K A T H O L I E K E H O G ESC H O O L SI N T-L I E V E N DEPARTEMENT GENT TECHNOLOGIECAMPUS GEBROEDERS DESMETSTRAAT 1 9000
GENT
M attéo B ryckaert Academic year 2009-2010
O ptimization of H P L C apparatus and method
IUT A de Lille Departement Chimie Rue de la Recherche – BP 179 59653 VILLENEUVE D’ASCQ
SU M M A R Y To do validate my diploma from the Institute of Technology of Lille I had to do a training period. I decided to do mine abroad and in an English speaking environment. I had the opportunity to do it in Gent at KaHo Sint-Lieven Hogeschool in the analytical laboratory under the supervision of Annick Boeykens, Kurt Haerens, Patrick Demeyere and Luc Pinoy. During my training period I discovered HPLC and learned a lot of things about it, how it works, what can be done with it, what problems you can encounter and much more. This report will sum up my stay in the lab and my studies.
A C K N O W L E D G E M E N TS I would like to thank the following people that helped me during my staying in Belgium and in my work: - M. Sc. Annick Boeykens for trusting me with this project, overseeing me during my training course helped me when I needed, and allowed me to learn a lot. - Ph.D. in Chemistry Patrick Demeyere for accepting me in the university and in his laboratory, giving me the opportunity to study HPLC and discover the world of research. - M. Sc. Kurt Haerens for his help in the lab. - Mr. Ahmed Mazzah for giving me the opportunity to study abroad. - Bie Van de Casteele for helping me finding an accommodation and helping me during the length of my stay. - All the staff from the multiple labs who helped me during my stay and my studies. - Davy Bonny for his support and his help in my research. - Ph. D. Luc Pinoy for trusting me with this project and helping me in my research. - Arnaud Caillier and Roxana Queste for giving me the opportunity to go abroad and helping me getting this trainning course and everything with it.
CONTENTS 1
T H E R E C E I V I N G I NST I T U T E A N D L A B O R A T O R I U M ............................. 5
2
PR O B L E M SE T T I N G A N D PR O J E C T SU BJ E C T .......................................... 6
3
L I T E R A T U R E ST U D Y ........................................................................................ 7
3.1
THEORY OF HPLC ................................................................................................ 7
3.1.1
Introduction of H igh-Performance L iquid C hromatography ........................... 7
3.1.2
Wor king mechanism .............................................................................................. 7
3.1.3
F actor that affect liquid chromatography ........................................................... 8
3.1.4
T ypes of H P L C ....................................................................................................... 9
3.1.5
T ypes of detector .................................................................................................. 14
3.1.6
C hromatogram and efficiency ............................................................................ 16
4
M A T E R I A L A N D M E T H O DS .......................................................................... 18
4.1
HPLC APPARATUS ............................................................................................. 18
4.2
HPLC MANUAL .................................................................................................. 19
4.2.1
Starting the H P L C ............................................................................................... 19
4.2.2
C hromquest .......................................................................................................... 20
4.3
METHOD .............................................................................................................. 21
4.3.1
Procedure.............................................................................................................. 21
4.3.2
A lteration of the original procedure .................................................................. 23
4.4
TROUBLESHOOTING ........................................................................................ 23
4.5
OLD COLUMN VS NEW COLUMN .................................................................. 24
5
R ESU L T A N D D ISC USSI O NS .......................................................................... 29
5.1
DETECTION OF COMPONENT Y AND COMPONENT X (IN EACH COLUMN) ............................................................................................................. 29
5.2
TEMPERATURE STABILITY (OLD COLUMN) .............................................. 30
5.3
REPEATIBILITY/REPRODUCIBILITY ............................................................. 31
5.4
ODD RESULTS .................................................................................................... 33
6
C O N C L USI O N .................................................................................................... 35
7
SO U R C ES ............................................................................................................ 36
5
1
T H E R E C E I V I N G I NST I T U T E A N D L A B O R AT O R I U M
The Katholieke Hogeschool Sint-Lieven is a catholic university which appeared with the merging of 8 institutions in 1995. It has study fields in multiple areas such as biotechnology, health care, teacher training, etc… It has multiple research labs and is in cooperation with multiple company for its researches especially in chemistry and biotech. I have worked in the analytical laboratory of KaHo Sint-Lieven in cooperation with an external laboratory.
6
2
PR O B L E M SE T T I N G A N D PR O J E C T SU BJ E C T
KaHo Sint-Lieven wanted to come up with a way of separation between component X and component Y from a sample. Therefore there was a need to have the ability to measure the amount of component X and component Y in samples. Component Y already had a way of being measured using ion chromatography and component X didn’t have any direct way to measure it. Furthermore the analytical lab of KaHo Sint-Lieven recently acquired a new HPLC machine and wanted it to be operational and with a manual for the next year for the students to be able to work on it. By combining the need of a way of measuring the component X and setting up the new HPLC my project subject appeared.
7
3
L I T E R AT U R E ST U D Y
3.1
THEORY OF HPLC
3.1.1
Introduction of H igh-Performance L iquid C hromatography
HPLC is a liquid chromatography technique which as his name suggests is efficient. The difference between HPLC and other liquid chromatography is that it usually gives narrower peaks, gives better separation and better limits of detection. Its disadvantage is the fact that it is more expensive and requires more skilled operators. It requires a difference in retention time and a good efficiency to be able to separate 2 compounds. It’s improvement from regular liquid chromatography is a better separation of the compound through the use of improved columns and components. The selection of the column and of the mobile phase is what will determine the quality of the result because they determine the separation of the components (2).
3.1.2
Wor king mechanism
An HPLC is usually made of a solvent reservoir, a pump, an injector, an analytical column, a detector, a recorder and a waste reservoir. Figure 1 shows the HPLC configuration used during the project.
F igure 1
The HPLC used
8
An HPLC analysis starts with the solvent reservoir. The solvent is used to carry the sample in the system. It is first filtered by an inlet solvent filter to remove the particles. Then there is the sample injector in which the sample will be injected using a syringe, manual or with an auto sampler. The sample injector is made with a sample loop of a predetermined size (and volume) to be able to repeat the injection precisely. This injector uses a loop system to allow the sample to enter the flow path. The loop system works in 2 parts; presented in Figure 2. First, in load sample modus, the loop is filled with the sample and then; in inject sample modus, the sample is send to the column.
F igure 2
Functioning of the loop
Then the mix of sample and solvent goes through the column in which the components are separated. There are multiple ways to separate the components; this will be discussed later. Then the separated compounds pass through a detector and then go in the waste reservoir. Depending on the type of detector different parameters are used to spot when the compounds go through it, it can be absorbance, the refractive index, the fluorescence or the conductivity. In this project a UV detector was used. Then the detector sends the data to a recorder (often a computer) which interprets the data (1).
3.1.3
F actor that affect liquid chromatography
3.1.3.1
The analyte
First of all the chemical to be analyzed, has to be a liquid that can be injected in the column. There has to be a difference in retention between the compounds to be separated.
9
3.1.3.2
The mobile phase
In HPLC the retention of solutes will depend on interaction involving the mobile and stationary phases. 2 types of mobile phases can be considered. The strong mobile phase, which is a pure solvent, or a solution, that quickly elutes a retained analyte from a column. This situation occurs when the analyte favors staying in the mobile phase rather than on the stationary phase, meaning if the analyte is more soluble in the mobile phase or has low interaction with the stationary phase. The weak mobile phase elutes a retained analyte more slowly. It occurs when the analyte has better solubility in the stationary phase than in the mobile one or when the mobile phase promotes a good interaction of the analyte with the stationary phase. Changing the composition of the mobile phase is a good way to change the retention time of analytes. The use of a constant mobile phase composition is known as an isocratic elution. It is the simplest and cheapest way to work. The solvent is usually a mixture of a solvent with a weak mobile phase and one with a strong mobile phase to get a mobile phase with an intermediate degree of retention for the analyte. The result is the fact that it makes it difficult to elute all solutes with a good resolution in a fair amount of time. Which lead us to the second type of elution, the gradient elution. For this method is started with a weak mobile phase to allow the weakly retained solutes to elute more slowly and then over time a stronger mobile phase is added to elute more quickly the highly retained solutes (2).
3.1.4
T ypes of H P L C
There are 3 different types of liquid column chromatography depending on the column and the nature of the stationary phase.
3.1.4.1
Adsorption chromatography
Adsorption chromatography is also known as liquid solid chromatography. It works by separating the products based on their absorption to the surface of the support as shown in Figure 3. It uses the same material as the stationary phase and support. It involves the binding
10
of an analyte (A) on the surface of the column and the fact that it would be in competition with the products in the mobile phase (M) for the binding sites.
F igure 3
Presentation of adsorption chromatography
The process can be represented with the following equation: A+ n M-Surface ↔ A-Surface + n M Therefore the retention of the analyte will depend on the binding strength of A to the support and the area of the support. It will also depend on how much mobile phase is displaced from the surface by A, the strength with which the mobile phase binds to the support. The strength of the mobile phase is characterized by the elutropic strength (ε°). It indicates how strongly a particular solvent or liquid mixture would adsorb to the surface of a given support. If the elutropic strength is high the liquid will strongly adsorb to the given support, which will prevent the analyte from binding to the support. Therefore a liquid with a large elutropic strength would be a strong mobile phase for the support because the analyte, unable to bind, would elute quickly from the column. Silica (SiO2) is the most popular support for adsorption chromatography. It is because silica is polar and therefore will most strongly retain polar compound. A strong mobile phase for silica will also be polar. Sometimes also alumina (Al2O3) is used for the same reason but it can retain some polar solutes so strongly that they are irreversibly adsorbed to the surface. Sometimes carbon based materials are used as non-polar support to allow the column to retain non-polar solutes with a strong mobile phase that is non-polar. A longer column for any of these supports will lead to stronger analyte retention.
11
Adsorption chromatography is widely spread due to its low cost and the fact that silica and alumina work well in separating compounds in an organic solvent. It is also useful in separating geometrical isomers and chemicals from a given class of substances. But adsorption chromatography can also create some problems for analytical applications. Those problems are the heterogeneous nature of the surface on silica or alumina and the ability for these surfaces to act as a catalyst for some chemical reactions. They also require high quality solvents for consistent elutropic strength. And they can also create non reproducible retention for polar compounds (2).
3.1.4.2
Partition chromatography
For this type of chromatography, presented in Figure 4, a stationary phase is coated on a solid support. It is based on the difference between the solubility of the components in the mobile and stationary phase, it usually is silica but can also be other materials.
F igure 4
Presentation of partition chromatography
There are 2 types of partition chromatography: normal-phase chromatography and reversephase chromatography. The main difference is the polarity of the stationary phase. Normal-phase chromatography or NPLC uses a polar stationary phase; therefore it retains polar compounds more strongly. The weak mobile phase in NPLC is a nonpolar liquid. Reverse-phase chromatography or RPLC uses a nonpolar stationary phase, which is opposite to the polarity of the stationary phase. It is the most common type of LC because the weak mobile phase is a polar solvent such as water.
12
The retention of solutes in these types of chromatography can be explained by the solubility equilibrium in the following equation: A mobile phase ↔ A stationary phase
It can be described by a distribution constant called KD: KD= It can be directly related to the solute’s retention factor (k) k=KD(Vs/VM) Where Vs is the volume of the stationary phase in the column and VM is the column void volume. The strength of a mobile phase in partition chromatography can be measured using a solvent polarity index (P). The value of P is low for a nonpolar solvent and increases with the polarity of the solvent (2).
3.1.4.3
Ion-exchange chromatography
The stationary phase has an ionically charged surface of opposite charge to the ions in the sample. Solutes are separated by their adsorption onto the support as shown in Figure 5.
F igure 5
Presentation of ion-exchange chromatography
13
The reaction can be described with competition of a sample cation (A+) and a competing cation (C+) for a negatively charged ion-exchange site on a support as shown in the following reaction: A+ + Support-(C+) ↔ Support- (A+) + C+ The equilibrium constant KA,C for this kind of ion-exchange reaction is called a selectivity coefficient. It can be calculated using the following equation: KA,C= It will describe how effectively the analyte will compete with the competing ion for the sites. As it increases, the retention will get higher. There are several factors which will affect the retention of charged analytes: -
the nature and accessibility of the ion-exchange groups on the support;
-
the type and concentration of analyte ions;
-
the nature and concentration of the compositing ions in the mobile phase;
-
the pH of the mobile phase, if any of the products used are weak acids or bases.
There are two types of stationary phases used in ion-exchange chromatography. The first type is a “cation-exchanger” which has a negatively charged group and is used to separate positive ions. The second type is an “anion-exchange” which has a positively charged group and is used to separate negative ions. For the support silica can be used if it has been modified to contain charged groups on its surface. Another support that is commonly used is polystyrene (2).
3.1.4.4
Size exclusion chromatography.
It is a technique that separates substances based on the difference in their size. The column is filled with porous material with controlled pore sizes. The product then goes through the column and the molecules travel speed depends on their size, the biggest one will be the first to come out as shown in Figure 6.
14
F igure 6
Presentation of size exclusion chromatography
The ideal support for SEC is a porous material that doesn’t react with the solute. Many of the products used for the other LC can be used. The most important thing is that they have a porous structure. The range of pore sizes will determine the size of the compounds that they will be able to separate (2).
3.1.5
T ypes of detector
There are several types of detector available for HPLC. Some of them are specific some are general. They can measure the refractive index, absorbance, fluorescence, conductivity or electrochemical properties. Each of them has a specific detection limit and can detect specific compounds. Here are some examples of the most common general detectors.
3.1.5.1
Absorbance detector
This detector measures the ability of a liquid to absorb light at a particular wavelength. It can be used for all compounds that absorb light. The mobile phase goes through a “flow cell” which allows the mobile phase and analytes to pass through the detector in a continuous manner. There are three types of absorbance detector, a fixed-wavelength absorbance detector, a variable one and a photodiode-array detector (PDA). The first type is used to monitor at a specific wavelength depending on the light source chosen depending on if you want to work in UV or visible. The second one allows the monitored wavelength to be changed over a wide range. The last detector uses multiple detectors cells to monitor the absorbance of the solution at multiple wavelengths (2).
15
The photodiode-array detector works by using 2 different lamps, a deuterium lamp and a tungsten-halogen lamp. Each lamp has a specific wavelength range and by combining the two with a beam combiner we can work on the combination of both ranges. The beam combiner is used so that the beam from the tungsten-halogen lamp is parallel and coincident with the light from the deuterium lamp. The combined beam is then focused with a lens on the flow cell. The light beam is then partially absorbed as it goes through the sample and then exit into a beam shaper. The beam is then diverted by a folding mirror to a fixed grating. The folding mirror is used to shorten the optical bench. The spectrum from the grating is then focused to a photodiode array which sends the data received to the computer. All of this can be further understood by watching Figure 7 which shows the detector optical system (4).
F igure 7
3.1.5.2
The UV detector
Refractive index (RI) detector
This detector measures the ability of a liquid to bend light. Therefore the refractive index will change as the composition of the liquid will change. It will detect the difference of the refractive index of a solution compared to a reference solution. It is one of the most universal detectors for HPLC. The only requirement is that the compound has a different refractive index from the mobile phase (2).
3.1.5.3
Evaporative light-scattering detector (ELSD)
The ELSD can be used for any compound that is less volatile than the mobile phase. The ELSD first turn the mobile phase into a spray of small droplets, then the solvent evaporate from the droplets and the small solid particles that contain the nonvolatile compound remain.
16
They are then passed through a beam of light where they scatter some of the light. The number and size of the particles are determined through the measurement of this scattering. There are also some more selective detectors such as the fluorescence detector which can only be used with fluorescent chemicals. The electrochemical detectors measure the ability of an analyte to undergo either oxidation or reduction and therefore can only be used for electrochemically active compounds. The conductivity detector measures the ability of the mobile phase, and its content, to conduct a current in an electrical field and therefore can only be used with ionic compounds (2).
3.1.6
C hromatogram and efficiency
After an analysis a chromatogram is obtained; a chromatogram is given in Figure 8 . Each peak can be described with its retention time tr, the time at which it is at its highest point, and its area. On a chromatogram the different compounds that are in the sample can be seen, each peak usually represents one compound. They can be determined by watching their retention time and their amount determined with their area.
F igure 8
Chromatogram
17
Some parameters can have an influence on the chromatogram obtained in the end; the most important are the column and the mobile phase. Depending on each, peaks can appear or not. The both have an influence on the separation of the products. The other parameters that can be changed are the flow and the injection volume. The flow doesn’t have a lot of influence on the result in the end; it changes the retention time and how narrow the peaks are. The injection volume just changes the height of the peak and if the peak is saturated or not.
3.1.6.1
Band Broadening and theoretical plate
To get the best result a good separation of the compounds has to be achieved; and therefore sharp and symmetrical peaks, meaning a small band broadening. The band broadening can be described using wb and wh which are respectively the baseline width and the width at half height of the peak as can be seen in Figure 9. Those values can be linked with the standard deviation, usually wb = 4σ and wh = 2.355σ. To describe how good the separation is the number of theoretical plates (N) is used. Its general definition is N = (tr/σ)². Therefore N = 16(tr/ wb)² or N = 5.545(tr/ wh)². The higher N the better the peak form is.
F igure 9
Baseline width and width at half height
There can also be written about the height equivalent of a theoretical plate (HETP or H). Its value can be calculated using N in the following equation: H=L/N with L the length of the column. A small plate height is good for the peak form (2).
18
4
M AT E R I A L A N D M E T H O DS
4.1
HPLC APPARATUS
The HPLC used during the experiment is assembled with 4 parts and presented in Figure 10. The solvent reservoir, which can hold up to 4 bottles with different liquids. Each bottle has its own tubing with an inlet solvent filter allowing us to do a gradient if necessary.
Solvent reservoir
Auto sampler
Pump
F igure 10
UV detector
HPLC equipment
The pump, which pumps the solvent from the bottle in the solvent reservoir, can be set to do a gradient. The pumping program can be installed directly on the pump or by using the computer to create the program and then send it to the pump. The auto sampler, which can hold up to 3 trays of 40 1.5mL vials each, and has a column oven. The vial is taken with a robotic arm and is then put under a syringe which will take the set amount of sample and then the rest of the syringe is filled with a solvent (taken from a bottle). It is then injected in the column which can be preheated. The UV detector measures the ability of the liquid that comes out of the column to absorb light at a precise wavelength or between two wavelengths. The detector sends the data to the computer to be integrated.
19
All parts could be monitored and set using the computer. The pump and the auto sampler could also be set directly on the instrument (1).
4.2
HPLC MANUAL
4.2.1
Starting the H P L C
-
Turn on the computer and all 5 parts of the HPLC.
-
Log for the computer: administrator, password: adm_001.
-
Launch Chromquest, log: engineering, password: abt1305!
-
Put your eluents bottles to the sonic bath to degasify.
-
Shake the tubing in the bottle to remove the air bubbles on the filter.
-
Check that the valve is opened on the pump.
-
Press “Purge” on the pump and set your eluent composition in the “Blend” area and a flow of 10 for 2-3 minutes.
-
When purge is over the pump should indicate “init”. Wait until “ready” appears.
-
Close the valve and then wait for the pressure to go up and stabilize.
-
Check that the waste is flowing properly from the UV detector.
-
Click on “Menu” on the auto sampler; go on “commands”, press “enter” and again on “Flush Sample Syringe”. Set the amount; 500 μL is usually enough. While it is flushing check for air bubbles in the tubes and in the syringe. Repeat the operation as many times as needed to remove the air bubbles.
20
4.2.2
C hromquest
-
Click on the “method” tab and on “instrument setup”.
-
In this folder all the parameters for the analysis can be changed.
-
For the trigger “External” is recommended.
-
When going on the p4000 tab there is a table with the amount of each solvent at a few given time. The last time in the table should be the end of the run. If no gradient is used, two lines with the same thing except for the time should be filled in.
-
Every time something is changed, a window will pop out called “Method Audit Trail”. Type “Test” in the space given and click on “apply to all” and then “Ok”.
-
Save the method.
-
Go on the “control” tab and click on “Download Method”.
-
Click on the “sequence” tab and “Edit”.
-
Here the sequence of the analysis can be filled in.
-
If a calibration is done: In the “level” column type numbers from 1 to the number of calibration samples. The “Run type” should change by itself to “Calibration”. In the “Vial” column type the vial for each sample in the auto sampler to analyze. In the “Volume” enter the volume to inject (knowing that the sample loop is 100 μL). Enter the sample ID for each. For the method click on the gray and green button in the squares and fetch the method to use for each sample. Then enter the filename for each. Enter the amount of sample in each vial. Change the ISTD amount if needed (Internal Standard) and the multiplier.
-
To measure an Unknown solution: the “level” column type 0, the “Run type” should change to “Unknown” fill the rest of the column as for the calibration except in the “sample amount” don’t type anything.
-
Save the sequence.
21
-
Go on the “control” tab and “Instrument stats” to check if all parts of the HPLC are ready.
-
Click on the 2 green arrows to launch the sequence just entered or the blue arrow of only one analysis has to be done. When clicking on the blue arrow don’t forget to put the sample amount and the vial to use.
4.3
METHOD
4.3.1
Procedure
Component X and component Y are retained and separated on a C18 column with the help of an ion pairing reagent. Detection is performed by UV for component X in combination with RI for the component Y. The solvent used for the HPLC is a mixture of: 2.1628 g of 0.1 mol/L sodium Octane Sulphonate (OSA), 1 g of H3PO4 85 %, 1000 mL of ultrapure water and 100 mL of methanol. All reagents must have analytical grade purity. The column advised is a Zorbax SB-C18 150 mm x 4.6 mm x 3.5 μm. It is a column for reverse-phase HPLC. The flow in the column has to be 1 mL/min which would give a column back pressure of 132 bar. The injection volume has to be 25 μL. The column has to be at 30 °C and the optical unit of the RI detector at 40 °C. The wavelength for the analyses is 210 nm. The retention time for component X is supposed to be 4.5 minutes and 12.0 for component Y. Component X is analyzed with an UV detector; it has to be calibrated at 210 nm with standards from 200 to 2000 ppm. The component Y is analyzed with a RI detector; it has to be calibrated from 250 to 7000 ppm. All standards are made using component X and component Y chloride of purity pro analytical and ultrapure water. Before being able to measure the amount of component X in a sample, a calibration curve has to be done. A calibration curve is a curve made from the area of a few samples with a known concentration. On the x axis, the concentration is expressed and in the y axis the area is expressed. To create it several samples are made with a known amount of component X (in this case 5). Analyze them to see the area of the peak for the product measured. Put all those value in a table (or the program can do it for you) and ask the program to create trend line.
22
When the equation of the curve is calculated, the amount of the product in a sample can be calculated using its area (if the goodness of the fit is high enough (r²), r²≥0.95 at least). In Figure 11 an example of a calibration curve made with the new column is given. First the area of each peak is obtained with the HPLC and then the amount of product in each sample is added to Table 1.
Table 1.
Area and concentration of different standards
Area 4241732 2079689 1065823 526936 265887
Concentration (ppm) 3181,5 1590,75 795,375 397,688 198,844
Using this table a calibration curve can be drawn.
Calibration curve 4500000 y = 1328,x R² = 0,999
4000000 3500000
Area
3000000 2500000 2000000 1500000 1000000 500000 0 0
500
1000
1500
2000
2500
3000
3500
Concentration (ppm)
F igure 11
Calibration Curve
The calibration curve has to be checked with a sample called an “unknown” which is a sample with a known amount of product but the detector doesn’t know it. If the result found is acceptable the calibration curve is good.
23
With this calibration curve the amount of component X of any sample can be determined as long as it is done using the mobile phase used for this calibration curve.
4.3.2
A lteration of the original procedure
All the analyses were done using only an UV detector; therefore a peak for component Y was never obtained. A peak was searched between 192 and 218 nm but none were found around the expected retention time. At the beginning another column was used. This was to test the procedure with one of our own column. It was a SC18R5 250 mm x 4.6 mm x 5 μm. The difference was that the retention time of component X was around 8.5 minutes. One other difference is the fact that the loop used was a 100 μL one instead of a 25 μL. As the temperature of the sample has been discovered to be very important, all analyses were done at room temperature.
4.4
TROUBLESHOOTING
In this part different problems, often encountered, will be explained and solutions given. -
Problem: The curve on the computer is drawn at 3 time the normal speed (for example it reaches 3 minutes of analysis in 1 minute). Solution: Stop the analysis and wait for the length of the analysis is supposed to last to be sure that the entire product went through the column and start the analysis again.
-
Problem: The curve doesn’t start on the computer but both the auto sampler and the pump says “RUN”. Solution: Stop the analysis and wait until the normal time of the analysis has past; this to be sure that all the product went through the column. Check that all the parts of the HPLC are well connected to the computer and between them and start the analysis again.
24
-
Problem: An unexpected peak appears. Solution: Check if the preceding analysis wasn’t stopped too soon. Check that the sample and solvents are not too old. There could also be some bubbles in the tubes from the solvent or in the syringe, purge both of them until there are no more bubbles. Or there could be an unwanted product in the sample.
-
Problem: The same sample doesn’t have the same area at all after multiple analyses. Solution: The sample isn’t at room temperature and as it gets warmed its area diminishes, wait for it to be at room temperature and start the analysis again. Or if the sample isn’t homogenous enough, shake it.
-
Problem: The expected peak is not appearing. Solution: Check that there is still enough liquid in the vial and if possible add more. If the problem persists the compounds expected might not be present in the sample or not in sufficient quantity to be detected with your current detector.
-
Problem: Odd looking curves are appearing. Solution: Check if all the connections are well made and start the analysis again; check chapter 5.4.
4.5
OLD COLUMN VS NEW COLUMN
After a while a column change was decided because the results were getting less and less replicable, more and more odd results were appearing and a baseline noisier and noisier. Also the peak for component X was made by two peaks and the combination of the two spread over 3 minutes to elute. Last but not least the length of an analysis was too long if multiple analyses were needed. The new column gave very good result with a retention time halved and a peak spread almost divided by three. The baseline noise has also been decreased a lot. A comparison of the results obtained with both columns will now be done. Since the retention time shifts a little bit with the concentration two examples will be used. The data for the old column will be extracted from Figure 12 and for the new column from Figure 13.
25
F igure 12
F igure 13
Component X peak in a S C18R5 250 mm x 4.6 mm x 5 μm
Component X peak in a Zorbax SB-C18 150 mm x 4.6 mm x 3.5 μm
26
The component X has a retention time of 8.3 minutes with the old column and the peak has a baseline width of 4 minutes. Using those values the number of theoretical plates can be calculated: N = 16(tr/wb) ² = 16(8.3/4)² = 68.89 plates for a 25 cm column Using this value we can also find the height equivalent of a theoretical plate which would give a good tool to find the best column: HETP = L/N = 250 mm/68.89 = 3.628 mm With the new column the component X has a retention time of 3.768 minutes and a baseline width of 1.512 minutes. With those values the number of theoretical plates can be found: N = 16(tr/wb) ² = 16(3.769/1.512)² = 99.42 plates for a 250 mm column Therefore its HETP is: HETP = L/N = 150 mm/99.42 = 1.509 mm Another important factor is the repeatability of the results. To study it a sample was run 10 times in each column. The curves obtained are Figure 14 for the old column and Figure 15 for the new one.
F igure 14
Repeatability in a S C18R5 250 mm x 4.6 mm x 5 μm
27
F igure 15
Repeatability in a Zorbax SB-C18 150 mm x 4.6 mm x 3.5 μm
For the old column Table 2 shows the results.
Table 2.
Results for repeatability on the old column
1st peak Average
2nd Peak
Ret time (min) Area Ret time (min) Area 8,35 3656911 9,425 1760927 8,428 3773081 9,485 1574588 8,227 3540559 9,498 1118997 8,298 3760707 9,542 1571478 8,378 3681478 9,58 1257804 8,252 3767281 9,65 1630420 8,243 3635596 9,64 1498459 8,402 3616562 9,675 1178684 8,37 3813182 9,718 1342499 8,283 3642347 9,778 1304181 8,3231
3688770,4
9,5991
1423803,7
Variance
0,005123433 7461870410
0,012593656 45325129927
Standard deviation
0,071578162 86382,11858
0,112221458 212896,9937
28
The variation of the retention time isn’t an issue, it is rather constant. The problem is with the most important part of the HPLC, the area of the peaks. Here the variation is more important. There is more than a 7 % gap between the value between the smallest area and the largest one for the first peak, and almost 40 % for the second peak. For the new column the results in Table 3 are obtained.
Table 3.
Results for repeatability on the new column Ret time (min) Area
Average
4,045 4,05 4,047 4,095 4,05 4,027 4,007 4,003 4,015 3,998
2561904 2686583 2646182 2625886 2596855 2571059 2650159 2545673 2639266 2562415
4,0337
2608598,2
Variance
0,000879789 2253418277
Standard deviation
0,029661235 47470,18303
The variation of the retention time isn’t an issue either even thought it has smaller changes than in the old column. The overall variation in the area is also smaller as the variance proves it too. The gap between the lowest and the highest values is about 5 %. The new column has a HETP more than two times lower which indicates it is a better column. Also the repeatability is much worse with the first column. And even with those improvements, the reduction of the retention time and the baseline were good enough reasons to change the column.
29
5
R ESU LT A N D D ISC USSI O NS
5.1
DETECTION OF COMPONENT Y AND COMPONENT X (IN EACH COLUMN)
Component Y was never detected with neither column due to the absence of a RI detector on the HPLC. Component X was detected with both; the retention time was different though. For the first column the retention time was around 8.5 minutes, for the second one it was 4.5. When detecting component X a shift in the retention time was monitored as the concentration was changing as can be seen in Figure 16 for the new column.
F igure 16
Shifting of the retention time depending on the concentration
30
5.2
TEMPERATURE STABILITY (OLD COLUMN)
After odd results were obtained when measuring samples that were taken out of the refrigerator not long before the analysis the question of the temperature stability was raised. To study it a sample was kept in the refrigerator for one night. In the morning it was taken out of the refrigerator and analyzed soon after. The sample was analyzed 20 times. Each analysis takes 12 minute so the time interval between two analyses is 12 minutes. The curves obtained are represented in Figure 17. It can be seen that as the temperature of the sample increases the area diminishes until the sample reaches room temperature.
F igure 17
Area change due to samples’ temperature changes
31
The data obtained are gathered in Table 4.
Table 4.
Temperature stability for the old column
1st peak
Injection time (min)
Ret time (min) 8,148 8,232 8,227 8,242 8,305 8,335 8,289 8,43 8,42 8,417 8,35 8,428 8,227 8,298 8,378 8,252 8,243 8,402 8,37 8,283
Area Ret time (min) Area 5871501 8,763 5719463 5208503 8,792 4804872 5158893 8,983 3690874 4676535 9,043 3379026 4511184 9,13 2874787 4252267 9,235 2539420 4047071 9,26 2303464 4082707 9,333 1988766 4045565 9,335 1945223 3659705 9,377 1999922 3656911 9,425 1760927 3773081 9,485 1574588 3540559 9,498 1118997 3760707 9,542 1571478 3681478 9,58 1257804 3767281 9,65 1630420 3635596 9,64 1498459 3616562 9,675 1178684 3813182 9,718 1342499 3642347 9,778 1304181
0 13 26 39 52 65 78 91 104 117 175 188 201 214 227 240 253 266 279 292
2nd Peak
The retention time did not change much but the area changes a lot. It was determined that the time it took for a sample to reach the room temperature without a water bath is about 3 hours.
5.3
REPEATIBILITY/REPRODUCIBILITY
There is a difference between repeatability and reproducibility. Repeatability is the ability of a single person or instrument to get multiple measurements with variation smaller than some agreed limit. The data has to be collected in the same place, and within a short period of time. It measures the success rate of an experiment.
32
Reproducibility is the ability of an experiment to be replicated by someone else or on another instrument than the original one. The results have also to be similar (5). The reproducibility couldn’t be calculated but the repeatability was. I considered the repeatability being r = (max reading-min reading/min reading) x 100. Using the data from 0 the repeatability of the old column is: r = (max reading - min reading / min reading) x 100 = ((3813182 - 3540559) / 3540559) x 100 = 7.70 % Using the data from 0 the repeatability of the new column is: r = ((2686583 – 2545673) /2545673) x 100 = 5.53 % The repeatability of the experiments is rather good and even better with the new column. Keep in mind that the repeatability is for the entire HPLC, each part has an influence, the pump, column, detector, etc... Another way to check the repeatability is the Residual Standard Deviation (RSD) can also be calculated using the following formula: RSD =
x 100
Using the data from 0 and 0, the value of the RSD for the old column is: RSDold column = (86382 / 3688770.4) x 100 = 2.34 % The RSD for the new column is: RSDnew column = (47470 / 2608598.2) x 100 = 1.82 % A good RSD is suppose to be lower than 2 % therefore only the new column allowed the HPLC to have a good repeatability.
33
5.4
ODD RESULTS
During the experiments some odd results were obtained. They are considered odd by the fact that they appeared multiple times for no apparent reasons, some of them disappeared for no apparent reason and most of them couldn’t be really explained. The first odd result is the large shift of the retention time based on the concentration of the samples as seen in Figure 16. The second one is the change of the area of the peak based on the sample temperature as shown in Figure 17. It is an odd result because if the temperature should have an influence it would be on the viscosity of the sample, therefore the retention time would change, not the area. Also the column oven is always at 30 °C therefore the temperature of the sample can’t have any influence on the separation detector. The third one is a rather random one, it didn’t appear often. It is, as shown in Figure 18, the appearance of multiple negative peaks. On the chromatogram there are 2 curves, both made from the same sample. The peak for component X appears in the same way for both; the only difference is the baseline.
F igure 18
Appearance of multiple negative peaks
34
The fourth and last odd result is one that appeared very often. It is as shown in Figure 19, a much noisier baseline and a peak going very high. The form of the peak is the same but there is a very high increase and decrease just before and after the peak. It could be the computer changing the sensitivity of the detector but no proof of it was found.
F igure 19
High peak sensibility
35
6
C O N C L USI O N
After a lot of work and multiple tests it was concluded that analyzing component Y just with an UV detector was impossible. On the other hand the study of component X was a success; it is far more accurate than the previous indirect way and results have a fair repeatability. The samples analyzed for an external laboratory gave the result they expected there for the separation of the product must be working as they expected, at least for the component X part. There aren’t many odd results and don’t appear often and all of them aren’t permanent so they aren’t a big concern. Using this HPLC shouldn’t be a problem for next year students and it is a good thing that they will discover it and encounter issues and problems as, in the future when they work. For future work it might be interesting to study: -
The stability of component X stability over time;
-
Do all the study done in this document using an RI detector to study component Y;
-
Compare the results obtained by ion chromatography and HPLC to see which is better.
36
7
SO U R C ES
(1) Chrisitaen Elke, These: “Determination of chlofenvinphos residues in apples by GC-NPD and HPLC-UV” (2) David S. Hage and James D. Carr, “Analytical Chemistry and Quantitative Analysis” (3) Thermo Separation Products UV6000LP Detector Reference manual (4) Thermo Separation Products Spectra SYSTEM & SpectraSERIES GRADIENT PUMPS Reference Manual (5) http://www.engineeredsoftware.com/papers/msa_rr.pdf
