency
Retention driftingColumn temperature stabilityConsistency of mobile phase preparationWhether equilibration is adequate
Weeping at a fittingThread standard and insertion depthQuality of the tubing endDamage to the sealing face
Rising baseline noiseBubbles in the flow cellMobile phase impuritiesA damaged outlet frit letting packing escape

14.3 Storage solvents

Phase typeLong-term storage solventAvoid
C18 / C8 / phenylNeat acetonitrile or methanolProlonged storage in pure water
HILIC90% acetonitrile / waterHigh aqueous content
Normal-phase silicaHexaneSolvents containing water
Ion exchangeDilute salt solution with a preservativeAqueous solutions without preservative
SEC / GPCThe storage solution specified for the mediumChanging solvent systems without a transition

14.4 Problem attribution matrix

← Swipe to see the full diagram →

Fig. 20 · Problem attribution matrix: part × stage Stage of use runs along the horizontal axis and hardware part down the vertical; each cell gives the main problem for that combination Goods-in inspection Installation Routine operation Shutdown storage Deviation Handling Tubing and ferrules Fittings and threads Column head and frits Packed bed Guard Column Check both bore and outside diameterConfirm material compatibility Angled cut / burrsReusing a ferrule across standards PEEK swellingStainless steel pitting Usually no action needed Locate the leak (Fig. 16) Check the thread standardCheck the insertion depth Mixed standardsThread damaged by over-tightening Sealing gradually lost, causing weeping Store properly after removal Replace the fitting and ferrule End plugs in place Tightening before seating createsdead volume before the column Particulates accumulatingBuffer salt crystallizing Sealed away without flushing out salt Strong-solvent flush / backflushReplace the inlet frit Re-measure against the QC report Reverse installationGoing straight to a high flow rate Exceeding the pH or temperature limitChain collapse at high aqueous content Prolonged storage in pure water Bed collapse is usually irreversibleFollow the retirement process Whether the phase chemistry matches Two fittings introduceadditional extra-column volume Overdue for replacement Store as for the analytical column Replace the cartridge when the pressurehas risen about 20% Red cells are where problems concentrate; amber cells are those to keep an eye on.

How to read it:The installation column has the most red cells, consistent with Fig. 1 — connections are where general problems concentrate. The goods-in column is mostly confirmation, which is cheap and pays well.

15 · Frequently asked questions

The questions below are drawn from common technical support enquiries, grouped into selection, use, lifetime and mobile phase. A shorter version is atFAQ

Selection

C18A is the general-purpose grade, with higher retention and loading, suited to most reversed-phase applications; C18C has a wider usable pH range (about 1.5–12) and suits demanding mobile phases or small peptide separations; C18D is intended for hydrophilic compounds, retains stably under high aqueous conditions, and can be used directly with LC-MS without an ion-pairing reagent. Establish the sample polarity and mobile phase pH first, then choose the grade.
A guard column is advisable where: (1) the sample matrix is complex (serum, urine, food extracts); (2) more than 20 injections are run per day; (3) the sample may contain particulates. A guard column slows degradation of the analytical column's inlet frit and the front of the bed, and a cartridge costs far less than a new analytical column. How much life is gained varies with the matrix and the standard of sample preparation.
The base medium is the same; the particle size differs. Analytical columns commonly use 3 µm or 5 µm and preparative columns 10 µm or 20 µm. Larger particles give lower backpressure and suit high-flow preparative work; smaller particles give higher efficiency and suit fine analysis. Choose according to the objective — resolution first, or throughput first.
It can be fitted where the pressure allows, but the measured efficiency is usually below what the column can achieve. A UPLC column's volume is of the order of a tenth that of a conventional HPLC column, and the connecting tubing bore, flow cell volume and data acquisition rate of an HPLC system introduce relatively large extra-column broadening. To realize its efficiency, change to narrow-bore connecting tubing and raise the acquisition rate at the same time. SeeSection 2

Use & Operation

Equilibrate with 10–20 column volumes of mobile phase before injecting. If the column is stored in methanol, transition to aqueous conditions via 50% methanol/water before pure water, rather than flushing directly at high aqueous content and risking C18 chain collapse (hydrophilic C18 and polar-embedded phases can go straight to pure water).
Yes. The arrow on the column marks the direction in which the bed was consolidated, and flow should follow it. Reverse flow applies force in the direction that loosens the bed, which can create voids and lower efficiency and broaden peaks, usually irreversibly. Confirm the arrow points towards the detector before installing.
Do not respond by applying more force. Check in this order: (1) does the thread standard match (stop if the first two threads feel tight on entry); (2) is the tubing seated against the bottom of the fitting; (3) has the ferrule been swaged onto a fitting of another standard; (4) is the tubing end square and free of burrs; (5) is the fitting's sealing cone scratched. Most persistent weeping comes from (2) and (3). SeeSection 6
Yes. Phosphate, ammonium acetate and similar salts left in the column and tubing can crystallize and block the frit, or corrode stainless steel parts. The sequence is: (1) flush for about 10 min with 5%–10% organic in water; (2) raise the organic proportion to 50%; (3) change to neat methanol or acetonitrile for storage.

Performance & Lifetime

Under normal conditions an analytical column will take roughly 2,000–5,000 injections (inferred on this page; it varies considerably with matrix and conditions). The factors that matter most, in order, are: whether sample preparation is thorough, whether mobile phase pH and column temperature stay within range, whether a guard column is fitted, and whether the column is flushed correctly after each run. Tracking the rise in column pressure relative to baseline and the change in plate number is more reliable than counting injections.
Common causes: (1) particulates accumulating at the head; (2) buffer salt crystallizing; (3) protein or lipid deposited at the head; (4) bubbles in the mobile phase. In order: replace the guard cartridge; remove the column and confirm the system pressure is normal; flush forward with a strong solvent; if it still has not recovered, consider backflushing or replacing the inlet frit. The decision path is inSection 13
Check in this order: (1) is the column temperature stable (retention changes by roughly 1%–2% per 1 °C); (2) is the mobile phase prepared consistently (organic proportion, pH, method of preparation); (3) is the column fully equilibrated (at least 10 column volumes); (4) is the flow rate accurate (pump calibration); (5) is the column pressure within its baseline range.
Look at the peak shape of the unretained component first. That peak undergoes no retention, so its broadening and tailing come mainly from extra-column volume and flow path geometry. If that peak is equally affected, the problem is usually outside the column; if only retained components are affected in a way related to compound properties, it is usually inside. You can also replace the column with a zero dead volume union and inject, to compare against the system alone.

Mobile phase and tubing

(1) Do not use non-volatile buffers such as phosphate or borate; (2) formic acid (0.1%), acetic acid (0.1%), ammonium formate and ammonium acetate are suitable; (3) acetonitrile (better ionization efficiency) or methanol is usual as the organic phase; (4) choose a column that needs no ion-pairing reagent (hydrophilic C18, HILIC); (5) a flow rate of 0.2–0.4 mL/min is usual, to reduce ion suppression.
Not recommended. Phosphate gives a low UV background but is incompatible with MS; ammonium acetate is MS-compatible but gives a slightly higher UV background. Where one method must serve both UV and MS, use ammonium acetate or ammonium formate with a column that needs no ion-pairing reagent. Do not mix batches of mobile phase, as differences in composition cause retention drift.
Select by flow path section and sample type. Use stainless steel or reinforced PEEK for the high-pressure section before the column; PEEK from the column outlet to the detector; for biological samples and phosphorylated compounds, PEEK throughout or tubing with an inert surface treatment. PEEK swells in tetrahydrofuran, dichloromethane and dimethyl sulfoxide and degrades in concentrated nitric or sulfuric acid; stainless steel is at risk of pitting with halides or strong acids. The comparison is inSection 7
Both refer to the outside diameter, 1.59 mm and 3.18 mm respectively, which determines the fitting and ferrule to use. The bore is stated separately and determines the internal volume and extra-column broadening. Pre- and post-column connections in analytical LC mostly use 1/16″ outside diameter tubing, with the bore chosen to suit the column bore; 1/8″ outside diameter tubing is mostly for preparative LC, pump inlets and waste lines. Both must be given when ordering. SeeSection 4
For anything not covered above, send the column grade, lot number, system configuration, mobile phase composition and the symptoms to our technical team, or submit them through theonline enquiry page.