This solution can multiply lab productivity up to fivefold without sacrificing performance.

HPLC is a precise, sensitive, and quantitative separation technique. Its popularity rests on its ability to separate a variety of analytes, including organics, ions, polymers, and biomolecules. It is the premier analytical technique in the pharmaceutical industry and is used in all phases of drug discovery, development, and quality control. It is not uncommon for a pharmaceutical lab to have tens or even hundreds of HPLC instruments, although often samples still pile up.

Now, imagine you are a pharmaceutical lab manager facing more and more samples each month and having trouble meeting project deadlines. What would you do? Work longer hours? Buy more instruments and hire more chemists? There is an existing solution that will help immediately. This solution is fast LC, an HPLC variant that is easily implemented but often overlooked.

What Is Fast LC?
Fast LC performs rapid analysis using short (3–10 cm), 4.6-mm i.d. columns packed with small particles (3-µm), usually operating in the 1–3 mL/min range. In contrast to conventional 10–15 cm, fast LC can enhance lab productivity up to fivefold without sacrificing performance, reliability, and convenience.

The concept of performing fast chromatography is not new. In 1941, A.J.P. Martin and R.L.M. Synge postulated that “The smallest height equivalent of theoretical plate (HETP) should be obtained by using very small particles and a high pressure difference across the length of the column.” Their predictions were mostly realized when early meter-long pellicular 30-µm columns were superceded by shorter columns packed with 10-µm and then 5-µm microparticulates. By the mid-1980s, short columns packed with 3-µm particles were commercially available; however, they weren’t commonly used until recently. The concepts, performance, requirements, and applications of fast LC were widely reported in the literature (1, 2).

The reasons for using 3-µm particles in fast LC are evident from examining Van Deemter’s plots (HETP vs flow rate) for 3-, 5-, and 10-µm columns (Figure 1). The smaller 3-µm particles yield lower HETP or higher efficiency per unit length. The optimum flow rate (the flow rate that yields maximum efficiency or lowest HETP) is higher for smaller particles. The optimum flow rate is 1.8 mL/min for the 3-µm column versus 0.7 mL/min for the 10-µm column. The small-particle columns have less efficiency loss at high flow rates because they are less resistant to mass transfer (i.e., flatter Van Deemter’s curve).

Benefits of Fast LC
Switching to fast LC results in major benefits: faster analysis, faster method development, reduced solvent consumption, and increased mass sensitivity.

Faster analysis. Typical isocratic assays can be performed in 1–3 min and gradient runs in 3–10 min. The analysis, for example, of an analgesic tablet extract took 1.5 min when using a popular 3 x 3 C18 column (3-cm long, 4.6-mm i.d. packed with 3-µm particles). This column with ~4000 theoretical plates is most suited for the analysis of bulk or formulated pharmaceuticals. What if higher resolutions for more complex samples are required? Fast LC offers a way to increase resolution without switching to a longer column by increasing analyte retention. Retention is often measured in k´s or capacity factors. Typically, the mobile phase is adjusted so all analytes are eluted in the k´ range of 1 to 10, so peaks are not too highly retained and have excessive analysis time. Figure 2 shows how fast LC can be used in either the high-speed or high-resolution mode. In the high-speed mode at 50% acetonitrile, the four antimicrobial parabens are baseline resolved in 20 s. By reducing the organic modifier to 20% acetonitrile, the last component elutes at 10 min or a k´ of 134; however, the resolution between propyl and butyl parabens is increased from 3.4 to 10.8. Fast LC is unique in its ability to effect separations at very high k´s while maintaining reasonable analysis time and detectability.

In HPLC gradient analyses in which solvent strength is increased with time, resolution and analysis time are controlled by several factors, including column length and efficiency, gradient time, and flow rate. One of the most challenging gradient applications is peptide mapping, a powerful method for the structural determination and confirmation of proteins. In peptide mapping, a sample protein is selectively cleaved by an enzyme into peptide fragments, which are then separated by HPLC and subsequently collected for sequencing. A typical high-resolution peptide map takes 40–120 min. Figure 3 shows two tryptic maps of lysozyme. The top chromatogram is from a conventional 15-cm column in 45 min, and the bottom chromatogram is from an 8-cm fast LC column in 3.5 min (3). This example illustrates how optimizing the column and the gradient-controlling factors can drastically reduce analysis time without sacrificing resolution.

Rapid method development and validation. Because resolution in LC is controlled primarily by the mobile phase, HPLC method development is mostly a process of finding the right mobile phase and adjusting its solvent strength. Paradoxically, with all the HPLC hardware sophistication and computerization, most method development is still performed manually via trial and error. Figure 4 illustrates the method development steps used to resolve the active ingredient (labeled *) in a pharmaceutical formulation using sequential isocratic steps. The big advantages for fast LC here are quick feedback and rapid column equilibration. The analyst can change the mobile-phase conditions, wait 2 min for the column to equilibrate, re-inject the sample, and watch the resulting chromatogram for instant feedback. The total method development time shown in Figure 4 is only 15 min. In method validation, analytical performance data such as precision, accuracy, robustness, linearity, and sensitivity are gathered, and hundreds of assays are required. Here, fast LC methods can be validated in days rather than weeks.

Reduced solvent consumption and enhanced mass sensitivity. The shorter analysis times in fast LC translate to a 50–80% reduction in solvent use, which lowers the cost of purchase and waste disposal. In addition, the smaller column volume (e.g., ~0.3 mL for 3 x 3 columns) means less dilution of the analyte peaks, resulting in 3–5 times higher mass sensitivity (taller peak heights) using an absorbance detector.

Dispelling Common Myths
If fast LC is so great, why isn’t everyone using it? There are several commonly raised objections to fast LC, some true, some false:

• Higher column backpressure: True. Backpressure is inversely proportional to the square of particle diameter. However, the shorter column format allows most analyses to operate in the “normal” range of 1000–3000 psi.

• Lower column efficiency: Untrue. The 3-µm particle column has 60% higher efficiency per unit length than the comparable 5-µm counterpart. Because column length plays a vital role, a 3 x 3 column with 4000 plates is less efficient than a 15-cm, 5-µm column with ~10,000 plates. Therefore, a longer 3-µm column (10-cm with ~12,000 plates) is recommended for samples requiring higher resolution such as peptide mapping, metabolic profiles, or stability-indicating assays of formulations in which all expected impurities and degradants are baseline-resolved.

• Batch-to-batch reproducibility problems: Untrue. This is a universal problem for all analytical columns and the number one headache for HPLC analysts. Significant improvements in reproducibility have been made in recent years with the advent of high-purity silica with lower trace metals and improved bonding chemistry.

• Short column lifetime: Partially true. Some early fast LC columns with distributor inlet plate design (which accentuate injector pressure pulses) did have shortened lifetimes. Better column hardware and packing techniques, however, have significantly increased column longevity to acceptable levels (>2000 injections). Nevertheless, the densely packed 3-µm particle columns are inherently more susceptible to clogging problems from sample- or mobile-phase–borne particles or contaminants. In this regard, the judicious use of a guard column (placed before the column) and a scavenger column (placed before the injector) is highly recommended for prolonging column lifetime (4).

Understandably, most pharmaceutical analysts operating under a cGMP (current Good Manufacturing Practice) environment cannot and would not change the designated column, because it might require a method revalidation. Therefore, fast LC technology must be adopted in early drug development before regulatory submissions. Other reasons cited for staying with the conventional 5-µm column are concerns about method portability and the comfort of built-in safety margins to resolve unexpected process impurities.

There are strong indications that fast LC is becoming more popular. A 1997 survey showed that 17% of analysts are already using fast LC, and 59% of current column users are interested in future purchases (5). This interest stems from the shift to accelerated drug discovery with emphasis on high-throughput screening of combinatorial libraries and pharmacokinetic analysis by LC/MS or LC/MS/MS. The high equipment cost of LC/MS provides a powerful incentive to increase sample throughput by fast LC.

System Requirements
Fast LC columns produce sharp and fast eluting peaks. For instance, a 3 x 3 column can generate peak widths of less than 30 µL or 0.5-s peaks that are adversely affected by extra-column broadening. To reduce these effects, the instrumental bandwidth or dispersion of the HPLC system must be minimized by using a small detector flowcell, shorter 0.007 in.-i.d. connection tubing, and a low-dispersion injector (e.g., Rheodyne 8125) (1, 2). Most existing systems can be modified successfully by replacing the standard detector flowcell with a micro flowcell (< 3 µL) and bypassing the heat exchanger preceding the detector flowcell. Note that smaller flowcells often have shorter path lengths, leading to some compromises in sensitivity. For fast gradient analysis, the delay volume inherent in most multisolvent LC pumps (~2 mL for low-pressure mixing pumps) can be a limiting factor. Secondary considerations are the injection turnaround time of the autosampler (<1 min), its sample tray capacity for overnight runs (>200 vials), the detector response time (<0.5 s), and the sampling or spectral scan rate of the data system (ideally >10 points/s for data system or >5 scans/s for diode array or mass spectrometer).

Trends in Fast LC
HPLC columns are becoming more efficient and reliable with less peak tailing and improved batch-to-batch reproducibility. The same holds true for fast LC columns, which operate like other packed columns and benefit from the improvements from high-purity silica and novel bonding chemistries. Improvements in monolithic column technology by using in situ polymerization techniques might one day offer a viable alternative to packed columns. Actual column configuration will always be dictated by the instrumental bandwidth. Low dispersion systems will allow the use of shorter, narrower columns packed with even smaller particles. The use of 3-mm and even 2-mm, 3-µm columns will lead to higher solvent savings, sensitivity, and better LC/MS compatibility. The adoption of sub-3-µm particles will lead to even faster separations.

Fast LC is obviously desirable as it provides a means of addressing the ever-increasing sample load with minimal investments. Today, fast LC is precise, economical, and practical. It is successfully implemented by taking deliberate steps such as replacing flowcells, installing precolumns, and paying more attention to operating procedures. Perhaps this is a small price to pay for a three- to fivefold increase in productivity.

---

References

1. Dong, M. W.; Gant, J. R. LC-GC 1984, 2, 294.
2. Gant, J. R.; Dong, M. W. Pharm. Technol. 1987, 11 (10), 44.
3. Dong, M. W. Tryptic Mapping by RPLC. In Advances in Chromatography; Brown, P. et al., Eds.; Marcel Dekker: New York, 1992; Vol. 32, pp 21–52.
4. Dong, M. W.; Gant, J. R.; Perrone, P. A. LC-GC 1985, 3, 786.
5. Majors, R. LC-GC 1997, 15 (11), 1008.

---

Michael Dong is senior principal scientist with Purdue Pharma. He has more than 10 years of research, applications, and marketing experience with a major HPLC manufacturer and has published more than 60 journal articles. He has a Ph.D. in analytical chemistry from the City University of New York. Comments and questions for the author may be addressed to the Editorial Office by e-mail at tcaw@acs.org, by fax at 202-776-8166 or by post at 1155 16th Street, NW; Washington, DC 20036.