Analytical methods employed on the pharmaceutical industry are mainly chromatographic procedures (e.g., HPLC and GC). With this in mind, the method development tips presented herein will focus on those chromatographic techniques. 

HPLC

HPLC procedures fall into four basic categories—reverse phase, normal phase, ion exchange and gel permeation/gel filtration. Another common type of HPLC technique, ion-pairing, is a reverse phase procedure. Since about 80% of all HPLC work is done using reverse phase chromatography, the following discussion will be on it. In developing a reverse phase HPLC procedure, the following parameters should be considered:

• What analytes are involved?  The goal of most analytical methods is to measure, quantitatively, the amount of Analyte(s) in a sample matrix. For raw materials such as active drug substances, the purity of the raw material is determined. For finished drug products or intermediates such as granulations, the content of active ingredient needs to be determined.

For raw material assays, sample preparation is generally fairly simple and straightforward, since the raw material itself usually represents the sample matrix in its entirety. The analysis of active ingredients in finished products, by contrast, requires a sample preparation that takes the matrix (inactive ingredients) into consideration. It must be determined, either by knowledge of the chemistry or laboratory studies, what interferences, if any, are due to sample matrix and how they can be overcome such that the active ingredient content can be accurately measured.

• What method of detection will be used?  The method of detection should be chosen based upon the chemical structure of the target analyte(s). If the target analyte(s) are good UV absorbers, and there are no other UV-absorbing components in the matrix (placebo), then UV detection should be used. If the target Analyte(s) are UV active, and there are other UV active components in the matrix, then the chemist must be sure that the analyte or analytes of interest, are separated from placebo components, and from each other, if there are more than one active drug substances in the formulation being tested. If the target analytes are not UV active (carbohydrates for example), then other detection methods such as refractive index or electrochemical should be considered. It is recommended that UV detection be attempted first and that other detection methods be used only when absolutely necessary. Refractive index detectors, for example, require steady flow rates and constant temperature throughout the chromatographic system, thus increasing cost (column heater needed) and decreasing efficiency (long column equilibration periods). In addition, refractive index detectors can only be use with isocratic analyses.  If an Analyte is a poor UV absorber such as a carboxylic acid, try using low wavelengths between 190 nm and 210 nm where almost everything has some UV activity. However, when using these low wavelengths, lack of matrix interference is extremely important due to potentially high background noise at low wavelengths.  In addition, low UV wavelength chromatography limits the choice of mobile phase to solvents that have a low UV cutoff such as water and acetonitrile.

• Preliminary screening:  An efficient way to speed up the methods development process is to perform a preliminary UV screening (by UV spectrophotometer) of all product components, both actives and inactives. Doing a UV scan from 360 nm to 190 nm on individual solutions of each products ingredient, allows the chemist to select the best wavelength for the analysis, one that gives the best sensitivity for target analytes (usually the UV maximum) and avoiding those that might be problematical in terms of interference. Additionally, if there is more than one active ingredient in the product to be tested, the separation of which is not possible or extremely difficult to achieve, and one active has a UV absorbtion maxima at a wavelength where the other active has no absorbtion, the analysis can be designed using a dual-channel detector, each channel set at the wavelength best suited for the respective analytes. This technique is known as wavelength masking.

• Selecting a column: Reverse phase columns come in wide variety of sizes and flavors—methyl, butyl, octyl, octadecylsilane, phenyl, amine, cyano, and others. Each has a functional alkyl silane bonded to silica, varying in their retentiveness and order of component elution, depending upon which functionality is involved. When selecting a column, start with an octadecylsilane (C₁₈) column. If this column is too retentive (long retention times) but the desired separation is achieved, try a shorter chain silane such as an octyl (C₈) column, or just use a C₁₈ column that is shorter in length.  If the desired separation can still not be achieved, try a different functionality such a phenyl, amino of cyano for example. Select a column length that results in the shortest run time without sacrificing resolution, peak shape quality and reproducibility.  For basic compounds (many pharmaceuticals active ingredients fall into this category), try using base-deactivated versions of the above columns. This will improve peak shape and resolution.

• What mobile phase will be used? If at all possible, stick to simple mobile phases such a mixtures of water and methanol or water and acetonitrile.

Reverse phase HPLC mobile phases often consist of a mixture of water (water containing a buffer or other additive) and an organic modifier such as acetonitrile, methanol or tetrahydrofuran. If the target analytes are nonpolar (poor water solubility), separation can almost always be achieved with a mixture of water and methanol or a mixture of water and acetonitrile, or a mixture of all three. For more polar materials, the aqueous component of the mobile phase should be buffered at a well-controlled pH, based on the analyte’s pK value, so that the analyte is more organic and less ionic at the pH of the buffer.   At a pH where the analyte if more ionic, it elutes rapidly because its solubility in the mobile phase is more dominant that its attraction to the column packing. Conversely, at a pH where the analyte is more organic and less ionic, its retention is enhanced due to its greater affinity to the column packing at that pH. A carefully selected mobile phase pH can be very effective in achieving separation of components with different pK values. The methods development chemist should choose mobile phase pH based on pK values of analytes coupled with actual laboratory experiments.  Also, be sure to operate within the pH limits of the column.

Occasionally, it is not possible to use a pH where the Analyte becomes retentive. When this occurs, an ion-pairing reagent can be added to the mobile phase. These reagents are non-UV active, long-chain organic compounds containing acidic or basic functionalities, such as an aklysulfonic acid or an akly quaternary ammonium compound. The ion-pairing reagent forms an ion-pair with the analyte (kind of like a weak salt), thus slowing the migration of the analyte through the column, because the long alkyl chain of the ion-pair reagent is highly attracted to the column packing.

            The last and final installment of Methods Development will describe how to design an HPLC separation and will address topics such as sample preparation, instrument parameters and GC methods development.

An analytical method is nothing more than a written list of instructions for performing a laboratory test procedure.  A well-written analytical method can be performed by any chemist of average ability by simply reading the method without any further clarification. The method should list all the reagents and equipment needed and should provide a step-by-step procedure with a detailed explanation of calculations and results units. In addition, if spectra or chromatograms are generated, a sample spectrum or chromatogram should be included as part of the written method.

            In addition to being well-written, first and foremost, an analytical method should work as intended and be practical in terms of functionality and efficiency. It should be robust enough to withstand slight variations in operating parameters, and rugged enough so that it can be used routinely and reliably by different analysts, on different days, in different labs and on different equipment.

The Method Development Process:

            Developing an analytical method, whether simple or complex, is a process based on science but is often practiced as art. The novice chemist is stuck with science, because he or she does not have the experience to develop the science into an art. As chemists gain more and more experience and practice in the development of analytical methods, their intuitiveness will evolve to the point where the science becomes an art form. The normal evolution is as follows:

• Novice chemist—Pure science
• Intermediate chemist—Science plus some art
• Experienced chemist—Art with science intuitively embedded\

In addition to science and art, a certain amount of luck is also involved.

The Billiards Approach:

            A good pool player is always looking ahead to the next shot, scanning the table for position in order to maximize the number of consecutive balls that can be sunk before yielding to the other player. Analytical methods development, figuratively speaking, also requires looking ahead to the next shot.

            The chemist who is developing a method must consider several factors before and during the process: the method’s intended use, who will be using the method, the analytical time cycle and cost.

The Method’s Intended Use:

            Methods should be developed the “fit the bill” so to speak, without losing sight of the intended application. For example, a chemist whose expertise is in HPLC methods development might lean toward using HPLC when designing analytical test methods. However, is a simple titration or other basic procedure will achieve the same end, then the simpler procedure should be selected as the method of choice. Similarly, if a complex analytical procedure is required to perform a particular analysis, then it should be used, rather than simpler techniques that might not offer the needed specificity, accuracy or sensitivity.

            Chemists involved in the methods development process can avoid overkill or under kill by asking, “What am I trying to accomplish?”  For example, in a mixture of carboxylic acids (acetic, formic, malonic and succinic), if one needs to determine the total acidity for neutralization purposes, then the method of choice will be a simple acid-base titration. On the other hand, if the exact composition of the mixture must be known to compute mass balance or yield for example, then a more sophisticated method such as ion-exchange HPLC separation of the carboxylic acids must be selected.

            The method development chemist must always stay focused on the intended application, as this will result in the development of analytical methods that are appropriate and sensible in terms of time and manpower allocation

            .

Who Will Be Using It?:

            An elegant analytical method, one with many technique-intensive steps and novel twists, will be next to useless if its intended end user is untrained or minimally trained in analytical chemistry (i.e., a production operator or Q.C. inspector).

            The methods development chemist must always keep in mind who will be the end user of the method. Suitability for its intended use is not the only criterion for an analytical method. It must also be designed around the skill level of the end user.  A good example is water testing, specifically, the determination of chlorine content in water. An analytical procedure designed for an experienced chemist might involve an amperometric titration of chlorine with standard phenylarsine oxide solution, while a method  designed for plant operator for instance, would probably be limited to using a swimming pool chlorine test kit.

Analytical Time Cycle:

            An analytical method, in addition to its suitability and skill-level components, must have a time cycle that fits its intended application.  A method that requires two (2) days to run because a long sample digestion or extraction procedure for example, is of little use to a Q.C. lab that needs results the same day.  The methods development chemist must determine, up front, how fast a method needs to be prior to its design and development.  The chemist can then design a procedure that can be executed within the required time frame.

Cost:

            Another important consideration in designing and analytical method is its impact on laboratory costs.  Part of the methods development process must include a consideration of cost in terms of the types of reagents selected, the quantities used, the required instrumentation, column selection (for chromatographic procedures), and glassware sizing. For example, when designing an HPLC method, methanol is much less expensive than acetonitrile; therefore, if either will work, then methanol should be the solvent of choice.

            Similarly, when determining how to perform sample and standard preparations, use minimum volumes of solvents and the smallest possible glassware sizes without sacrificing accuracy. For example, if a method calls for a standard solution having a final concentration of 0.2 mg/mL in methanol, one could weigh 100 mg of standard and take it to 100-mL of methanol, resulting in a stock solution containing 1.0 mg/mL, which is then diluted 10.0-mL to 50.0-mL in methanol to give a final concentration of 0.2 mg/mL. The total volume of methanol used per preparation is 150-mL.

            Alternatively, one could prepare a standard solution containing 0.2 mg/mL by weighing 50 mg of standard, taking it to 50.0-mL in methanol, and then dilute 5.0-mL to 25.0-mL.  This procedure uses only 75-mL of methanol per standard preparation. If an analytical balance of sufficient accuracy is used, a 50 mg weighing is perfectly acceptable.  If one were a big spender, the 0.2 mg/mL standard solution could be prepared by weighing 200 mg of standard and taking it to 1000-mL with methanol.  The decision to do a single step preparation versus a weighing plus a dilution must be based upon both cost and accuracy.

            In addition to cost considerations, the selection of reagents must also be driven by availability and safety.  A chemical that is only available from one source, and/or not always available, should not be considered unless there is no alternative. Similarly, extremely hazardous materials such as carcinogens,, mutagens, and acutely or chronically toxic materials, particularly those whose safe handling is a burden to productivity, should be avoided whenever possible.

            The methods development process should also consider the selection of instrumentation and, if the procedure is chromatographic, the selection of columns. If a method can be done by titration, why tie up an HPLC system unnecessarily?  When selecting a chromatography column, select the cheapest one that works. This does not necessarily mean the cheapest column in price, although this is sometimes the case. It refers to the cheapest one in terms of overall time management. If a 150 cm C₁₈ column works as well as a more expensive 250 cm C₁₈ column for a particular separation, then buy the 150 cm column. On the other hand, if one has a choice between an $800.00 ion exclusion column that will perform a separation directly with minimal sample preparation, and a $250.00 C₁₈ column that requires ion-pairing to effect the desired separation, the ion exclusion column might actually be cheaper in terms of overall sample throughput, labor and cost of reagents.

            A detailed and technical treatment of method development in terms of chemical theory is beyond the scope of this article. Our goal is to provide guidance and suggestions for the practicing chemist, and to help jump-start the methods development process, particularly for chemists who are new to the world of methods development and analytical research and development in general.

            Part II of this article will focus on the nuts and bolts of HPLC and GC methods development.

This installment deals with some assorted instrumental analysis techniques that might of use in the analytical laboratory.

GAS CHROMATOGRAPHY:

Check the gas supply pressure regularly; a daily checklist is suggested to monitor primary pressure gauges for helium, air, hydrogen, etc. If more than one instrument is being serviced by a single carrier gas cylinder, make sure that there is enough head pressure to each instruments gas controllers. Using helium as an example, set the cylinder’s secondary pressure gauge (feed to instrument) to 80 psi.  For each additional instrument, add an additional 20 psi.

Clean flame ionization detectors by setting the detector temperature high (300-350°C) and then making several injections of Freon. The flame will convert the Freon to hydrofluoric acid (HF), which will, in turn, effectively clean the detector, removing even siliceous materials. Never turn off the gas flow to columns unless the column temperature is below 60°C (40°C is recommended).  Periodically replace or clean injector liners. Glass liners should be soaked in IN sodium hydroxide for at least one hour, then rinsed with IN sulfuric acid, then DI water, and finally with methanol. Residual methanol can be removed with vacuum or a gentle stream of nitrogen before returning it to the instrument.

Replace the septa daily. When using thermal conductivity detectors, never apply filament current without a flow of carrier gas through the detector. To do so will burn out the filaments, resulting in a very expensive repair bill, not to mention down-time.

When performing manual injections, clean the syringe thoroughly after use, using appropriate solvents to remove any sample, particularly those containing salts or siliceous materials that could freeze up the syringe. A properly cleaned and maintained syringe should last for many months. If a lab has an inordinate number of syringe failures (breakage, bent needles and freeze ups), try issuing a personal syringe to each chemist as opposed to using syringes from a common stock.  It will soon become evident which person or persons is responsible more most of the syringe failures.  Finally, always use two wrenches together when removing or installing columns.

 HIGH-PRESSURE LIQUID CHROMATOGRAPHY (HPLC) TIPS:

During the routine practice of HPLC, always use two wrenches together when removing or installing columns, or when tightening or loosening any other fittings, such as injection valve and detector inlet fittings.

Failure to do so may results in damage that could cost several thousand dollars, particularly if one were to snap a braze weld on tubing going into an expensive detector cell.

If bubbles are generated and all else fails, tie a small knot in the tubing coming out of the detector. The additional back pressure created by the knot may solve the problem. This is above and beyond any backpressure regulators that might be connected to the system.

If an HPLC column exhibits poor resolution or high back pressure, remove the fitting from the head of the column, exposing the packing. Chances are the packing will look dark or even black. Using a small spatula, gently scrape packing from the surface until it is pure white in color. Replace the fittings and reinstall the column. The column performance should be markedly improved. If this doesn't work, try installing the column backwards and let it run until a minimum pressure has stabilized. The reverse the column and run it normally.

Precede each analytical run with at least six conditioning injections. These injections, which do not count as part of the analytical run, will help ensure that system suitability will be quickly attained upon commencing the analytical run.  System suitability should not be initiated until conditioning peak responses have reached equilibrium.

FLAME ATOMIC ABSORPTION TIPS (Flame AA)

Never try to light a nitrous oxide-acetylene flame when using an air-acetylene burner head. Most instruments have safety interlocks to prevent this, but should one succeed, the resulting explosion will be impressive. Do not use an acetylene tank whose pressure is below 75 psi. Acetone is used as a solvent for acetylene. If a tank is used at too low a pressure, acetone may get into and destroy the instrument's gas control boxes. This repair can run in excess of $4,000.00

Periodically aspirate 10 percent nitric acid (HNO3) into the instrument. This ensures good cleanliness and optimum performance. Finally, reserve glassware for AA work just for AA work and use appropriate reagents.  Use only metals-grade acids and deionized water. All glassware used for AA work should be soaked in 10% metals-grade nitric acid, then thoroughly rinsed with deionized water, then methanol and then air-dried. Oven drying can activate the surface of glass vessels, causing losses of analyte by absorption.  Standard stocks should be stored in linear polyethylene containers and working standards should be prepared fresh daily

TRICKY WATER DETERMINATIONS

Two tricky water determinations that come to mind are Karl Fischer titrations of bases such as amines and the determination of low levels of water in matrices that react with the Karl Fischer reagent.

Karl Fischer Base Titration

When titrating bases such as amines with the Karl Fischer reagent, use a mixture of 10 percent acetic acid in methanol as the titration medium. The acid ties up the base so that it does not interfere with the determination of water in the sample.

Low Level Water Determinations Not Well Suited for Karl Fischer Titration

An elegant method for the determination of low-level moisture in materials that react with the Karl Fischer reagent is as follows:

·        Place the sample into a suitable round-bottom boiling flask.

·        Add dry toluene (toluene that has been stored over metallic sodium) to the sample.

·        Connect the flask to a suitable condenser and receiving vessel such as an Erlenmeyer flask. The receiving vessel should be almost totally immersed in a dry ice-acetone bath.

·        Distil the toluene into the collection vessel

·        Seal the receiving vessel, allow it to warm and titrate the collected toluene with Karl Fisher reagent.

·        Run a blank consisting of distilled toluene without sample.

The water in the sample is azeotroped with toluene and collected as ice crystals in the cold receiver. The small amount of water is then titrated with Karl Fisher reagent.

            The last installment in this series will give a potpourri of assorted additional laboratory tips that might be useful in day to day work.

The next few installments of articles will present a collection of technique tips—provided to help make laboratory life somewhat easier and more fun. They are presented in no particular order and cover a wide variety of laboratory situations.

Back Titrations

             When performing back titrations (i.e., adding an excess of one titrant to a sample and then titrating the unreacted first titrant with a second titrant), there are two possible approaches. The first is to use two titrants, each of which has been standardized; the second is to use two titrants, only one of which has been standardized. For example:

           25.0 mL of 0.5N potassium hydroxide (KOH) is added to a sample. The sample is refluxed for a period of time, cooled to room temperature, and the unreacted KOH is titrated with 0.5N HCl. The milliequivalents of KOH consumed could be calculated as follows:

(mL KOH x Normality KOH) -(mL HCl x Normality HCl)

However, if a blank, consisting of 25.0 mL of 0.5N KOH was titrated with 0.5N HCl in addition to the sample titration, then only the HCl needs to be standardized. In this case, the expression for calculating the milliequivalents of KOH consumed is as follows:

                        (mL HCl Blank -mL HCl Sample) x Normality HCl

Back titrations using only one standardized solution eliminate the need for standardizing a second solution. While one must titrate a blank, a single blank can be used to calculate multiple samples. Overall, the single, standard solution-blank titration approach is faster and less prone to error .

Consistent Endpoints

            When performing manual titrations, an easy way to stop at the correct endpoint consistently is to create an endpoint color reference. Before beginning any titration involving an indicator, prepare a solution consisting of each reagent to be used in the titration, but omitting the sample ( e.g., water plus indicator for example). This color reference blank should be the same volume and contained in the same type of vessel as that used for titrating the sample. For example, if a sample of about 100 mL total volume is titrated in a 250 mL Erlenmeyer flask, the color reference blank should be about 100 mL contained in a 250 mL Erlenmeyer flask. Once the blank is prepared, titrate dropwise ( one drop at a time) until the indicator changes color. Set the color reference aside. When titrating each sample, stop at that same color. This technique is particularly useful for indicators that have transition colors, such as methyl orange and crystal violet.

pH Electrode Care

    When calibrating pH meters with standard buffer solutions, the inability to apply a slope correction to a buffer usually indicates that the electrode needs to be reconditioned. Other symptoms that reconditioning is needed include drifting pH/millivolts or sluggish response. Assuming that the electrode is not damaged, the following seven-step procedure will almost always bring a pH electrode back into good operating condition:

1. Empty the electrolyte filling solution.

2. Rinse the inside of the electrode with several small volumes of hot tap water to dissolve and remove any crystallized salts that may have formed. An eye dropper is handy for this purpose.

3. Rinse the inside of the electrode with several volumes of deionized water, followed by filling solution.

4. Fill the electrode to the proper level with electrolyte filling solution ( e.g., saturated potassium chloride [KC1] for example).

5. Hold the bulb of the electrode under hot tap water for several seconds, making sure that any film or salts are removed. If a film still remains, rinse the bulb with a suitable solvent, such as methanol to remove the film.

6. Immerse the electrode in 1N HCl for about 15 minutes. .

7. Rinse the electrode with deionized water and store it in either buffer solution or tap water. Never store an electrode in deionized or distilled water.

The electrode is now ready for use and should perform like new.

Preparing Standard Caustic Solutions That Last

 Aqueous NaOH or KOH Solutions

                   When preparing standard solutions of aqueous NaOH or KOH, prepare the solution in a beaker. Bring it to a boil and continue boiling for about two min. Remove the beaker from the heat source and immediately cover it with a piece of clear wrap, securing it to the beaker with a large rubber band. After the solution has cooled to room temperature, filter through glass wool into the final container prior to standardization. Prepared in this way, the NaOH or KOH solution will maintain a constant normality for a longer period of time. Boiling removes any carbon dioxide from the solution, while the plastic seal prevents reabsorption of CO2 during cooling. Minimizing carbon dioxide content reduces the formation of carbonates, which lowers the normality of caustic solutions over time.

Alcoholic KOH or NAOH

     Alcoholic KOH or NaOH solutions that seem to last forever and never turn yellow can be prepared by using alcohol (ethanol) that has been distilled over metallic zinc (Zn). Simply placing alcohol in a boiling flask with granulated Zn and distilling into another container will produce aldehyde-free alcohol. This aldehyde-free alcohol, when used to prepare alcoholic KOH or alcoholic NaOH, will produce solutions that are water-white and that will stay water white for long periods of time.

TLC Plates That Work Great

       Before using a TLC plate, put it in a developing chamber with the developing solvent to be used for the analysis and let the solvent migrate to the top of the plate. Remove the plate from the developing chamber, allow to air dry, place it in an oven at 105°C for 1 hour, and let it cool in a desiccator. Store the plate in the desiccator until ready to use. The plate will perform excellently, enhancing tight-round spots, maximizing chromatographic resolution, and minimizing interference from impurities in the silica base.

            The next installment will deal with non-aqueous titrations and chromatography.

For as long as anyone can remember, the traditional way of teaching experimentation was to use the One Factor at a Time (OFAT) method. For example, if there are three variables (time, temperature and shaking speed for example), one would keep time and temperature constant while varying speed to study its effect, then keeping speed and temperature constant while varying time, and so on.

The problem with this approach is that there is no evaluation of the interactions between variables. It is not possible to determine whether an effect is due to a single variable or due to the interaction between two or more variables (confounding or aliasing).

If one needs to look at multiple variables simultaneously to determine which single variables and which interactions have statistically significant effects, then one should use statistically designed experiments (DOE) as the method of choice.

Click on the  SIX SIGMA METHOD VALIDATION tab.

This installment presents good laboratory techniques for performing potentiometric titrations measurements and for performing volumetric Karl Fisher titrations.

Karl Fisher Titrations (Volumetric):

Accurate measurement of water content ranging from percent down to parts per million can be readily achieved using proper Karl Fisher titration techniques. The following equipment and reagents are needed:

·         Karl Fisher titrator capable of amperometric endpoint detection.

·         Karl Fisher reagent, preferably one that is pyridine free, having titre of about 5 mg water per milliliter of reagent.

·         Suitable solvent (usually anhydrous methanol).

·         DI water.

The first step in performing a Karl Fisher titration is to standardize the Karl Fisher reagent just prior to use, i.e., determine the titre of the reagent in milligrams of water per milliliter of reagent.  This is done as follows:

• Follow the manufacturer’s instructions for operation of the Karl Fisher titrator.
• Blank about 50-mL of anhydrous methanol with Karl Fisher reagent.  With most titrators, Karl Fisher reagent will be added until an acceptable level of drift is attained.
• Place a small dropping bottle or vial containing about 1-mL of DI water, equipped with an eyedropper or similar dispensing device, onto the pan of an analytical (5-place) balance.  A microliter syringe containing about 30 uL of water may also be used.
• Tare the balance to ZERO DISPLAY.
• Dispense one drop of DI water into the blanked methanol, or in the case of a syringe, dispense the entire volume into the titration vessel through a septum.
• Titrate with Karl Fisher reagent to the same endpoint (amperage or drift criteria) as that used to blank the methanol. Be sure to minimize exposure of the titrant to air.   A trap containing molecular sieves, attached to the Karl Fisher reagent bottle, is a good way to do this.
• Reweigh the dropping bottle or vial or syringe and the record the weight of water used (absolute value of the negative balance display).  NOTE: This technique is called weighing by difference.
• Perform the water titration in triplicate.
• Calculate the water factor (F) as follows:

milligrams of water         =  F (mg/mL)

mL Karl Fisher Reagent    

The % RSD for the three factor determinations (personal preference) should equal or less than 1.0%.

            For titration of samples:

·         Follow the manufacturer’s instructions for operation of the Karl Fisher titrator.

·         Blank about 50-mL of anhydrous methanol.

·         For liquid samples, weigh by difference as described above, or dispense by volume, depending upon the expected water content of the sample. Direct pipeting can be used for either water by volume or water by weight. For percent by weight, use the density of the sample to calculate weight from volume.  Sample size guidelines for a titration of approximately 5-mL are as follows:

Percent Water             Sample Weight (grams)

10–20%                       0.1

5–10%                         0.2

1-5%                            0.5–2.5

<1%                             2–10

NOTE:            For sample having very low water contents, a diluted version of Karl Fisher reagent can be used.   Diluents are commercially available as are Karl Fisher reagents with very low titres.

·         Dispense the sample directly into the Karl Fisher titration vessel containing blanked methanol.

·         For solid samples, use a sample boat (usually supplied with the Karl Fisher Apparatus).

·         Tare the weighing boat, plus an amount of sample containing about 20–30 mg of water, to ZERO DISPLAY.

·         Slide the sample from the weighing boat directly into the Karl Fisher titration vessel containing blanked methanol.

·         Titrate the sample with Karl Fisher reagent to the same endpoint (amperage or drift criteria) as that used to blank the methanol.

·         Calculate the water content of the sample as follows:

(mL titrant) x F x 100  =         % water, w/w

                                   mg of sample

·         For determination of water by volume, samples may be pipetted or dispensed by syringe into blanked methanol and then titrated. In such cases, the water content of the sample is as follows:

(mL titrant) x F x 100 =          %water, w/v

    mL of sample

            The above procedure is that most commonly used, however, there are a number of variation. For example:

·         A mixture of 10% acetic acid in methanol is used for determination of water in nitrogen containing compounds such as amines.

·         Alternate solvents can be used such as anhydrous toluene stored over sodium metal. This is useful if the sample must be dissolved prior to running the Karl Fisher titration. Since toluene over sodium metal is absolutely anhydrous, it contributes no blank.

Potentiometric Titrations:

Potentiometric titrations are titrations based upon measurement of change in millivolts or pH versus volume. The titrator determines the endpoint or endpoints automatically thereby eliminating the need for subjective observations of indicator color changes.  Titrations performed by automatic titrators have a wide range of applications such as acid-base titrations, dead-stop titrations, argentometric titrations, Karl Fisher titrations and complexometric titrations. Titrations can be performed in a variety of solvent systems ranging from water or methanol to more exotic solvents such as liquid ammonia and liquid hydrogen cyanide.  The following basic equipment and reagents are needed:

·         Automatic titrator capable of measuring and recording (on chart paper) change in electrode response versus volume, and capable of automatic endpoint(s) determination.

·         Suitable electrode system (dependent upon application).

·         Suitable solvent  (dependent upon application).

·         Suitable standardized titrant.

·         Appropriate additives, such a nitric acid for argentometric titrations.

General procedure is as follows:

• Prepare the sample as per its analytical monograph. The sample should be contained in a beaker of sufficient size so that the electrode or electrodes can be properly immersed ands there is adequate volume available for samples plus titrant.
• Setup the titrator with the proper titrant.
• Flush the lines and burette several times with titrant.
• Make sure that there are no bubbles in the lines or in the burette.
• Immerse the electrode(s) and burette tip into the sample solution and begin stirring. The electrode(s) should be immersed as much a possible without hitting the stirring, as such immersion will help minimize electrical noise and maximize shielding.

NOTE:            Do not use any plastic objects on the stir plate (such as a cat-food can cover or coffee can cover, as the static charge from these items may make it impossible to titrate anything.

• Begin titrating, slowing down the rate of addition as the endpoint is approached. NOTE: Most automatic potentiometric titrators have a built-in algorithm that provides for reduction in the rate of addition as the endpoint is neared.
• After the titration has been completed, remove the electrode(s) and burette tip from the sample solution. Rinse each with DI water and store as per manufacturer’s recommendations.
• Refill the burette.

This series of articles is meant to act as an instructional program of basic laboratory techniques that hopefully, will contribute to accuracy and efficiency in your analytical laboratory. This installment will deal with weighing and pipeting.

            Please note that these techniques and guidelines will usually but not always correspond to the USP method of doing things, but then again, if the USP way was the only way, we would probably have to kill ourselves. However, in the pharmaceutical industry, USP procedures generally prevail and should be followed whenever possible. The techniques presented herein are based on sound laboratory practices and are meant to provide general guidance—not to supercede or replace any approved standard operating procedures.

Weighing with Analytical, Semi-Micro and Top-Loading Balances:

1.                  Make sure the balance pan, chamber and surrounding area are clean and dry.

2.                  Tare the balance so that the digital readout is as follows:

·         0.00 + 0.01 grams for a top-loading balance.

·         0.0000 + 0.0001 grams (0.1 mg) for an analytical balance.

·         0.00000 + 0.00001 grams (0.01 mg) for a semi-micro (5-place) balance.

3.                  Select the proper balance to use according to the following criteria:

·         For weighings equal to or greater than 5 grams, use a top-loading balance.

·         For weighings between 100 mg and 5 grams, use an analytical balance (4-place).

·         For weighings between 10 and 100 mg, use a semi-micro balance (5-place).

4.                  Accurately weigh the sample according to the following guidelines:

Powders & granulations that do not easily gain or lose weight (water or solvents)-

·         Tare of piece of glassine weighing paper on the balance pan (place a piece of glassine weighing paper on the balance pan and press TARE). The balance display should read zero.

·         Distribute sample evenly onto the weighing paper, keeping the sample as centered as possible on the balance pan.

·         Record the weight.

Solids that readily gain or lose weight (moisture or solvent) upon standing-

·         Tare a covered weighing bottle on the balance pan.

·         Remove the cover, placing it on the balance pan.

·         Add sample to the weighing bottle until the approximate correct amount is indicated on the balance display.

·         Without delay, place the cover back onto the weighing bottle and record the actual sample weight.

Liquid samples (by difference)-

·         Place a quantity of liquid in excess of that needed for the analysis into a container, such as a dropping bottle.

·         Place the dropping bottle containing the sample onto the balance pan and tare the balance display to zero.

·         Transfer the quantity of sample needed for the analysis  from the weighing bottle into another vessel such as a beaker or volumetric flask.

·         Reweigh the sample bottle. The sample weight is the absolute value of the negative number displayed. For example, if the balance displays -0.1033 grams, then the weight of sample is 0.1033 grams. HINT: Assume 20 drops per milliliter when transferring liquid samples.

Liquid samples (direct weighing)-

·         To a suitable stoppered vessel such as a volumetric flask, add an amount of trapping solvent (solvent in which the sample is soluble) equivalent to about 25% of the vessel’s volume. For example, if a 100-mL volumetric flask was used, then 25-mL of trapping solvent would be used. 

·         Place the stoppered flask containing the trapping solvent directly onto the balance pan and tare the display to zero.

·         Remove the flask from the balance pan and set it on a clean dry surface. Remove the stopper and set it aside on the same clean, dry surface.

·         Add sample to the flask, by volume, estimating the volume needed to achieve the desired weight by using weight and density to calculate volume.

·         Without delay, restopper the flask, swirl to mix, place the stoppered flask back onto the balance pan and record the sample weight.

NOTE:            An example of such an application is preparing alcohol standards using water as a trapping solvent.

Pipeting (To Deliver Pipets):

1.                  Make sure that the pipet is clean and dry.

2.                  Place the tip of the pipet below the surface of the liquid. Using a pipeting bulb or other suitable device, suck liquid into the pipet top about one inch above the mark.

3.                  Allow the pipet to drain completely to a waste container.

4.                  If any drops of liquid cling to the inner walls of the pipet, discard that pipet, select a different clean pipet and start over from Step 1.  NOTE:  A pipet that has drained properly will look dry inside. Allow a pipet to drain by gravity, never blow out the liquid.

5.                  If the pipet drains properly, with no sign of liquid remaining on its inner walls, then suck up liquid a second time to just above the mark and let drain to waste container.

6.                  Once again (third time) suck up liquid to just above the mark.

7.                  Using an absorbent tissue or other suitable material, wipe the outside of the pipet completely dry, making sure that no liquid remains.

8.                  Slowly drop the liquid level in the pipet into a waste container until the liquid meniscus is level with the mark.

9.                  Wipe off the pipet tip one more time.

10.              Allow the contents of the pipet to drain by gravity into the sample receiving vessel. The drop remaining in the tip is supposed to be there, so don’t blow it out.

11.              For nonviscous samples, allow the pipet to drain for about 15 seconds after the liquid has been dispensed.

12.              For viscous samples such as syrups, allow the pipet to drain for about 45 seconds after the liquid has been dispensed.

13.              Do not blow out the pipette or rinse the inner walls. Volumetric pipet designed to deliver (TD), are calibrated to deliver by spontaneous drainage.

Pipeting (To Contain Pipets):

There are occasions when a liquid is such that it simply will not drain without leaving residual material on the walls of a pipet.  For such samples, one must use a To Contain (TC) pipet. The TC pipet looks just like a TD pipet except that it usually has two marks.

The lower mark is for the TC mode and the upper mark is for the TD mode.  In the TC mode, sample is drawn exactly up to the point where the liquid meniscus is level with the lower mark. The sample as then allowed to drain, and the inner walls of the pipet are rinsed with a solvent such as DI water to assure quantitative transfer of sample.  The lower mark makes the drop in the tip unnecessary and thus the pipet must be rinsed to assure complete transfer of the sample contained therein.

The next installment in this series will feature techniques for performing sample handling and quantitative transfers plus dilutions and manual titrations.

One of the most difficult tasks in the pharmaceutical industry is how to justify the relaxing of a specification, particularly for holders of NDAs or ANDAs.  Tightening of specifications is easy; just do it and include the change in the annual report.  If, on the other hand, one needs to relax a specification, there must be good scientific rationale for such a change that includes not only process capability, statistical process control data and current industry practices, but also takes into account the impact of the change on consumer safety and drug efficacy.

A case study involves a former client who is an NDA holder for a popular and successful OTC product. The dose uniformity specification on their principal product was 97.0 to 103.0 percent of label claim, versus an industry norm of 85 to 115 percent of label claim [United States Pharmacopeia (USP), General Tests, Uniformity of Dosage Units <905>].

Unlike assay, where an out of specification result can be overcome if the OOS investigation reveals an explainable cause for the failing result, uniformity of dosage units is a tiered test, i.e., there are several acceptance levels, any of which, if exceeded, constitutes failure of a batch without recourse. For example, the USP states that for tablets,

“the requirements for dosage uniformity are met if the amount of the active ingredient in each of the 10 dosage units as determined from the Weight Variation or the Content Uniformity method lies within the range of 85.0% to 115.0% of the label claim and the Relative standard deviation is less than or equal to 6.0%. If 1 unit if outside the range of 85.0% to 115.0% of label claim and no unit is outside the range of 75.0% to 125.0%, or if the Relative standard deviation is greater than 6.0%, or if both conditions prevail, test 20 additional units.”

For the USP test, Level 1 is 85.0% to 115.0% of label claim and a 6.0% relative standard deviation, and Level II is 75.0% to 125.0% of label claim.  If during the testing of the initial 10 tablets, one or more tablets falls outside the range of 75.0% to 125.0% of label claim, then the batch fails. No further testing is permitted.  In the case of the subject client, Level I was 97.0% to 103.0% of label claim, with a relative standard deviation limit of 3.0%, and Level II was 93.0% to 107.0% of label claim.  This highly restrictive specification was based upon three validation batches submitted for NDA approval.  The result was that between five and seven batches per year were rejected for dose uniformity, each with a retail value about one million dollars.

The client specification was clearly out of line with industry practice. In addition, this overly tight specification did not offer any advantage in terms of drug safety or efficacy.  It was therefore decided to aggressively pursue an attempt to relax the dose uniformity specification.

From a regulatory standpoint, according to 21 CFR 314.70, “Supplements and other changes to an approved application”,section (c)(2)(iii), “Relaxation of an acceptance criterion or deletion of a test to comply with an official compendium that is consistent with FDA statutory and regulatory requirements” can be described in the annual report. This section incidentally applies equally to compendial and noncompendial products.   The client’s regulatory group was fairly conservative and opted to go with a preapproval supplement (PAS) for the specification change.

We looked at the last 140 consecutive batches of product produced and entered all of the does uniformity data into a spreadsheet.  Since ten tablets are tested per batch, we wound up with a data set of 1140 individually tested tablets. Using statistical analysis software, process capability data was generated using the current specification limits.

FIGURE 1 – CAPABILITY ANALYSIS FOR 97.0% TO 103.0%

At 97.0% to 103.0%, the process spread is much wider than the specification limits. The process is not capable and within a 97.0% to 103.0% range, more than 10% of tablets would fail.  An analogy that is useful for understanding process capability is to consider the process a car, and specifications as a garage.  In Figure 1, the car is wider that the garage and cannot fit; is not capable of fitting.  The process capability index (Pp), defined as Upper Specification Limit minus Lower Specification Limit divided by 6 times the process standard deviation (USL-LSL/6 Sigma), should be at least 1.5.  In this case, Pp = 0.58.  Looking at Figure 2:

FIGURE 2 – CAPABILITY ANALYSIS FOR 95.0% TO 105.0%

AT 95.0% to 105.0%, the process is still not capable, with a potential failure rate of 0.7% (14000 tablets in each 2 million tablet batch). The car still won’t fit in the garage. At 93.0% to 107.0%, the widest limits for Level II:

FIGURE 3 – CAPABILITY ANALYSIS FOR 93.0% TO 107.0%

In this case the car is capable of fitting in the garage, but if it is not centered, one might hit either the right or left wall.  Here, the degree of centering is important. The value Ppk, which is the process capability adjusted for centering, should be 1.5 or greater. When Pp = Ppk, the process is perfectly centered.  In this case, Ppk = 1.23. The result is that there is still significant failure rate of 877 ppm. For Six Sigma performance, quality levels must be equal to or less than 3.4 defects per million opportunities, in this case, 6.8 or fewer tablets per batch. Incidentally, Ppk =, the specification limit closest the mean, minus the mean, divided by 3 sigma. In Figure 3, Ppk = (99.3939-93)/(3 x 1.73) = 1.23.

NOTE:   Pp and Ppk are used instead of Cp and Cpk in order to more accurately express overall capability (long-term). Calculations for Pp and Cp are identical as are calculations for Ppk and Cpk.

It was calculated, that for this process to yield six sigma performance for dose uniformity, that the Level I specification would have to be 88.0% to 112.0% of label claim. Looking at Figure 4 below:

FIGURE 4 – CAPABILITY ANALYSIS FOR 88.0% TO 112.0%

With a process capability (Pp) of 2.31 and a Ppk of 2.20, the process is not only capable, but also has sufficient room for drift off center without producing any out of specification tablets.

For this product, an 88.0% to 112.0% dose uniformity specification was appropriate. The proposed specification, still tighter than industry standard, but which adequately suited the client’s business needs, was approved by FDA, and with a Level II of 80.0% to 120.0% of label claim.  Relaxed relative standard deviation limits were also approved by FDA.

The annual review is a good time to look at all process data and to evaluate the need for change.  FDA actually encourages change based upon actual process performance.  The key to being successful with attempts at change is to have good scientific rationale supported by statistical analysis of data.

The next installment will demonstrate, step by step, how to get approval for extension of product shelf life.