When LC Methods Failed to Reproduce: What Could Be the Cause?

In HPLC, reproducibility is an inevitable and vital topic for both users and manufacturers. Fundamentally, perfect reproducibility can only be achieved when chromatographic conditions remain entirely unchanged—meaning not only the parameters recorded in our experimental logs, but every micro-variable within the entire system.

In this article, we divide the chromatographic system roughly into four primary components for simplicity: reagents, samples, instruments, and columns. To make this discussion more practical, we will exclude the unavoidable factor of random error and operate under the assumption that the analytical method itself is sound and possesses a reasonable degree of robustness.

Please note, that this article is based on the long-term experience accumulated by Welch Materials, and focuses primarily on application-level factors that are easily overlooked in daily practice, rather than restricting the scope solely to the column itself.

Needless to say, the quality of reagents is a fundamental prerequisite. Whenever circumstances permit, always opt for the highest quality reagents available. The negative impacts of subpar reagent quality often do not manifest immediately, but they significantly compromise system lifespan and method reproducibility over time.

Besides the quality of reagents, several other aspects also require attention:

1. Changes in Reagent Properties

Reagent properties do not remain static from the moment it is sealed in a bottle until it is completely used. Shifts occur time to time, especially after they are exposed to the environment:

• Volatilization: Highly volatile reagents will escape, changing the concentration of the solution.
• Carbon Dioxide Absorption: Alkaline reagents readily absorb CO₂ from the air, altering their pH and ionic strength.
• Moisture Absorption: High-purity solvents absorb ambient moisture. This is particularly detrimental to normal-phase chromatographic systems, where trace water drastically shifts retention times.

2. Inherent Instabilities of Specific Reagents

• Ammonium Acetate: This reagent is notoriously unstable, particularly regarding the pH of its prepared solutions. Even with freshly opened, high-purity bottles, bottle-to-bottle variation can be large enough to alter chromatographic outcomes.
• Purified Water: Water is a highly unique reagent in the laboratory. While other reagents are purchased ready-made at specific purity levels, water is typically generated on-site via purification systems or sourced from commercial purified drinking water. As we know, the physical and chemical properties of purified water can fluctuate drastically across many parameters, and even a water purification system displaying a resistivity of 18.2 MΩ·cm does not guarantee problem-free water; conductivity is merely one of many quality metrics.
  Welch Tech Tip: Water quality varies significantly across different global regions and even within large nations. Extra caution is required when attempting to replicate experiments across geographically distant laboratories.

A common misconception among users is that if the primary analyte of interest remains the same, the method should be universally applicable. However, the fundamental purpose of chromatography is separation: even if the target compound is identical, any variation in the other components—including impurities, excipients, matrices, and diluents—will alter the requirements of the chromatographic separation.

This explains why active pharmaceutical ingredients (APIs) and their finished formulations often require vastly different sample pretreatments and detection methods. When the sample matrix changes, the original chromatographic conditions frequently lose their suitability due to the following factors:

• Solvent Effects: Shifts in the sample solvent (diluent) or injection volume introduce varying degrees of solvent effects, which can distort peak shapes.
• Matrix and Excipient Variations: Alterations in the type or concentration of excipients and matrices not only impair detection but can also cause direct, irreversible damage to the column stationary phase. This explains why columns used in formulation and food analysis typically exhibit shorter lifespans despite rigorous sample cleanup.
• R&D Reaction Mixtures: In the research and development phase, the composition and concentration of reaction solutions fluctuate dramatically, making variations in chromatographic results completely expected.

As the most expensive component of a chromatographic system, instruments used by most laboratories are from major global brands. They are often highly trusted, and as a consequence, we often overlook instrument-induced variability.

Nevertheless, almost every chromatographer has encountered a situation where shifting a method to a different instrument, or even using the same instrument after routine maintenance, yields altered results.

1. Extra-Column Volume (Dead Volume)

The detrimental effects of extra-column volume on peak broadening are well known and needless to say.

We have also observed a unique phenomenon in practical applications: when the extra-column volume between the injector and the column inlet is relatively large, it allows the sample and the mobile phase to mix more thoroughly prior to reaching the column head. This can, to some extent, alleviate severe solvent effects.

However, modern high-performance instruments are often engineered with minimizeddead volumes, and as a result are occasionally more susceptible to solvent effects because of this reduced pre-column mixing.

2. Actual Mobile Phase Composition (under Isocratic Conditions)

Under isocratic conditions, the environment inside the column remains constant. As long as the instrument delivers an accurate solvent ratio, few issues arise. The two primary methods for mobile phase delivery behave differently:

• Manual Preparation with Single-Chamber Delivery: Provided that standard operating procedures are strictly followed, this method offers excellent stability.
• In-Instrument Online Mixing: Different instrument designs introduce variations in actual delivered ratios. Low-pressure mixing systems exhibit significantly higher errors when blending very small percentages of a solvent. Furthermore, variations in pump module designs and mixing logics (such as the utilization of correction factors versus pre-compression technologies) generate discrepancies in the actual mobile phase output between different instruments.

3. The Flow Rate in Gradient Conditions

Under gradient profiles, reproducibility issues are significantly amplified. Because the mobile phase environment within the column changes continuously—and is non-uniform across different sections of the column at any single point in time—achieving ideal reproducibility requires strict control over the internal column environment at every sequential moment. This demands that the solvent ratio and flow rate delivered to the column inlet correspond precisely with time.

This makes flow rate critical in gradient elution, because it directly dictates the rate at which the mobile phase environment shifts inside the column. In contrast, minor flow rate variations under isocratic conditions rarely alter result, as the selectivity remains constant.

4. Gradient Profiles: "Not What You See"

When evaluating gradients, keep these critical technical factors in mind:

• Gradient Delay (Dwell Volume): The parameters configured in the gradient table dictate when the instrument modules execute an action, not when that specific mobile phase composition actually reaches the column inlet.
• Step-Wise Gradient Delivery: Although method software displays a smooth, continuous gradient curve (whether linear or curved), reciprocating piston pumps deliver fixed solvent proportions per stroke. Consequently, the actual gradient progression is a series of microscopic "steps" rather than a truly continuous curve.
• Instrument-Specific Gradient Curves: Due to differences in pump hardware design and electronic control logic, the actual delivered gradient curve inevitably varies from one instrument model to another. These variations are prominently magnified in shallow (gentle) gradient profiles.

Chromatographic behavior operates at a highly microscopic level, and even modern theoretical frameworks cannot fully explain every interaction. Although Welch Materials has been an established column manufacturer, we, in this section, merely analyze these factors based on macroscopic parameters and theoretical models to provide a systematic troubleshooting framework.

1. Column Hardware

While column hardware design strongly influences peak symmetry and flow distribution, the machining and manufacturing technologies of these components are highly mature. Therefore, quality fluctuations in column hardware are generally minimal, and their impact on reproducibility is limited as long as the hardware model remains unchanged.

2. Stationary Phase (Packing Material)

The following parameter analyses assume a controlled-variable framework:

Welch Quality Assurance: The carbon loading configurations of Welch products—particularly our classic series—are defined through extensive experimental validation and decades of market feedback. Within reasonable manufacturing tolerances, our processes minimize the impact of bonding heterogeneity on the vast majority of target compounds.

3. Column Packing Technology

The quality of column bed consolidation has a universal impact on all analytes. This factor is tightly controlled at the factory level through stringent column efficiency and peak symmetry testing prior to release.