HPLC SOLVENT COMPRESSIBILITY - REVISITED

 Twelve years ago I published a short article here (HPLC PUMP SOLVENT COMPRESSIBILITY VALUES) which described the importance of setting the correct solvent compressibility values in the HPLC pump's table. Developing HPLC methods which exhibit smooth, stable baselines, with little measurable signal artifacts (e.g. spikes, noise, oscillation) and minimal pressure fluctuations help insure reliable, repeatable methods. Taking steps to insure that the LC pump operates is setup properly for the method are part of following good chromatography fundamentals

Over the past month I consulted for three different clients who needed help in troubleshooting various "pump stability problems". In all three cases, each HPLC system showed extreme pump pressure cycling, cavitation, noise and instability over time. Pressure fluctuations of 10% (or in one case, 10-30% Ripple values) were observed in several different HPLC methods that were used. One of the very first areas to check for problems with pump pressure instability is mobile phase degassing.  

Proper operation of the HPLC pump requires that efficient degassing of all mobile phases is performed before the liquids enter the pump head. 

Failure to properly degas liquids often results in pump cavitation, check valve sticking and baseline instability. An Inline vacuum degasser or continuous Helium sparing should be used to degas all mobile phase solution for use in HPLC (not sonication or vacuum filtration which perform poorly to solve degassing issues). 

In one of the three cases, the HPLC degasser was found to be broken and long overdue for service. Cleaning and servicing the degasser cured the problem and the method that once showed pressure ripple of >10% now shows no baseline disturbances and very low ripple of ~0.1% at ~ 70 bars system pressure. 

Before I was called in to assist each client, the clients had replaced numerous parts, including: pump seals, check valves, mixers, solvent frits and still had the same baseline instability issues (no change). As recommended by me, two of the clients had their very old degassers cleaned and serviced (as they were long overdue for service), but still had some baseline and pump instability (servicing the degassers improved the baselines, but the pump was not running as it should). In both cases, the cause for the remaining pump instability was quickly identified by me on-site (many problems can be quickly diagnosed on-site).

• The client had incompatible solvent compressibility values stored as part of their HPLC methods. This resulted in huge baseline disturbances, spikes, cavitation and occasional loss of prime. 

One of the clients normally ran methods containing high percentages of ACN (with some water) for their sample methods, but a few months earlier had switched to running with gradients containing high percentages of methanol. The solvent compressibility values stored in their system were appropriate for WATER, but they never updated them when they used the same method file to run samples in mostly methanol solutions (which need different compressibility values). Though they all had been using HPLC for many years, they had not received basic HPLC instrument training to know how to adjust and optimize these and other important instrument settings for EACH method (they were overwriting each new method, a common new user mistake, when making changes). Once we changed the method's solvent compressibility value to a more compatible one (in their case, for methanol), the baseline smoothed out in just a few minutes and all of the pressure instability issues went away (*they had replaced several thousand dollars worth of perfectly functioning parts trying to solve this issue before I arrived). Professional training in how to use and operate any HPLC instrument should always include how to set and optimize the compressibility value(s). Make sure you know how to incorporate the correct value in each new method that you create. Always spend up-front time to optimize each method for the application before you use it to analyze real samples. The initial time spent getting everything to run smoothly and reliably will improve overall accuracy plus save money and time.

• Note: In a low-pressure HPLC single-head pumping system with multi-position solvent selector valve (e.g. Most ternary or quaternary systems) one value is allowed, but in a true, dual-head binary pumping system each of the two pump-heads may have a separate field to input the solvent compressibility values.

The importance of inputting the correct and applicable solvent compressibility value(s) into the pump's settings, for each solvent used is one of many steps in creating an optimized HPLC method. There are no universal values, but the instrument manufacturer will have included a generic value in the pump's compressibility settings field. Should you use this generic value?  What are the chances that a randomly selected value used as a 'place holder' in the software is the correct value for your method?  Just as with flow rate, solvent composition, run time, stroke volume, wavelength etc., entering (and saving) the correct solvent compressibility value into EACH method helps to optimize the pumping performance. You will want to select an appropriate value FOR EACH AND EVERY HPLC METHOD YOU CREATE and use (and be sure to save the method with a unique name). Start by loading your HPLC method into the system, then look at the solvent compressibility value(s) used. Are they correct? Change the value(s) shown to values that are appropriate for your method. It is OK to experiment and try different values (we encourage it!). Monitor the S/N levels of the baseline noise for comparison. The instrument manufacturer should provide a table of suggestion solvent compressibility values for use with their system [For HP/Agilent systems, you can see an example table at the link I provided in the first paragraph of this article or review the operator's manual for more information].

• A related article to this one that you may find useful is; "Diagnosing & Troubleshooting HPLC Pressure Fluctuation Problems (Unstable Baseline)";
  

A Case of Changing Solution pH. Formic Acid Stability in Solution (Methanol)

Real life examples help to better illustrate problems that I am called in to troubleshoot for clients. As a professional scientific consultant, many of my clients have spent months (sometimes years) trying to solve an analytical problem on their own before I am brought in to make the diagnosis and propose a solution. Many years of working in a wide range of scientific fields allows me to identify problems quickly and efficiently saving clients the most money and allowing them to resume work on their projects.

This was the case during a recent consult for a major cannabis testing laboratory. They were having a great deal of difficulty obtaining reproducible results for their analytical testing screens (14 compounds in their analysis with a need for repeatable and accurate results). Variations from 25% to 50% were observed run-to-run over the course of seven days. They assured me they were doing everything in the same way. To begin the troubleshooting process, we started by looking at the actual data gathered and the actual method(s) used to acquire the data. These were evaluated to see if they followed good practices and techniques, also to make sure they had SOP's in place which were clear. Good SOP's must include enough detail to allow anyone reviewing them to prepare samples, standards and/or solutions in the exact same way. Additionally, the HPLC instrumentation was checked and tested to verify it was performing as designed.

After reviewing their training and methodologies on-site, a number of areas of concern were quickly identified. One of the most likely reasons for the variation in values over time was found to be caused by a common mistake in the preparation of mobile phase solutions for the HPLC system. To save time, the client's scientists prepared all organic solvent solutions in advance (~ one month or more), then filtered and stored them at room temperature. For example, their solutions of 0.1% formic acid in HPLC grade Methanol were pre-mixed and stored in glass one liter bottles. These bottles were then put aside, for an average of one month before use. This finding proved key as someone with proper HPLC training would be aware of a well known problem when formic acid is left in pure organic solvent, especially methanol, over time (less so with ACN). Briefly, the formic acid content degrades quickly over time and is often found to be only half of what it was initially after just three or four days (If you have not done so already, this is a simple and useful experiment to run in your lab, monitoring the acid level by titration, not with a pH meter, over time at room temperature in methanol)! This degradation continues over time reducing the amount of acid in solution. If the acid is added to the solution to enhance ionization (i.e. LC-MS; LC-MS/MS) or provide acidification to maintain the sample in a fully ionized form, then as the level of acidification decreases, so does the solution's ability to maintain it. In other words, your HPLC method may change over time (resulting in an in-valid method).

•  I have always promoted the importance of making and using freshly prepared mobile phase solutions (daily), especially where any aqueous solutions are used (to prevent degradation of additives and/or bacterial or fungi growth). However, this precaution does not normally apply to many pure organic solvents, but there are a few very important exceptions to this, formic acid and methanol in this example. 

Changes were made to their SOP's to insure that future solutions of formic acid in methanol were not prepared in advance, but instead, fresh on the day needed only. This coupled with a few basic improvements to their column washing, equilibration and overall training resulted in %RSD of only 0.3% for future analysis runs.

 
As a side note, I have been asked why solutions of formic acid in methanol are sold commercially for HPLC use? I have no answer to this, but respectfully remind everyone that just because something is offered for sale, does not mean it should be purchased. Ask yourself if the item is appropriate for your application? It may not be suitable for your use or application. 

BTW: Please be sure to flush your HPLC system of all organic acids (e.g. acetic, formic) after use and do not leave them in the HPLC system overnight. Even 1% levels of organic acids may be corrosive to stainless steel. 

Backpressure Changes, Pressure Drop from HPLC Tubing Selection (0.007, 0.005, 0.010")

In previous articles we have discussed how the choice of column particle size directly changes the system backpressure. Smaller particles generate higher back-pressures. We have also discussed the importance of HPLC tubing selection to minimize delay volume and diffusion within the HPLC's laminar flow path. Let us now focus on how the tubing's internal diameter and length impacts the total HPLC back-pressure (or pressure drop) observed. 

Key Points:  

1. Try to optimize the plumbing of your HPLC system.  
2. HPLC Tubing lengths between connections (or HPLC modules) should always be as short as possible. 
3. Pressure drop is dependent on the tubing length and inner diameter. Doubling the inner diameter of the tubing will decrease the pressure by a factor of 16.

Once the HPLC tubing connection lengths have been minimized, the next critical dimension which affects band broadening, delay volume and peak-width is the internal diameter (ID) of the tubing. The tubing selected should be narrow enough to reduce the undesirable spread of the peak(s) inside the tubing, but not be so narrow or restricted to result in clogs or obstructions (which is why good chromatography guidelines should be followed insuring that each sample is fully dissolved and filtered before injection). Commonly used tubing ID’s for most analytical HPLC systems are: 0.010” (0.25 mm), 0.007” (0.17 mm) or 0.005” (0.12 mm). By far, 0.007” (0.17 mm) is the most commonly used size for modern analytical HPLC analysis as it offers a compromise between low delay-volume and modest back-pressure (with fewer clogs). However, in addition to the much lower internal volumes which accompany the narrower ID’s, the pressure drop measured across equivalent lengths of tubing may change dramatically and this should be noted during set-up, selection and operation. Take the time to learn what "normal" backpressures are under specified conditions.

 

Understanding how the HPLC system backpressure changes as the internal diameter of the tubing varies is extremely useful in troubleshooting a number of common HPLC problems.

Let us compare the pressure drops measured across three popular HPLC tubing ID’s of the same length (40 cm) using common HPLC mobile phase solvents. This table will help illustrate the observed backpressure changes that the tubing ID and liquid have on the pressure drop.

PRESSURE DROP (in bars):

SS Capillary Tubing, 40 cm length, flow rate 1.000 mL/min.

Mobile Phase / Tubing ID	Water	ACN	MeOH	MeOH/Water (1:1)	IPA
0.010” (0.25 mm)	0.7	0.2	0.4	1.2	1.5
0.007” (0.17 mm)	2.7	1.0	1.6	5.1	6.2
0.005” (0.12 mm)	10.4	4.0	6.3	19.1	24

Note: Pressure drop is also a function of tubing length so if we halve (1/2) the length of tubing used, we also will reduce the pressure drop by one-half. 

Note the four-fold change that narrowing the tubing ID has at each ID reduction. The change is more dramatic when viscous solutions are used (i.e. MeOH/Water or IPA). If you re-plumb any part of your HPLC system with new tubing, then awareness of this physical change will assist you in troubleshooting many types of HPLC problems (to know which types of pressure changes indicate a real problem and which types of pressure changes are normal). Changes to the overall length or ID may result in noticeable changes to the total system backpressure. As an experienced chromatographer knows, when HPLC solvents are mixed together (e.g. gradient analysis) the pressure does NOT always follow a linear progression. In some cases, a reaction occurs between the solutions resulting in an overall change to the final viscosity of the mixture which may not be expected or understood by novice chromatographers (e.g. mixtures of MeOH/Water and ACN/Water are very well know examples which show these properties). 

 

You can download a free, more detailed table of 'HPLC Tubing Backpressure Examples' in PDF Format at this link:

HPLC Solvents, Acetonitrile and Methanol, Key Differences and Properties

Widely used in RP HPLC method development, Acetonitrile (ACN) and Methanol (MeOH) are the two most common solvents you will use with water or aqueous buffers to develop methods. So, besides the fact that Acetonitrile is well know to have a higher elution strength / capacity than Methanol [*but NOT at high organic concentrations (e.g. 95% Methanol vs 95% ACN) where Methanol has a higher elution strength than Acetonitrile does], what other properties should chromatographer's be aware of? Let's discuss a few that all chromatographers should know.

PREPARATIONS of MIXTURES (A/B):
First, a few comments about the preparation of mobile phase solutions. 

• Only the purely aqueous portion can be correctly adjusted for pH. Do not try and measure or adjust the pH of the organic or organic solution mixture. 

     There are two common methods of preparing binary mixtures (V/V) of mobile phase solutions.

• Method #1 is to fill a volumetric flask with a specific volume of the "A" solution, then fill the flask up to the line with the "B" solution.

• Method #2 is to fill a graduated cylinder (or volumetric flask) with a specified amount of "A" solution; fill a second graduated cylinder (or volumetric flask) with a specified amount of the "B" solution and then mix the contents of both together.

Whichever method you use, please fully document it in your HPLC method so anyone reading it will be able to accurately reproduce it. The two methods described above are both correct in design, but will result in solutions with different properties.

ABSORBANCE of UV LIGHT:
For HPLC grade solvent (*we should always use HPLC or LC-MS grade solutions in HPLC analysis), ACN has the lowest absorbance (~ 190 nm) of the two making it well suited for low UV applications. HPLC grade MeOH has a slightly higher UV cut-off, around 205-210 nm, limiting its use in the very low UV ranges. *Methods which require low UV wavelengths (<230 nm) should not use Methanol as the primary solvent.

SOLVENT SOLUBILITY:
There is a significant difference between ACN and MeOH in their ability to dissolve many types of buffer salts AND samples. These differences may be critical during method development as higher salt concentrations could lead to plugs, clogs or precipation. 

Solubility of the Mobile Phase:

• A common reason for gradient runs to show poor reproducibility or to fail may be associated with running high concentrations of buffer combined with high concentrations of organic solvent. Most aqueous / organic solutions containing salt solutions of less than 10 mM concentration are not likely to precipitate under most gradient conditions (running to a max of 95% organic, not 100%). If high percentages of organic solvent are mixed with more concentrated buffer solutions, then the higher salt concentrations may precipitate out of solution during the analysis (resulting in clogs, leaks, plugs and/or inaccurate results). Be cautious when mixing organic solvents and buffers together for gradient analysis. Make sure the solutions used will stay in solution and be stable at all concentrations used. Also verify that the buffering capacity is still present when high organic concentrations are used (as your buffer will be diluted). *Not sure if the salt will stay in solution? Just mix up a sample at the same concentration for a test. Look at it. Is there any turbidity or particulate visible? You should have your answer.

• Methanol's overall better solubility characteristics (better than ACN) mean that it often does a better job of dissolving most salts (esp NH4, K and Na) at higher concentrations resulting in better performance and less precipitation.

Solubility of the Samples (changes to Peak Shape, Selectivity & Retention):

• A fundamental requirement of liquid chromatography is that the sample fully dissolves in the mobile phase (initial mobile phase). Dissolve the sample in the mobile phase or in a slightly weaker strength solution (not a stronger solution) before analysis. This insures it will be loaded onto the head of the column as a concentrated slug improving peak shape and RSD. If the sample does not fully dissolve in the mobile phase then you are not in fact analyzing the whole sample. Another area where Methanol may be superior to ACN can be found in its ability to fully dissolve more types of samples. This improved solubility may result in better overall peak shape. Methanol also has different selectivity, often better than ACN (not just the elution strength) which may result in peaks eluting at different retention times than expecting. This is another reason why we always try different mobile phase mixtures containing either ACN or MeOH when developing RP methods. Please never assume that one solvent will be better than the other. Too many novice chromatographer's use only ACN as their main organic solvent for method development. Please don't make their mistake as such a strategy indicates a lack of practical experience and knowledge. You must first try them both separately (ACN & MeOH) to evaluate the results with your own sample (best to start with comprehensive gradients at different pH values, as applicable). You will be rewarded for putting in the initial time to test both types of solutions as no simulator has yet been developed which can predict a truly accurate result with your own sample(s). You may be surprised to learn how many samples show better peak shape and performance using MeOH solutions. If no improvement is seen, document it and move forward with more confidence.

BACKPRESSURE & OUTGASSING:

• ACN is less viscous than MeOH ( 0.34 vs. 0.54 viscosity, respectively) and if used alone will result in lower column and system back-pressures overall. Less gas will dissolve into ACN vs MeOH. Mixtures of ACN and Water will also exhibit an endothermic reaction (cooling the solution) which can trap gas inside the solution. If you pre-mix your mobile phase, let it rest for several minutes after preparation.Mixtures of ACN and Water will show a pressure max around 70% ACN (*This is an unusual characteristic well worth learning).
  

• MeOH is more viscous than ACN alone. It also has an unusual property where a 50/50 mixture of MeOH and Water will result in a much higher system and column back pressure than either MeOH or Water alone will (*ACN has a similar property, but the peak pressure occurs between 60-70%). The effect with methanol is very Gaussian with a peak pressure observed with a 50/50 mixture. An exothermic reaction results from an initial mixture of the two solutions (MeOH and Water) releasing some gas. When preparing solutions it is best to allow the solution to rest for a few minutes to out-gas before topping off or using in the HPLC system.

I hope that this short discussion about some of the differences between these two popular HPLC solvents will aid you in developing better quality HPLC and LC-MS methods.

Reference: Table of HPLC Solvent Properties

CHIRAL HPLC / SFC METHOD DEVELOPMENT (Alcohols):

We are experts in chiral HPLC and SFC method development of pharmaceutical samples and have operated a contract separations labs for nearly two decades. We have learned a great deal about developing fast and reliable racemate separation methods, often where other companies have failed. The knowledge we have gained has allowed us to develop over ten thousand new chromatographic methods for our clients. This has made our company the leading expert in the chiral separations field. 

I would like to share a tip with you regarding the use of different alcohols in chiral method development. That tip is to experiment with different alcohols during the method development process (*Please make sure the alcohol is compatible with your column!). Many of the normal and reversed phase chiral columns can be used with some unconventional alcohols to achieve excellent separations. These alcohols are often used isocratically at 100% concentration for HPLC methods and at levels ~ 10 to 20% for many SFC methods. We have had a great deal of success using 100% pure Methanol for HPLC methods on normal phase style chiral columns (though 100% Ethanol is still one of the best alcohols to initially choose). For SFC methods, Methanol, Ethanol and Butanol (plus mixtures of these) are still some of our favorite co-solvents.Note: SFC needs these alcohols to make the compressed CO2 more polar. Since most chiral drug compounds are resolved using Normal Phase chromatography, SFC is still limited in what it can resolve chromatographically vs HPLC. SFC will never replace conventional HPLC as SFC is far more limited in application (restricted in its range of polarity), but SFC is a worthwhile and important technique to supplement HPLC separations. 

• Here is a list of some popular alcohols (HPLC grade) worth using in your chiral method development: 

Methanol; Ethanol;1-Butanol; 2-Propanol; 2-Butanol and Acetonitrile (I know this last one is not an alcohol, but it is often overlooked in chiral method development. It works where other solvent systems fail for both HPLC (100%) and less so for SFC (10% as a co-solvent with Methanol). 

Many of the above liquids can be used at 100% concentration, but others require mixing with Hexane or Heptane to yield lower concentrations of alcohol. *Remember to always consult with the column manufacturer first to determine which solvents are safe to use.

• A note about acids. Weaker is often better in chiral method development. Examples: TFA at 0.01% and Acetic or Formic Acids at 0.1% concentration are often strong enough to ionize most compounds.