Troubleshooting HPLC Gradient Valve / Proportioning Valve / MCGV GPV Leaks. How to Identify Them.

HPLC pumps which utilize low-pressure mixing VALVES are known by names such as: "Ternary" (3-solvents) or "Quaternary" (4-solvents) pumps.These types of HPLC pump configurations use a single, high-pressure pump head coupled to a multi-port / proportioning valve and represent some of the most popular and versatile pump configurations offered. Featuring random access to multiple solvent bottles (more than two is always better), lower operating costs and less maintenance work provides you with one of the best platforms to develop new methods on. I highly recommend them for most, but not all, HPLC applications (vs. Dual pump, high-pressure "Binary Pumps").

• If your HPLC system utilizes a single, high-pressure pump head coupled to a multi-port valve, then please remember that in addition to pump head maintenance, regular maintenance of the multi-port / proportioning valve is also required.

A few weeks ago I was hired by well known Pharma company to solve a gradient method problem that I was told has stumped their best scientists for almost one year. The client presented me with their validated UHPLC method which suddenly developed a shift in retention time of all peaks. The shift was significant, about 10% of the previous values over a 20 minute run, and had been observed on two different, but similarly configured HPLC systems in their lab. Changing the column to a new one showed no change on either HPLC system. They were out of ideas.

• Before I reveal the cause of the trouble, let us briefly think about what types of changes can result in a small, repeatable shifts of peak retention times. Four common ones that come to mind are: 

(1) Flow Rate changes;

(2) Column Temperature changes;

(3) Column Fouling;

(4) Mobile phase composition changes. 

Start the troubleshooting by ruling out the easy causes first (#1, 2 and 3 above).  

• (1) Flow Rate: When the actual flow rate is in question, start by measuring it manually Never trust the instrument's display screen value or the software's value for flow rate. Measure it. An easy way to measure the flow rate involves timing the amount of liquid that exits the HPLC detector line after a defined period of time. For example: If your flow rate is set at 1.000 ml/minute, then using water, measure the time it takes to fill a 10mL graduated cylinder to the 5 mL line. It must take exactly 5.00 minutes (= 1.00 mL/min). Run this flow test on each pump channel.
  
• (2) Temperature: The HPLC method should be run under controlled column temperature conditions. Verify this. Retention times are a function of temperature (i.e. cooler temps usually result in longer retention times, warmer = shorter). The temperature should be stable (~ 1 or 2 degrees C).
  
• (3) Column Fouling: To prevent fouling, wash the HPLC column with a solution that is STRONGER than the mobile phase after each analysis. Use fresh, clean solutions. Verify that the samples are dissolved in the mobile phase (100% dissolved) and filtered before injection. Verify that the injection volume is less than ~3% of the column volume and the concentration of the sample is not too high (avoid saturating or overloading the column). Solubility is very important for both the sample and any additives used in the mobile phase (to prevent precipitation). Anything that "fouls" the column support will directly effect the retention times and often the peak shape too. Be aware of these causes and take action to avoid them.  *Replacing a suspect column with a new one is often an inexpensive way of troubleshooting a "peak" problem. Always have a NEW spare column on hand for testing. *Columns are consumable items.
  

• (4) Mobile Phase: Changes to the actual amounts of additives, pH or final composition of the mobile phase may impact peak retention times (sometimes, the peak shape too). After all, the final composition used was developed for the purpose of establishing a reliable and reproducible method of analysis. It must be controlled. We must take steps to insure the mobile phase preparation and delivery are accurate. Always prepare fresh solutions each day (esp. all aqueous solutions!). pH values may change after a few days (e.g. even in MeOH / acidic solutions), bacteria/mold/algae grow quickly in many solutions, even in the refrigerator, so only prepare what you need for the day. Evaporation of more volatile solvents (in pre-mixed solutions) can change their actual concentration (always protect them from heat and evaporation).
  

*There is another way that the mobile phase composition can change which often goes unseen. It can change during delivery to the column. The HPLC's low pressure proportioning valve that allows us to easily select and use different solvents can develop small internal leaks, resulting in valve cross-flow leakage. This cross-flow leakage allows liquid (or air, if the line is not connected) to be drawn out of one channel and into another, changing the actual mobile phase composition. This happens because the valve seals, esp if they have been left unused for a long time, can change shape (e.g. shrink) and begin to leak over time. Often the amount of leakage is very small (ul/min), but depending on the method, a small change may result in a significant change to the chromatography.

I reviewed the client's method parameters and concluded that the method met good chromatography fundamentals. Checking the flow rate (using a graduated cylinder) confirmed the flow rate was accurately shown. A review of their mobile phase preparation procedures and methods also appeared OK. Degassing of mobile phase and column temperature were also satisfactory. 

As I looked more closely at the two running HPLC instruments they used, I began to quickly zero in on the most likely problem. 

• A long stream of air bubbles were observed exiting the HPLC pump's gradient valve leading into the high pressure pump head, but no air bubbles were seen exiting the degasser's outlet line (IOW: The vacuum degasser may or may not be the cause, though it is critical to insure the degasser is clean and fully serviced before use. Have the degasser professionally serviced first before proceeding with troubleshooting. Using a damaged degasser will make it difficult to use the pump or run any valid tests as degassed solution is needed). This was observed on several of their HPLC systems, including the two used for this method. The fittings connecting the lines from the degasser module to the valve were correctly connected (as a loose connection would cause air to leak in and must be quickly ruled out). 

The cause was from one or more of the unused gradient valve positions leaking air into the flow path, changing the mobile phase composition. Of four possible mobile phase lines available (A,B,C,D), the client only had two lines connected to mobile phase bottles (A,B) with the remaining two lines left open to the air. The internal valve seals in the unused 'C' and 'D' valve positions had deformed, shrinking in size, sticking,leaking, allowing air to flow into the mobile phase on one of the channels. This resulted in a change of the organic composition % used in the method (due to a cross-flow leak), changing the peak retention times (as the actual mobile phase composition used in their gradient was different). I directed the HPLC pump's outlet line to waste, placed all of the solvent pickup bottle lines (A,B,C,D) in a beaker filled with IPA and allowed the pump to run pure IPA at 1 mL/min across each channel, one-at-a-time (100%), for ~ 20 minutes to re-hydrate the internal gradient valve seals. This was repeated with each valve position, then all of the lines were placed in fresh mobile phase solution, primed and flushed. The system was restarted and the method now ran showing the expected peak retention times. Instructions were provided which included regularly using all of the channels and valve positions plus flushing weekly to maintain valve operation. Use ALL of the lines and flush the valve(s) through all positions, one-at-a-time, on a regular basis. If prolonged flushing with pure IPA does not fix the leak, then it is time to replace the valve. All valves eventually wear out and must be included in maintenance inspections and checks. This is especially true when you purchase your HPLC system at an auction or from an 'equipment' reseller. Never assume that the 10+ year old HPLC valve is OK. Test it first (e.g. Acetone tracer test).

 

Acetone Tracer Test: If you suspect that a cross-flow leak exists on a gradient valve, then one method I use to check for leakage is to mix up a "Tracer" solution of pure organic (often ACN) that has 1% Acetone mixed in (for RP methods). Remove the column and replace with a restriction capillary. Place the tracer solution on the valve position you suspect may be leaking at an appropriate flow rate and set it for 0%. Run one of the other channels with 100% (pure ACN in this example) and monitor the UV (265nm) for the presence of acetone. If the acetone leaks into the channel you are using, it will be easy to observe on the UV trace. So called "bubble" tests (introducing and monitoring the position of a gas bubble into the low pressure solvent line) are not reliable leak detection methods for small leaks. Use a tracer such as acetone to find the leaking channel(s). You can read more about these types of Valve Leakage tests in this article (Click Here).

The Three Most Common HPLC Questions and How To Solve Them

The three most common HPLC related questions I am asked each week can be summarized below. Test your basic chromatography knowledge. Before reading the answers, see if you can answer them correctly on your own.

• "What Is Causing the HPLC Baseline, Pressure or Peak Retention Time(s) To: Wander, Change, Drift, Vary or be Unstable?"

• "How Should I Wash or Regenerate My HPLC Column?"

• "How Can I Tell if the Sample is Retained On the HPLC Column? or What Does It Mean When No Chromatography Took Place?"

Let us address each question in order and attempt to provide accurate answers (I have included links after each question to articles with more detailed explanations).

What Is Causing the HPLC Baseline, Pressure or Peak Retention Time(s) To: Wander, Change, Drift, Vary or be Unstable?

• Retention times must be reproducible from run to run.The causes of an unstable baseline and/or changing peak retention time(s) are often related. Common reasons include: Column temperature fluctuations, inadequate mobile phase mixing or degassing, leaks, dirty column, sample overload, lack of pH or buffering control (weakly ionizable samples can be very sensitive to changes). *Full Article link with detailed answers, here.

How Should I Wash or Regenerate My HPLC Column?

Note: Before proceeding with any column regeneration or cleaning procedures, always refer to the specific advice provided by the column manufacturer. Approved maintenance and cleaning instructions can often be found in the product guide or booklet which comes with the new column. Additional information can be found on the vendor's website or by contacting them directly.

• Two issues must be addressed to answer these types of questions. (1) Always wash your column with a specific column wash solution which is stronger than your analysis solution. The use of a stronger solution (In this context, "stronger" means better at dissolving the samples and faster at eluting them from the column) as the wash solution requires regular use to maintain the column. Failure to regularly wash your column may result in compounds accumulating on the column over time (fouling the column) resulting in poor reproducibility, higher back-pressures, contamination and/or poor peak shape. (2) Next, always wash your column after each analysis. This should be a separate step, not incorporated into your analysis method. The analysis method should not include the column re-equilibration steps at all. A second, separate wash method should always follow each analysis method which includes the rinsing of the column with a "stronger" solution for an adequate period of time, then adjustment back to initial conditions where re-equilibration can take place to get it ready for the next analysis run. These are fundamental guidelines of good method development and follow well established principles. Developing methods in this way should increase the lifetime of your columns and improve the reproducibility of results obtained (better %RSD run-to-run).

For more information on washing bound proteins off RP HPLC columns, please refer to this linked article found here.

How Can I Tell if the Sample Is Retained On the HPLC Column? or What Does It Mean When the Sample Comes Out At or Near the Column Void Volume?

• Chromatography is a tool which when used properly adds one or more additional dimensions of physical or chemical characterization information to your analysis data. It does so first by using on-column RETENTION. Samples must be run under conditions which allow the material to interact with the chromatography support for a period of time. We define this time as the retention time. A sample which does not interact at all with the column support material will elute off the column early (and not be retained) at the "column void time" (or column dead time). We refer to this void time as the "T zero" time. When a sample elutes at or near the T zero time, no chromatography has taken place and no method has been developed. It is as if the HPLC column was not used. How do you know what the "T zero" time is (it will be different for different methods)? You must first calculate the HPLC column's dead volume. Once you know the column dead volume and flow rate, you can calculate the T zero time. A scientifically valid HPLC method will include conditions which retain the sample on the column for a long enough period of time to insure that it is interacting with the support. This allows for separation from other compounds to take place and is the purpose of chromatographic resolution. Without this retention mechanism, you are just flow-injecting the sample past the column and skipping all chromatography. It would be far simpler to just place the sample in a spectrophotomer cell as no retention or additional data would be obtained using that technique.

• When first learning liquid chromatography, two of the very first calculations you must learn to use in HPLC are: Column Dead Volume (aka: Column Void Volume) and the K prime of a sample (aka: Peak Capacity Factor). Do you know how to calculate these? They are calculated and reported for each method used. You should be able to tell anyone who asks you what the values are for each method. A chromatographer must know and understand them before using an HPLC system or running a method. They are also critical to method specificity and proper validation. Here are links which after reading and practicing, should make you an expert in these two fundamental calculations. 

"Determination of HPLC Column Dead Volume / Dead Time (T zero)";

"K Prime (also known as: Capacity Factor or Retention Factor): One of the Single Most Important HPLC Parameters of All".

So, how did you do answering these basic questions? If you have put in the needed study time and practical experience to learn and use these fundamentals of high-performance liquid chromatography, then you should have been able to easily provide correct answers to all three questions. If not, then it is time to go back and study up on those basic liquid chromatography texts and article links, plus get more supervised hands-on time with the instruments.

Diagnosing & Troubleshooting HPLC Pressure Fluctuation Problems (Unstable Baseline)

Few things in chromatography are more frustrating than dealing with large pressure fluctuations (>1% ripple). If the pump pressure is unstable, and fluctuating up and down, then it will negatively impact your ability to analyze, measure and integrate sample peaks in a reliable manner. A smooth, flat baseline is needed to run and develop methods, collect the data (peaks), integrate and report the results which are reproducible. Baseline instability during an analysis may lead to the entire analysis being declared invalid.

So what causes the HPLC pressure to sometimes fluctuate in a wild manner up and down on your HPLC system? Unfortunately, many things... Most result from poor training, incorrect operation techniques, but some are maintenance related so be sure and keep your chromatograph in excellent condition. Maintain a logbook for each instrument and record what types of maintenance and service have been performed over-time, with the date and list of parts used/replaced. Additionally, maintain a preventative maintenance schedule (e.g. every six months) to inspect and clean the entire HPLC system to check condition, verify operation and minimize unproductive down time. 

HPLC Pump or System Pressure Fluctuation Causes and Solutions:

• Air / gas In the Liquid or Mobile phase (Failure to Degas Mobile phase OR loose fittings) --- Air gets into the system due to a leak or from gas trapped in the mobile phase. Find and correct the cause of the leak and/or degas the mobile phase (use continuous Vacuum degassing or a Helium sparging system only). Leaks are the most common cause of instability, but insufficiently degassed solution is a close second. Make sure your degasser is working 100% correctly (they require professional servicing every 5 years). HPLC pumps require degassed mobile phase for reliable operation.

• Loss of Prime. Improper Priming of the System --- Failure to flush ALL of the lines with freshly degassed mobile phase, before use (every day), will often result in all kinds of instability problems until all of the old gas-filled mobile phase has bee purged from the system. *This could take many column volumes of liquid. Make sure you account for any vacuum chamber volume too. Properly prime the pump heads before use.

• Sticking Check Valve(s) --- If air is exiting the pump outlet, the pump will not function properly. Both Inlet and Outlet valves should be inspected. Remove and clean the check valve(s). Be sure the pump is fully primed with liquid as the check valve might just have an air bubble in it (common on Waters, Thermo and Shimadzu systems). Sometimes sonication of the valve for ten minutes in a beaker containing warm solvent does the trick (e.g. MeOH or IPA/Water). Though very rare, ACN has a bad reputation for polymerizing in solution. If the system has sat unused for a long time OR was not properly flushed out when last used, it is possible that particulate matter may clog the flow path. Small sticky particles may form (ACN polymerization) and cause the check valve to stick inside the housing (use fresh, filtered solvents only to prevent these problems). Clean and inspect any suspect valve first. Replacement of the check valve may be needed in some cases to restore operation. Note, this problem of "sticking" check valves is most likely to be an issue in HPLC pumps with mechanical (gravity or spring) check valves with ruby balls, not modern style active inlet check valves ("AIV") which are electromechanical (solenoid valves) and are very reliable, much less susceptible to these problems. In any case, verify operation of all valves while under pressure (backpressure is needed for them to function correctly).
  

• Worn Pump Piston Seals --- Commonly observed as rapid up/down spiking on all channels and an inability to maintain or produce backpressure (the pump will often prime with no problem, as this is done at low-pressure). Run a formal pump high-pressure leak test at max pressure to confirm (remove the column and replace with a calibrated backpressure restriction line for all testing). Clean pistons and replace piston seals to repair (you should have spare pistons and seals on hand). *Seals are a maintenance item so expect them to wear out and need regular replacement.

• Flow rate too low (may be inappropriate for system). Running at a flow rate that is below the optimum range of the specific instrument (i.e. System rated for 200 to 2,000 uL/min, but run at 100 to 200 uL/min or at the limit of the range) may result in an unstable baseline. The cause may be due to pump cavitation, loss of prime, non-optimized piston stroke volume.

• HPLC System Back-pressure too low to maintain prime in system. Most types of analytical HPLC systems require a minimum system back-pressure of 40 or more bars to maintain enough pressure (mechanical compression) on the component parts to run in a reliable fashion (*Water's Article number: 32564 states the back-pressure must be at least 1000 psi for their Alliance systems). Too low a pressure often results in a loss of prime, cavitation, mixing problems, turbulence and poor reproducibility. Correct sizing of column, particle size, flow rate and mobile phase composition should all take into account achieving enough back-pressure on the system to maintain a stable baseline throughout the entire analysis. Monitor the system back-pressure at all times for stability. High quality research grade HPLC systems are often capable of maintaining stable isocratic flows with less than 1% ripple and 0.2% ripple common ("ripple" is a term we often use to describe the pump's pressure output over time relative to the baseline (S/N)).

• Mixing Problem (gradient or isocratic online mixing) --- If your active mixer or proportioning valve (AKA: Gradient valve) is defective or dirty, then one or more of your mobile phase channels may not be getting to the pump. Air would most likely be mixing with the mobile phase causing the unstable flow. Clean or replace the valve. Note: Always try flushing the gradient valve with pure IPA, then DH20 for about twenty minutes. This sometimes restores operation by wetting and flushing the internal seals (which may dry out).

• Wrong Pump Solvent Compressibility Settings --- In HPLC we routinely subject different liquids to very high pressures which can result in measurable liquid compression. The degree of actual compression for each liquid varies, but the modern HPLC pump can compensate for this to improve the accuracy of the mixing and flow delivery.  Most pumps provide for user adjustable solvent compressibility values. If the value input varies a great deal from the actual liquid in the system, then it can result in pressure fluctuations. Example: Water has a value of 46, but Methanol 120. Using the wrong value can cause instability.  

• Poor Solubility, Mobile Phase  --- Sometimes the mobile phase which has been prepared (or mixed online) is not 100% soluble. This could be due to an inorganic salt additive which has not gone into solution or failure to fully mix and filter the mobile phase before use. Ultrasonication, a bit of heat and stirring for 20 minutes can help to get everything dissolved. 

• Dirty inline filter --- A fouled or partially plugged filter can disrupt the normally smooth flow into a turbulent one. Some are installed as part of the pump (i.e. HP/Agilent brand pumps) and should be changed out every month (Yes, for the PTFE frit, replace it once a month with a new one). Other systems use these pre-filters downstream of the pump before the injector. Clean or replace all filters frequently. If used in your system, these are regular maintenance items and should be part of a general 'PM' program.

• Dirty Solvent Pickup Inlet Filters: These can become obstructed or fouled over time (esp. if used with aqueous solutions!). Just as with any built-in filter, the multiple solvent inlet pickup filters should be cleaned or replaced on a regular basis to prevent particulate or any material which may contaminate or restrict the flow path from entering the system. Mobile phase pickup filters are often 10 to 20 um and connect to the bottom of the low pressure (e.g. Teflon) solvent lines in each bottle. If you use 316 Stainless steel filter (recommended for organic solvents), they should be removed, cleaned in an ultrasonic bath, rinsed and replaced monthly. If you use sintered glass or other disposable type filters (often used with aqueous solutions), they should be disposed of on a regular basis and replaced with new ones (replacement, not cleaning is recommended because sintered glass can not be sonicated and should be disposed of to prevent bacterial, mold or fungal contamination). A quick way to check if one filter is causing the pressure to fluctuate is to remove the filter from the one line, then re-test the system. If the problem goes away, then returns when you re-install the filter back on the line, the filter may be obstructed (replace it),

The above list includes some of the most common reasons for unstable baselines. Other non-pump related causes would include a bad / old detector lamp(s) or contaminated mobile phase. To find the cause, test and verify the operation of each component part of the HPLC. Troubleshooting Advice: Test one part at-a-time, before moving to the next part. Never assume anything, test, re-test and verify or prove at each step.