Proper sample prep ensures HPLC success: high-quality filtration in sample and mobile phase separation is crucial to guarantee reproducible HPLC and UHPLC data.
Effective sample preparation not only removes particles, it also helps reduce the sample complexity, thereby making chromatographic analysis easy. Particles held up on a column can also leach contaminants into the sample, thereby increasing background noise.
Filtration is one of the most commonly used sample preparation steps prior to chromatography. Even though filtration is very common and may appear to be a simple technique, selection of the correct filtration device is important.
Three common types of filtration techniques used prior to chromatography are ultrafiltration, microfiltration and depth filtration.
In ultrafiltration, separation is based on molecular size. This approach is typically used to separate/fractionate dissolved proteins and nucleic acids based on their molecular weight.
Microfiltration, based on membrane pore size, is commonly used for removal of particles and cells or sterilization of solution prior to analysis. Particles larger than the pore size of the membrane are retained.
Depth filtration is used for removal of larger or aggregated particles. Membranes used for depth filtration include glass fiber or polypropylene wool, which have large pore sizes; the tortuosity of the filtration path enables large particles to be retained.
Several key membrane characteristics must be considered when evaluating filters for sample preparation:
* Pore size: A rating based on the bubble point of the membrane, which indicates particle retention of the membrane.
* Porosity: The volume of open space in the structure of the membrane, which impacts dirt-holding capacity and flow rate.
* Chemical compatibility: The compatibility of the membrane material with various solvents and modifiers used for HPLC.
* Analyte binding: The binding characteristics of the membrane.
* Extractables: The impurities introduced by the sample coming in contact with the membrane.
Pore size and porosity
The pore size of a microporous membrane is typically measured using a technique called bubble point. Bubble point is the minimum pressure at which a wetted membrane allows a constant stream of air bubbles through it. Since bubble point is inversely proportional to the pore size, a higher bubble point indicates a smaller pore size and vice versa.
For ultrafiltration membranes, pore size is typically represented by retention characteristics of the membrane. This is expressed by the nominal molecular weight cutoff and is normally determined using dextran solutions of precise molecular weight.
In reality, the pores in a membrane are never round or cylindrical, but irregular. Pore size only gives a general idea regarding the particle retention behavior of a particular membrane. Scanning electron micrographs of various 0.45-micron membranes show that pore morphology is completely different across vendors and results from differences in manufacturing processes.
Percent porosity among membranes also differs. As noted above, percent porosity is how much open space exists within a membrane; this in turn determines the flow characteristics of a membrane. For membranes like PVDF, PES or nylon, porosity is in the range of 70 to 80 percent, which means 70 to 80 percent of these membranes are filled with air. Tracketched polycarbonate membranes have very low porosity ranging from about 10 to 20 percent. This means that polycarbonate membranes do not allow for higher amounts of filtration through them.
The pore structure across the thickness of a membrane can impact filtration characteristics. Figure 1A compares two 0.22-micron membranes--a symmetric (PVDF) membrane and an asymmetric PES membrane with funnel-like morphology. The particle retention characteristics will be determined by the lower surface, which has a smaller pore size, whereas the upper surface having a larger pore size will allow for fast filtration. Figure 1B shows the impact on throughput of tissue culture medium with 10 percent fetal bovine serum. The PVDF membrane clogs after filtration of only about 1 L of sample, while the asymmetric PES membrane allows for filtration of nearly 3 L.
It should be noted that membranes from different vendors might have different pore structures that will result in different particle retention.
Figure 2 demonstrates the effect of reduced particle retention on a UHPLC back pressure. In this study, a mobile phase of 50 percent acetonitrile and water was filtered through various disk membrane filters. No sample injections were involved. UHPLC was run at 0.25 mL/min overnight and the back pressure was monitored throughout the run.
The mobile phase that was filtered through the polypropylene disk filter from vendor B allowed for particles to go through the filter, which could result in a clogged system. This back pressure increase could lead to a shutdown in the system in an overnight operation. In light of these variations, it is critical to distinguish between membranes from various vendors to make sure they meet the requirements of the application being performed.
In an extreme case, a membrane can completely dissolve in a solvent that is incompatible with the sample being filtered. More often, a membrane that is incompatible with a solvent leads to extractables entering the sample and impacting resolution and sensitivity. It is therefore important to know the compatibility of sample or solvent being filtered through the membrane to ensure the cleanest sample for chromatography.
Table 1 summarizes the chemical compatibility of membranes with various mobile phase components. Polypropylene and hydrophilic PTFE membranes show the most chemical compatibility of all the commonly available membranes.
Compatibility of the housing materials used in the syringe filter should also be considered. Housing materials are made of polypropylene or HDPE, high-density polyethylene.
Both of these have high chemical compatibility and therefore offer very clean samples for downstream analysis.
Any surface that comes in contact with the sample has the potential to bind analyte. Typically, analyte binding occurs because of weak nonpolar interactions between the analyte and the surface. It is well known that the glass vials used in HPLC analysis can bind polar analytes, such as amino acids or phenols, to different degrees. Different membranes show different binding behavior depending on whether the analyte is a protein, peptide or a small molecular weight analyte.
In the case of proteins, binding characteristics can vary widely from a low 10 [micro]g/[cm.sup.2] binding for a PVDF membrane to the other extreme of close to 250 to 300 [micro]g/[cm.sup.2] shown by a nylon membrane. Other membranes fall in between these two extremes. Similar membranes from different vendors can have very different protein binding characteristics (data not shown).
With small molecular analytes, nylon membrane-based syringe filters tend to show very strong binding characteristics. Figure 3 shows the binding of small molecule analytes to membranes. In this study, four different drug tablets were dissolved and filtered using either a nylon or a hydrophilic PTFE syringe filter. Filtrate volumes were collected and the concentration of drug in the downstream filtrate was determined. With nylon syringe filters, between 10 to 40 percent of the analyte was lost due to drug binding. Hydrophilic PTFE membrane-containing syringe filters provide quantitative drug recovery even for the first mL of filtrate.
Extractables are impurities released into the sample as the sample passes through a membrane filter. Extractables may appear as a distinct peak or may coelute with the analyte of interest, thereby complicating quantiation.
The level of extractables can be reduced by pre-rinsing the filter with the sample or solvent used for filtration. When pre-rinsing is not possible because of limited sample volume, it is recommended to use a membrane or syringe filter that inherently shows low levels of extractables as well as low levels of binding, such as hydrophilic PTFE-based filters.
Syringe filters are not the only source of extractables in the solvent or the sample. Any surface that comes in contact with a sample can introduce extractables, including the black piston seal in the syringe. A glass syringe with a Teflon luer may result in extractables when the stainless steel comes into contact with an acidic sample. Glass vials can also introduce various metal ions into samples, thereby contaminating the sample.
In summary, there are many filtration materials and device configurations available to meet specific application requirements. It is important to consider variables such as chemical compatibility and sample recovery as well as sample volumes and throughput requirements when choosing the ideal filtration product. In the end, high-quality filtration products for sample and mobile phase preparation are critical to enable reproducible HPLC and UHPLC data and optimal instrument performance.
by Vivek Joshi, Principal Research Scientist, EMD Millipore, Danvers, Mass.
Table 1: Chemical compatibility of various membranes. Hydrophilic PTFE and PP are the most chemically compatible membranes. Sample Mobile Phase Diluents Solvents MeOH ACN Water IPA DMSO DMF and Buffers PTFE E E E E E E PVDF E P E E P P PES G E P G P N Nylon N E E N E E MCE P P E G P P PP E E E E E E Modifiers 0.2% 10 mM Formic 0.1% Phosphate Solvents acid TEA Buffer and Buffers PTFE E E E PVDF E E E PES N E E Nylon P E E MCE E G E PP E E E Modifiers 50 mM 50 mM 0.1% Ammonium Solvents Ammonia TFA Acetate and Buffers PTFE E E E PVDF E E E PES P E N Nylon N P E MCE E G E PP E E E E = Excellent, G = Good, P = Poor, N = No Information
|Printer friendly Cite/link Email Feedback|
|Date:||Sep 1, 2014|
|Previous Article:||OEM components for biochromatography apps.|
|Next Article:||GC column set runs seven EPA methods.|