DE102004021060A1 - Pharmaceutical substance steroid production installation cross contamination reduction procedure uses special investigation techniques - Google Patents

Abstract

A pharmaceutical substance steroid production installation cross contamination reduction procedure allows the detection of steroidal substances at very low concentrations with analytically negligible concentrations of related products using special investigation techniques.

Description

The The invention relates to a method for reducing cross-contamination between pharmaceutical agents and impurities in production facilities.

The Production of pharmaceutical active substances or medicaments takes place preferably in multi-purpose plants, that is, in one and the same plant are successively substances or products with different toxicological and pharmacological properties.

to Avoidance of cross-contamination when changing from a product to a secondary product is efficient to clean and proof to ensure that the residues remaining in the production plant are not pose a potential risk. Evidence of the efficiency of cleaning processes is therefore an essential aspect of quality assurance in the pharmaceutical Production and a necessary contribution to increasing drug safety.

In the Food and Drug Administration's Inspection Guide on Cleaning Validation "are the basic requirements in the planning, implementation and Evaluation of cleaning validation included, so the creation a validation plan with justification of the chosen Cleaning method and acceptance criteria, the detailed description the validation procedure, the documentation of the applied analytical procedure, the description the sampling and the evaluation of the results.

The Importance of the ever-safe mastery of the cleaning processes In addition to the manufacturing process has been steadily increasing since the late 80s.

It is no longer sufficient at the end of a manufacturing process, in particular of medicines great pharmacological activity, and cleaning the equipment to make the statement that "no residue in the plant was detected ". This result is only in conjunction with validated cleaning methods and sampling methods and the sensitivity of the Analytical method to study the solutions after cleaning acceptable.

The The requirement to validate cleaning processes extends from the classical production of pharmaceutical active substances and Pharmaceuticals to newly developed biotechnology products. When Validation is a documented program called that certain process, method or system with a high Degree of safety delivers results previously obtained from the manufacturer specified acceptance criteria.

cleaning validation is the documented proof that cleaning a plant leads to consistent and reproducible results by: from critical and representative Parts of the system are cleaned after cleaning.

The execution of purification processes and the proof of their efficiency is in of the literature. Some examples are cited below.

For meclizine hydrochloride (USP26), as an oral drug in various dosages Treatment of musculoskeletal disorders in the market, a cleaning process and analytical method has been developed with which traces of this substance could be reliably detected. The stability of the active ingredient on stainless steel surfaces, in cleaning materials and in extraction solutions was examined.

Clarithromycin (USP26), a new semisynthetic broad spectrum antibiotic, has been developed to provide a simple and reproducible analytical method for the determination of lowest concentrations on different material surfaces. Rotsch and co-workers were able to reliably detect clarithromycin in quantities of 3 μg per 100 cm ² with this HPLC method coupled with extremely sensitive electrochemical detection.

For highly active substances, the proof of efficient purification, that is, the safe detection of these substances in the nanogram range is essential. The determination of the platinum complex compound cisplatin (cis-diamminedichloroplatinum (II)) (USP26) in the concentration range from 20 to 200 ng / ml by direct analysis of the solutions and to 0.5 ng / ml with appropriate sample concentration has been developed by Raghavan and coworkers. The need for cisplatin and other platinum compounds in the ppb range or to be able to reliably and reproducibly detect them, results from the very small but highly effective doses of these compounds, which are used as cytostatic agents in chemotherapy, for inhibiting the development and proliferation of tumor cells.

bisnafide (Drugs Fut 1999, 24 (3)), a bisnaphthalimide class compound, is also used in cancer therapy. Validated cleaning methods were and are in the development and production of this extremely cytotoxic substances are an inseparable component. The task for Segretario et al. consisted in the development of surface-specific cleaning procedures of glass and stainless steel materials that have come into contact with Bisnafid are, with the aim, the surface contamination to reduce to the trace range. In addition, sample name procedures were used and analytics for the detection of bisnafid in extremely low concentration ranges developed.

For losoxantrone (Drugs Fut 1998, 23 (8)), a compound of the anthrapyrazole class and used to treat breast tumors, Shea et al. developed a simple cleaning method that enables the detection of this highly active substance in areas below 1 ng / cm ² equipment surface. With the aid of this process, the contamination of losoxantrone in undesired concentration ranges can be reliably excluded for successive production processes.

Highly active substances and pharmaceuticals, such as penicillins, are increasingly being produced in plants intended exclusively for this product. This legally required "dedicated equipment" is an expression of the high level of attention given to drug safety and quality.The maximum permissible concentration of residues, in particular of highly effective carcinogenic or non-carcinogenic substances in medicinal products, has been determined by Conine et al Determining the acceptance limits of impurities in active substances and finished medicinal products depends on whether single use, long-term use or lifetime use of the drug is available Fentanyl (USP26), alfentanyt (USP26) and sufentanil (USP26) , all piperidine derivatives, are highly potent centrally acting analgesics.Lambropoulos, J. et al developed an analytical method for the detection of these highly potent drugs in concentration ranges of 0.01 μg / ml with the aim of purifying processes in the production of these subs to control and control dancing. ⁱ stainless steel surfaces contaminated with the above-described substances, by a suitable method, in which by means of textile materials, the impurities have been removed from the surface, more than 80% of the respective active ingredients could be found again.

The method used to determine the residue in the pharmaceutical Plant should be sufficiently sensitive and beyond deliver validated results. Mirza et al. studied the stability of methenamine hippurate (USP26), an agent used to treat chronic urinary tract infections, on stainless steel surfaces and showed that within only a few days only about 10% substance of the contaminated surface were found again. As causes were both instability of the drug as well as penetration of the solutions into the metal surfaces discussed. In addition to the usual Chromatographic and spectroscopic methods are becoming more and more more time-saving and less expensive Analytical methods, e.g. the 'total Organic Carbon 'method (TOC) in residue analysis applied. Extensive research with traditional products such as antibiotics and glucocorticoids, led to the establishment of this Analytical method in the pharmaceutical industry. TOC does however also decisive disadvantages compared to classical methods, which Method is not specific and only suitable for water-soluble samples. to Determination of acetylsalicylic residues Anlagenoberfächen A TOC method was developed, with a minimum of time acetylsalicylic acid (USP26) can be reliably detected in the ppm range. Holmes and Vanderwielen led Studies on the recovery of this substance on aluminum, Stainless steel and Plexiglas surfaces through, recovering from 25% for aluminum to 75% for Plexiglas were determined. It has been shown that for less potent agents like acetylsalicylic acid also the visual rating of equipment surfaces after cleaning to acceptable results leads.

In In the 1990s, the lawsuit became Food and Drug Administration / USA known to Barr Laboratories Inc./USA. Main charge was in addition to the release of products that are not the specification corresponded, incomplete and fake Documentations, missing validations of test methods and manufacturing processes the use of non-validated cleaning methods.

• – Verwendete Reagenzien zur Reinigung wurden nicht identifiziert.
• – Es erfolgte keine Prüfung auf Reagenzienrückstände nach Reinigung.
• – Studien zur Reinigungsvalidierung von pharmazeutischen Anlagen waren lückenhaft oder fehlten generell.
• – Die Reproduzierbarkeit der Reinigungsprozesse wurde bei Barr Laboratories nicht untersucht.

Barr was accused of the following points concerning her cleaning processes:

• - Used reagents for cleaning have not been identified.
• - There was no check for reagent residues after cleaning.
• - Studies on the purification validation of pharmaceutical plants were patchy or lacking in general.
• - The reproducibility of the cleaning processes has not been studied at Barr Laboratories.

In the verdict is called it: "It proof must be provided that the cleaning methods used are effective. Companies are allowed do not wait for contamination or product mixtures and only then recognize that their used methods are inappropriate. "Task of the present The invention is a method for reducing cross-contamination between pharmaceutical agents and impurities in production facilities to provide.

According to the invention Task by a method for reducing cross-contamination between active pharmaceutical ingredients and impurities in production facilities, especially for steroidal agents, whereby by appropriate investigation methods steroidal agents can be detected in a range up to 10-100 ng / ml and at the same time realizes that at a steroidal drug concentration from 0.1 mg / ml to 0.01 mg / ml of analytically negligible by-products.

About the Furthermore, it must be shown that by the cleaning conditions in the drug manufacturing as well as by the transport and storage conditions, the quality of the active ingredients not affected.

For the security suitability, i. the safety and efficacy of a medicinal product for the intended application is the responsibility of the pharmaceutical manufacturer, the goal with a reliable functioning quality assurance system that achieves all Areas that the quality drug or drug.

Next In the manufacturing process, processes such as e.g. the cleaning after production, storage of products and transport quality for a Product.

Of the Influence of plant material, such as glass, stainless steel and plastics, on the quality Pharmaceutical agents have been identified through recovery studies studied the goal of favoring special materials and solvents or exclude from the outset.

outgoing from the task the work includes the determination of Recovery rates of selected Active substances using different sampling methods, the be tested for their efficiency; as a result of this Investigation should be justified be which method for the substance classes in question should be used in practice.

Attachments for the production of pharmaceutical active ingredients on pilot scale and production scale, such as reactors and processing vessels exist mainly made of alloyed stainless steel. The influence of the surface texture, i. of the Roughness of stainless steel materials, on the adsorption tendency of the active substances and consequently the quality of the products should be investigated become.

Out The totality of examinations should be an optimal course of action in the purification of steroidal agents with regard to plant material, used solvent and sampling procedures are developed.

Within This work was a suitable analytical method for quantitative analysis Proof of selected develop steroidal agents in the lower ppm range.

Of the Proof of suitability of developed analytical methods, such as Reproducibility, precision, linearity, Detection and determination limit, accuracy, stability of the test solution and Robustness should be provided.

Also the possibility the contamination of pharmaceutical active substances by particles the reactor and container materials of the manufacturing and storage process were to be investigated.

in the Framework of the implementation of a quality assurance concept is the Goal of this work, understanding for the Deepen manufacturing, cleaning and storage processes, so be able to control the qualitative influencing parameters in a targeted manner, and connected with it to ensure a high quality of the active ingredients.

cleaning process

It is a fundamental requirement of the regulations of 'good manufacturing practice' (GMP) that the implementation of Cleaning procedure according to detailed cleaning instructions perform is. This requirement applies to both drug and drug manufacturers. Cleaning processes are processes for removing residues in Production equipment produced by the product previously produced in the same plant, through environmental influences or caused by maintenance.

A Manufacturing plant is considered pure, if a predetermined and scientific vested Acceptance limit is met. In determining the acceptance criteria, which is the responsibility of the pharmaceutical manufacturer, are the batch size that Dosage, toxicological properties and the size of the production facilities to take into account. "Nothing in the universe is truly clean ", i.e. Residues of Active substances, reagents or detergents can be prepared on the basis of state-of-the-art analytical technology be detected in lowest substance concentrations, too if in these concentrations for the residuals no biological efficacy is more detectable. It therefore makes sense to work closely with toxicologists and Medicines acceptable residue limits in active ingredients or medicinal preparations.

acceptance criteria

The maximum permissible Residual amount of an active substance in a synthesis plant depends on the effectiveness and amount of administration of the following prepared drug in a drug.

• – Sichtbar-sauber-Kriterium
• – 10-ppm-Kriterium
• – Dosiskriterium.

The following acceptance criteria are recognized in practice and are generally applied:

• - Visible-clean criterion
• - 10 ppm criterion
• - Dose criterion.

The Visible-Clean-Criterion states that all parts of a system must leave a visually clean impression, which means that visually no substance residues can be recognized without aids. Scientific investigations have shown that starting from a white substance, active substance residues in a concentration of 100 μg per 25 cm ^{2 are} still visually visible. However, due to the increasingly larger and more complex manufacturing facility, the evaluation of efficient cleaning is no longer sufficient only by considering the plant components.

The 10 ppm criterion states that no more than 10 ppm of an active substance be contained in another active substance or a medicinal product allowed to. The maximum allowed Contamination is dependent from the batch size of the follow-on batch.

The Dose criterion is applied in the case when the therapeutic Daily dose of an active ingredient in a drug is known. When using this acceptance criterion, no more than 1/1000 of the normal daily dose of an active ingredient in the daily dose be included in the subsequent product. Similar to the dose criterion defines the pharmacological criterion that says 1/1000 the minimum daily dose of a substance in another active substance be detectable, this being the maximum applied daily therapeutic dose enters into the calculation.

• 1. Teilfaktor: Es wird im allgemeinen angenommen, dass Wirkstoffe in einer Dosis von einem Zehntel ihrer normal vorgeschriebenen Dosis keine Wirksamkeit mehr zeigen.
• 2. Teilfaktor: Dieser Faktor ergibt sich aus dem Toleranzbereich des Reinigungsvalidierungsprogramms und wird auch als Robustheitsfaktor bezeichnet. Der
• 3. Teilfaktor wird als kalkulatorischer Sicherheitsfaktor eingefügt.

The factor 1000 results from the multiplication of the partial factors 0.1 × 0.1 × 0.1, which are defined as follows:

• 1st partial factor: It is generally assumed that active substances in a dose of one-tenth of their normal prescribed dose are no longer effective.
• 2nd partial factor: This factor results from the tolerance range of the cleaning validation program and is also called the robustness factor. Of the
• 3rd partial factor is inserted as an imputed safety factor.

For substances, which are in the developmental stage, is the dose criterion generally not applicable because within clinical trials only the dose determination takes place.

Gary L. Fourman and Michael V. Mullen developed a new method determining the acceptance limits, in which all three criteria, Dose criterion, 10 ppm and "Visible clean "- criterion, Fulfills become.

method for sampling

The Choice of a specific sampling technique of surfaces after successful cleaning hangs both on the properties of the substance to be purified, e.g. of the Solubility, as well as the system configuration. It is crucial if the product-contacting surfaces e.g. little porous and are easily accessible or whether the critical plant components consist of pipe systems and reaction vessels, which do not allow direct access.

The Sampling by rinsing or by wipe test, referred to as a final rinse or Swab method, and the study of a placebo lot on parts of the predecessor product are the recognized and applied in practice techniques, where the latter for cost and time reasons rarely used. For proof of purity of equipment in the production of pharmaceutical Drugs, the placebo process plays virtually no role, so only swab and final rinse procedures be taken into account at this point.

At the Swab test is performed with a suitable material, e.g. a cotton swab, a previously defined product-contacting surface mechanically cleaned. The material was previously mixed with water or a suitable solvent impregnated, after sampling, the swab, also referred to as Swab, extracted and the extract then on remaining traces of product examined. Choosing only a few sampling points under circumstances not representative for the Are considered to be critical in this method.

For less accessible Plant parts is the final rinse method of choice, at which flushed the entire system and the last rinse water analyzed analytically. In this method, the analytic Effort compared to the Swab test significantly lower, as the extractive Processing of the cleaning material is eliminated. Disadvantages of this method are that from the analysis results determined not on the perfect purity of the whole plant can be closed and the consumption of larger quantities solvent For cleaning is expensive, including what the environmentally sound disposal As.

Choice of analysis method

The Application of an analytical method for the determination of the active substance content in the cleaning solutions hangs in first of all, which acceptance limit for the substance residue to be determined by Manufacturer was specified.

Generally common the use of chromatographic methods such as high performance liquid and thin layer chromatography and spectroscopic techniques e.g. UV spectroscopy, wherein the high performance liquid chromatography to classify as the most sensitive, but also costly method is. The determination of the detection limit of the analytical method is as a necessary component within the entire cleaning validation program required by the FDA. An increasingly used lately Analysis method is the determination of the organic carbon content in rinse solutions by TOC method. The principle of this method is based on the oxidation of carbon to carbon dioxide, which subsequently is detected quantitatively. Advantages of the TOC determination are the high detection sensitivity and low effort for sample preparation.

The Food and Drug Administration pointed out that the TOC method is a non-specific method acts, that is, that between the possible organic Impurities can not be differentiated. this fact is of particular importance when it comes to detecting e.g. highly active or toxic substances. The residue of additionally used cleaning agents, e.g. of surfactants, can be determined by such methods as pH and conductivity measurements, Titrations and gravimetric techniques are detected, however These methods are characterized by a very low sensitivity out.

Analysis methods used to evaluate cleaning samples must be fully validated, this means, it is in addition to proof of reproducibility, precision, Accuracy and robustness also show that within the workspace a linear dependence between substance concentration and measurement signal exists and that the detected detection and quantification limits are far below the for the cleaning process specified acceptance limits are.

recovery

By Investigations on the recovery of substances on material surfaces is to provide proof that the sampling techniques used Swab or final rinse procedures, are suitable residues on abutment surfaces to remove quantitatively.

A material surface comes with a solution defined substance concentration contaminated and then purified. In the cleaning medium used should ideally be 100% of this Be contained substance. Due to different influencing factors, which are still discussed within this work, this was Ideal state not reached so far.

The Food and Drug Administration refers to this Facts clearly position: "We are seeing again recovery studies (at) 50%, (at even) lower than 50%, it is very hard to say that this is acceptable. "Recovery rates established substances below 50% present a risk in terms of evaluation all highly active agents. By still remaining precursor In the pharmaceutical plant, the quality of the secondary product becomes significant influenced, and it can come to sub-mixes.

The Development, specification and documentation of the used cleaning process, the acceptance limits, the sampling method, the analytical method and the investigation of the recovery rates of surfaces Simulated sampling are the prerequisites for carrying out the Cleaning validation. This overall process ensures that that cross contamination and product entrainment excluded can be.

This is a crucial aspect of quality assurance in the Wirkstoffund Drug manufacturing.

Selection of active substances and solvents

Selection of active ingredients

When Steroids becomes a group of naturally occurring compounds referred to from the skeleton of cyclopentanoperhydrophenanthrene derive and predominantly today be synthesized. The steroids include the sterols, bile acids, vitamins the D group, some alkaloids and glycosides, and the steroid hormones.

• – Estrogene (C₁₈-Steroide)
• – Androgene (C₁₉-Steroide)
• – Gestagene und Corticosteroide (C₂₁-Steroide)

The most important classes of steroid hormones are classified according to their number of carbon atoms and pharmacological characteristics as follows:

• - Estrogens (C ₁₈ steroids)
• - androgens (C ₁₉ steroids)
• - progestins and corticosteroids (C ₂₁ steroids)

estrogens and progestins are highly effective hormonal agents used for pregnancy detection and prevention and find widespread use in hormone replacement therapy.

estrogens affect the growth of female genitalia and are for the Training of external sexual characteristics responsible for the woman. Together with the progestins they control the cyclic changes on the female sex organs, the structure and the repulsion of the endometrium and the preparation of the uterus on the implantation of the fertilized ice. Besides, they work the metabolism, the bone structure, the blood clotting u.a. androgens, e.g. The testosterone and its derivatives are used to treat male Potency or fertility problems used.

Androgens cause the development of male genitalia, they are responsible for the formation of the secondary sexual characteristics of the man, they also have a protein-building effect and affect bone growth. The following steroid hormones are considered (Table 1): Table 1: Active substances used

in the The following text generally uses the trivial names.

The include substances described to the classic product range of Jenapharm GmbH or are active ingredients in the development. The selection was made on the basis of special structural features and the development status, in Table 2 are next to these two Criteria also described the main areas of application.

Table 2: Structural features and application of the active substances from Table 1

Selection of solvents

steroid drugs are practically insoluble in water, so this medium for the cleaning can be excluded. To one as possible Achieving efficient cleaning success is a solubility of at least 10 mg substance per milliliter of solvent, which corresponds to 'little soluble' according to Ph. Eur.

• – Ein großer Teil der Wirkstoffsynthesen erfolgt in diesen Lösungsmitteln.
• – Steroide sind im allgemeinen gut löslich in diesen klassischen Medien.

For the investigations, the solvents methanol, acetone and tetrahydrofuran were selected for the following reasons:

• - A large part of the drug syntheses takes place in these solvents.
• - Steroids are generally well soluble in these classic media.

The dissolution behavior of the active pharmaceutical ingredients fixed in 0 was investigated in the following solvents (Table 3):
Methanol for spectroscopy, quality Uvasol ^®, purity (GC ¹⁾ ≥ 99.9% acetone for spectroscopy, quality Uvasol ^®, purity (GC) ≥ 99.9% tetrahydrofuran for liquid chromatography, quality LiChrosoly ^®, purity (GC) ≥ 99.9%

The Determination of solubilities carried out gravimetrically: approx. 5g of substance were mixed with methanol, acetone or tetrahydrofuran at room temperature until a while Sediment was clearly visible. After settling the undissolved crystals was 1.0 ml of the supernatant saturated solution pipetted off, after distilling off the solvent, the weight was carried out of the residue.

The Cutoff values for Methanol, tetrahydrofuran and acetone were taken from the literature (Table 3).

As 'UV Cut off' is the respective wavelength in which a 10% transmission of light through 1 cm solvent possible becomes. The cutoff values of solvents limit the use of non-selective spectroscopic methods for analytical content determination of substances that only then possible is when the substance molecules absorb light above this value. This situation had a decisive influence on the analytical method development within this work.

Table 3: Solubilities of the active substances (section 0) in mg / ml in the solvents acetone, tetrahydrofuran and methanol at room temperature, cutoff values of the solvents

The subsequent recovery studies were conducted with high purity solvents carried out, commercially available as 'spectroscopy' or 'liquid chromatography' solvents. The purification of synthesis equipment in daily practice takes place with solvents the quality 'pure'.

At the Comparison of the specifications of the solvents 'for spectroscopy' and 'for liquid chromatography' with the quality requirements of pure quality methanol, acetone and tetrahydrofuran could be differences in terms of the gas chromatographic content and water content (Table 4).

Table 4: Specifications of methanol, acetone and tetrahydrofuran, for spectroscopy 'and, for liquid chromatography' and methanol, acetone and tetrahydrofuran, pure.

With The help of gas chromatography has been used in routine cleaning the synthesis equipment used with the within this work used high purity solvents compared with regard to their contamination profiles.

All tests were carried out on a GC system, Autosystem XL 'of Perkin Elmer Life Sciences GmbH, Rodgau-Jügesheim, using the following gas chromatographic conditions: Pillar: Permabond OV-1-DF-0.25, 25 m long, 0.32 mm inside diameter temperatures: Column space 50 ° C isothermal, injector 200 ° C, detector FID 300 ° C Carrier gas: He 5.0, 72 kPa form River: 1 ml / min Split 1:50 injection volume 1.0 μl

The analysis of the gas chromatographic examination revealed no significant differences with regard to organic impurities in the solvents used (were).

It Gas chromatograms of solvents became Acetone, pure, methanol, pure), tetrahydrofuran, pure, acetone for spectroscopy), Methanol for Spectroscopy tetrahydrofuran for Liquid chromatography created. Acetone 'pure' contained approx. 0.03% methanol; in tetrahydrofuran and methanol was the content of foreign solvents independently from the quality far below 0.01%, allowing the use of high purity solvents in the following recovery studies. In the course subsequent investigations were exclusively acetone for spectroscopy, Methanol for Spectroscopy and tetrahydrofuran used for liquid chromatography.

acceptance criteria at product conversion

The Determination of Acceptance Criteria, i. the permissible concentration of the impurity on precursor in the subsequent product in the synthesis plant is the responsibility of the pharmaceutical manufacturer.

• – Die Dosis für Substanzen in der Entwicklung, z.B. Carbamat J-956, wurde noch nicht festgelegt.
• – Bei der Berechnung der möglichen Grenzkonzentration einer Substanz in einem anderen Wirkstoff, z.B. bei der Produktumstellung von Dienogest auf Desogestrel (Tabelle 6), können sich unakzeptabel hohe Werte ergeben.

In the production of active ingredients by Jenapharm GmbH, the dose criterion and the 100 ppm criterion are used to evaluate the degree of purity and, based thereon, for the release of an apparatus. The latter is used in the following cases:

• - The dose for substances under development, eg Carbamat J-956, has not been determined yet.
• - In the calculation of the possible limit concentration of a substance in another active ingredient, for example in the conversion of dienogest to desogestrel (Table 6), unacceptably high values may result.

According to the dose criterion, the permissible limit concentration C _{g of} an active substance (precursor) in another steroidal active substance (secondary product) is calculated in accordance with Equation 1.

Equation 1: Permitted limit concentration C _g , Minimum applied daily therapeutic dose of the precursor (TD _min ), Maximum applied daily therapeutic dose of the derived product (TD _max ) C G = TD min / TD Max × 1000

The tolerable total amount MS as contamination of the subsequent batch results taking into account the batch size of the follow-on batch MF (Equation 2).

Equation 2: Tolerable impurity quantity in the secondary product (MS), batch size of the secondary product (MF) MS = C G × MF / 100

The listed Active ingredients are listed in Table 5 with the commercial preparations and their dose ranges listed.

Table 5: Prepared medicinal products of the considered active substances (Table 1) and dose ranges

The Calculation of limit values when applying the dose criterion accordingly Equation 1 gives the matrix of limit concentrations for conceivable Product transitions (Table 6).

Table 6: Limit concentrations (%) for all conceivable product transitions of the active substances considered (without carbamate J-956)

The for multi-purpose plants an active ingredient technology developed cleaning concept includes the definition of a so-called unit allocation or sum acceptance criterion for Evaluation of the degree of purity: A chemical synthesis plant sets consisting of several subsystems, such as reactors, filtering devices, Stills, jars, etc. together. Each individual apparatus will, if necessary, one Subjected to cleaning, the so-called Teilanlagenkriterium of 10 ppm is used to evaluate the cleaning level of a part.

The cleaning sum criterion of 100 ppm results from the sum of the unit criteria, if a maximum of 10 devices are connected to a synthesis system.

Within the development of the cleaning strategy of the drug synthesis plants Jenapharm GmbH has determined that at calculated limit concentrations> 100 ppm instead of Dose criterion the 100 ppm criterion is to be used as the cleaning sum criterion.

• – Produktwechsel zu Oxim J-867
• – Umstellung von Desogestrel auf anderen Wirkstoff

This results in an acceptance limit of ≤ 0.01% for possible product conversion variations of the active substances mentioned in 0, with the following exceptions:

• - Product change to Oxime J-867
• - Conversion of desogestrel to other active ingredient

In both cases are the maximum allowable Consider concentrations in accordance with Table 6.

Within the development of analytical methods for the determination of the active substances listed it requires that the detection limits determined for all considered Substances far below the calculated acceptance limits. In Assuming that the cleaning takes place with 10 l of solvent and the following batch with a batch size of 10 kg is produced, according to Table 7, the following limit concentrations, if in the synthesis plant residual contamination at the analytical Detection limit existed. The 10 ppm acceptance criterion is only at a follow-up approach of at most 100g, this batch size can in the daily Routine be excluded.

Table 7: Calculated limit concentrations of impurities at the analytical detection limit, batch size: 10 kg, amount of solvent for cleaning: 10 l

experimental part

analytical methods development

to Quantitative determination of the active substances mentioned in section 0 could theoretically spectroscopic and chromatographic analysis methods be considered. All six active substances are due UV-active, but the solvents used, especially acetone, due to their transmission behavior (Table 3, cutoff values) determination by a non-separating method not too.

• – Robuste reproduzierbare Durchführung
• – Anwendbarkeit auf alle Substanzen durch schnelle Variation von Eluentenverhältnis und Detektionswellenlänge
• – Kurze Analysenzeiten
• – Die Nachweisgrenzen für die zu untersuchenden Substanzen sollten weit unter den geforderten Akzeptanzlimits liegen.

Selective chromatographic methods such as high performance liquid chromatography (HPLC) and thin layer chromatography (DC) were thus favored. DC methods are only suitable for semi-quantitative determinations, the evaluation can be done visually or with the help of scanners. Due to the low sensitivity of thin-layer chromatography methods, these were not considered for the following investigations. In general, the detection limits for active substances such as dienogest and estradiol valerate in the DC are approx. 0.004-0.02 mg per ml. ^Ii The now preferred HPLC method should meet the following requirements:

• - Robust reproducible execution
• Applicability to all substances due to rapid variation of eluent ratio and detection wavelength
• - Short analysis times
• - The detection limits for the substances to be investigated should be far below the required acceptance limits.

The selective separation of said substances from their respective potential by-products, such as degradation products and synthetic by-products, was not required because of the minimal concentrations the test solutions all Impurities (Table 20 - Table 24) remain below the detection limit.

For the quantitative Investigation of steroid solutions was a reversed-phase chromatography on a non-polar modified Silica gel phase developed under the following conditions:

Pillar:

Hypersil ^® ODS (RP 18 - phase), Length: 125 mm, Diameter: 4 mm Average particle size: 5 μm, Supplier: Scientific Instrument Engineering Dr. med. Ing. Herbert Knauer GmbH, Berlin

Mobile phase:

Mixture of acetonitrile and water (Table 8)

Qualities used:

• - Acetonitrile gradient grade for liquid chromatography, quality LiChrosolv ^® , manufacturer: Merck KgaA, Darmstadt, purity (GC) ≥ 99.9%
• - Water: Ultrapure water, produced by post-treatment of demineralised water in a 'Nowapure select' water purifier, manufacturer: KSN Wassertechnik, Nistertal, conductivity ≤ 0,0625 μS / cm (20 ° C), total organic carbon content (TOC) <500 ppb Flow rate: 1 ml per min Column temperature: about 40 ° C Detection method: UV (DAD (photodiode array detector), wavelength range 190-800 nm)

All Investigations were carried out on a HPLC high pressure gradient plant the company Shimadzu Europa GmbH, Duisburg, consisting of two high-pressure pumps LC-10A, a SIL-10A autosampler, a column oven CTO-10A, a degassing module DGU-14A and a photodiode array detector SPD-M10A. The facility was over a system controller SCL-10AV using the chromatography software Class VP controlled, with the likewise the evaluation of the chromatograms took place.

On Reason for the different molecular structure and thus different polarity of the six active substances became the composition of the mobile phase varied. The detection wavelengths were recorded by recording the UV spectra in the range 210 nm - 350 nm on a UV spectrometer 'Specord 200 'of the company Analytik Jena AG, Jena, determined Table 8)

Table 8: Substance-specific HPLC conditions and retention data of all active substances according to 0

• 1 : (1): UV spectrum of estriol in methanol, conc. 0.05 mg / ml, d = 10 mm, A (specific absorption = absorption of a substance in a 1% solution at a layer thickness of 1 cm) = ( 1%, 1cm) = 72, ⎕ = 2070 l / mol * cm (⎕ _max = 281 nm) (2): UV spectrum of estradiol valerate in methanol, conc. 0.10 mg / ml, d = 10 mm, A (1%, 1cm) = 58, ⎕ = 2072 l / mol · cm (⎕ _max = 281 nm)
• 2 : (3): UV spectrum of dienogest in methanol, conc. 0.01 mg / ml, d = 10 mm, A (1%, 1 cm) = 674, ε = 20990 l / mol · cm (λ _max = 304 nm) (4): UV spectrum of desogestrel in methanol, conc. 0.08 mg / ml, d = 10 mm
• 3 : (5): UV spectrum of oxime J-867 in methanol, conc. 0.01 mg / ml, d = 10 mm, A (1%, 1 cm) = 526, ε = 23670 l / mol · cm (λ _max = 264 nm) A (1%, 1 cm) = 519, ε = 23320 l / mol · cm (λ _max = 299 nm) (6): UV spectrum of carbamate J-956 in methanol, conc. 0.01 mg / ml, d = 10 mm, A (1%, 1 cm) = 549, ε = 28560 l / mol · cm (λ _max = 272 nm) A (1%, 1 cm) = 489, ε = 25470 l / mol · Cm (λ _max = 289 nm) A (1%, 1 cm) = 549, ε = 28560 l / mol · cm (λ _max = 298 nm) A (1%, 1 cm) = 549, ⎕ = 28560 l / mol · cm (⎕ _max = 272 nm) A (1%, 1cm) = 489, ⎕ = 25470 l / mol · cm (⎕ _max = 289 nm) A (1%, 1cm) = 549, ⎕ = 28560 l / mol · Cm (⎕ _max = 298 nm)

Validation of the analytical procedures

Of the Proof that the analytical methods for the intended purpose had to be in terms of parameter selectivity, precision, linearity, Detection and quantification limit, accuracy and robustness for all six considered active substances within a defined concentration range be provided. The validation was carried out according to the ICH guidelines Q2A 'validation Analysis Methods: Definitions and Terminology 'and Q2B' Validation of Analysis Methods: Method'.

precision

The validation parameter "precision" indicates the degree of scattering of individual values around a mean as a result of random errors. The standard deviation s (equation 3) or the relative standard deviation s _rel (equation 4) apply as scattering _measures .

Equation 3: Calculation of the standard deviation s; x _i - single result, x _M - average, N - number of measurements

Equation 4: Calculation of the relative standard deviation s _rel (%) (= variation coefficient V _k )

The Scattering of the values is done by the analyzer itself (System precision) and by the sample preparation steps, e.g. Weighing, resolution and Dilution, Extraction, etc. (method precision).

system precision

Within In this work, the relative standard deviation was used as a quotient from standard deviation and mean of the peak areas of 6 consecutive Injections of a sample solution the concentration 0.001 mg / ml for each substance is calculated in methanol as per 4.1 (Table 9).

methods precision

six sample solutions the concentration 0.001 mg / ml for each substance in methanol was prepared separately and each chromatographed twice. The relative standard deviation of the response factors as a quotient peak area (Average of 2 injections per weight) and the corresponding Substance weight was calculated (Table 9).

Table 9: System Precision and Method Precision for all active substances according to 0; System precision: Relative standard deviation of peak areas (%) (6 consecutive injections of a substance solution); Method precision: Relative standard deviation of the response factors (%) (6 separately prepared substance solutions)

When assigning the standard deviation to a probability, a normal distribution of the individual values can be assumed, which can be tested with the help of various statistical tests, eg the rapid test according to David. ^{Error! Bookmark not defined.}

Equation 5: calculation of the statistical value W for the evaluation of the normal distribution according to David; R as the difference between the largest and smallest individual result (= span), s - standard deviation

Table 10: Test value W for the evaluation of the normal distribution of the measurement results according to David (P = 99%, f = 1), R = span, s - standard deviation, P - significance level, f - degree of freedom

To The statistical table of David is a normal distribution for one assumed probability of 99% for N = 6 (number of individual measurements) if the calculated value W is between 2.15 (lower limit) and 3.16 (upper limit). Table 10 shows that all Values of system and method precision can be assumed to be normally distributed.

Striking strongly deviating single analysis results could not be determined, so that on the execution of outlier tests, eg after Dixon or Grubbs ^{error! Bookmark not defined.} , was waived.

linearity

The linearity describes the proportional dependence the measurement signal, e.g. the peak areas or heights in chromatographic processes, from the concentration of examining substance.

Of the Workspace, for the assessment of linearity The analytical method used was to be chosen so that the concentrations used are within the calibration line were. Starting from a stock solution of 0.15 mg / ml separately for each active substance a dilution series to give the following concentrations: 0.12 mg / ml, 0.09 mg / ml, 0.06 mg / ml, 0.03 mg / ml, 0.008 mg / ml and 0.001 mg / ml. A double determination was made per concentration level, the mean values obtained were used for the calculation.

Table 11: Linearity test: peak areas (area units FE) as a function of the concentration of the analytes; Residual standard deviations according to equation 6 for the linear and quadratic calibration function

The calculation of the calibration functions y = bx + a as function 1.Grades and y = cx ² + bx + a as function 2.Grades as well as the residual standard deviations was done with the help of the software 'MVA Version 1.0', Novia GmbH Saarbrücken (Table 11) (Table 11) ( ). The residual standard deviation is a measure of the dispersion of the analysis values around the calibration straight line or curve and can be calculated from the standard deviation of the differences of the individual measured values from the corresponding regression curve.

So could for all active ingredients in the concentration range described the qualitative Statement be made that the calibration curve with a function 2nd degree is described better than with a 1st degree function, since the calculated residual standard deviations for the quadratic function are lower.

At the visual comparison of the first and second order regression lines does not readily indicate which function is preferable, Furthermore the correlation coefficient r gives very good agreement in all cases of the individual measured values with the calibration curve (generally greater than 0.99). However, it should not be used as the sole evaluation criterion for linearity become (5).

calibration 1st degree (linear) and 2nd degree (square) with confidence interval (95%) to evaluate linearity for the 1st degree (linear) and 2nd degree (square) substances with confidence interval (95%) to evaluate linearity for the Substances Estriol, estradiol valerate, dienogest, desogestrel, oxime J-867 and carbamate J-956 in the concentration range 0.001 mg / ml to 0.15 mg / ml was used.

The quantitative assessment of linearity was made by the almond test. ^{Error! Bookmark not defined.} The dif The variance variances were calculated (Equation 6) and used to calculate a statistical quantity that was compared with the F-Table. If the calculated test value was smaller than or equal to the table value F, a linear calibration function could be assumed.

Equation 6: Calculation of the difference of the deviation variances ⎕s ² , s _lin - residual standard deviation of the linear calibration function, s _qu - residual standard deviation of the quadratic calibration function, N _lin , N _qu - number of concentrations which have entered the linear or quadratic function ⎕s 2 = (N lin - 2) · s lin 2 - (N qu - 3) · s qu 2

Equation 7: Calculation of the statistical test variable PG for the evaluation of the linearity after comparison with the F-table (almond test)

The test parameters PG for the calibration functions of estriol, estradiol valerate, dienogest, desogestrel, oxime J-867 and carbamate J-956 were calculated and compared with the table value of the F-table with f ₁ = 1, f ₂ = 4 and P = 99% , All calculated values were smaller than the table value (21,200), thus the calibration functions could be assumed to be linear for all substances within the described working range (Table 12).

Table 12: Calculated test parameter PG for the assessment of linearity (adaptation test according to Mandel); Table value of the F-table (f ₁ = 1, f ₂ = 4, P = 99%): 21,200

With Help of the residual analysis, in which the vertical deviations of the Measurement results from the calibration curve as a function of the concentration could be considered a normal distribution around the zero line for the linear and quadratic calibration function with one exception (quadratic Function of oxime J-867) The use of the linear Calibration function for all six active substances in the considered concentration range is thus permissible.

residuals dependent on from the concentration for the linear and quadratic calibration function for the substances estriol, estradiol valerate, Dienogest, desogestrel, oxime J-867 and carbamate J-956 in the concentration range 0.001 mg / ml to 0.15 mg / ml were shown.

Detection and determination limit

When Detection limit will be the smallest still contained in a sample detectable amount of analyte, the limit of quantification quantified quantity. Between detection limit and limit of determination is the detection limit, which is generally twice the detection limit equivalent. The detection limit should be well below the required Specification or acceptance limits are and no more than 50% the lower specification limit.

Within In this work, the detection limit was determined by further dilution of the Linearity solution lowest Concentration (0.001 mg / ml) with methanol using the signal-to-noise ratio method determined (Table 13). The solution was diluted for so long until the substance of the diluted Samples no longer delivered a measurement signal. The concentration at the The relationship from signal and baseline noise was about 3, was used as detection limit accepted. The quantitative limit of quantification was threefold Detection limit calculated.

Table 13: NWG - Detection limits (mg / ml) for all active substances (section 0) in methanol, determined by signal-to-noise ratio method

Further in practice applied methods to determine the detection limit are the determination by means of calibration methods and the method of acceptable dispersion.

The Determination of the detection limit by means of the calibration curve method by recording a calibration curve in a concentration range, which corresponds approximately to ten times the expected detection limit. The.

The detection and detection limit was determined according to the calibration method (DIN 32645); NWG detection limit, EFG detection limit, VB confidence interval ( 4 )

A further and increasingly accepted method for determining the detection limit was recommended by EURACHEM. In this case, six-fold determinations of dilute calibration solutions of defined concentration are carried out and from these the relative standard deviation s _{rel is} determined. The concentration at which a standard deviation of about 10% is determined can be referred to as the detection limit of the method.

accuracy

The Accuracy of an analytical procedure is a measure of agreement of the obtained measured value with the correct result and thus gives the extent of existing systematic error. Systematic errors can be e.g. due to errors in sampling or preparation, due to inhomogeneities of the substance to be examined, as a result of faults in the measuring apparatus, etc. occur.

The Correctness of the analytical procedures took place within this Work by determining the recovery rates as a quotient from the measured concentration of the analyte and by accurate Weighing and dissolving defined concentration of the substance (Table 14). There were pro Substance six solutions the concentration of 0.001 mg / ml in methanol separately prepared and measured.

Table 14: Means of recovery rates (= accuracy) (%) for all active substances (Section 0) in methanol; Concentration 0.001 mg / ml; VB - confidence interval of the mean for a probability of 95% and assumed t-distribution, V _k - coefficient of variation

Further options for the exam the correctness are u.a. the comparison of the test results with a certified reference substance whose content is known under Using the t-test and, more rarely, comparing with another validated method.

robustness

By minimal variation of method parameters, e.g. the flow rate or the measuring wavelength, should the reliability of the analytical system (method robustness). About that In addition, studies of process stability, i. stability the test solutions, carried out.

methods robustness

The Robustness of the methods was determined by varying the measuring wavelength (± 1 nm), the flow rate (± 0.1 ml) and column temperature (± 5 degree), the composition of the mobile phase (± 1 part by volume) and by Using two different chromatography columns with same stationary Phase determined so that per substance six different chromatographic Parameter combinations (variants 1 to 6) were applied. Per Variant were 6 content determinations with sample solutions of concentration 0.001 mg / ml for every Substance is 0 in methanol. The salary results of the Variants 2 to 6 were used with the measurement results of variant 1 statistically evaluated by F-test and mean-t-test (equation 8, equation 9).

Equation 8: F test: calculation of the test variable F for comparison of the standard deviations of 2 test series; s ₁ , s ₂ - standard deviations F = s 1 2 / s 2 2 s 1 > s 2

Equation 9: average t-test: calculation of the test parameter PG for comparison of the mean values of 2 measurement series; s ₁ , s ₂ - standard deviations, x _M1 , x _M2 - mean values, N ₁ , N ₂ - number of measurements of series 1 and 2, s _d - mean standard deviation of both series

Table 15: Chromatographic conditions for determining the method robustness of the HPLC method using the example of estriol; Variant 1 - standard analysis method Table 8, p. 24); Variants 2 to 6 - HPLC method with modified parameters; Hypersil ^® 1 - Batch LH136, Hypersil ^® 2 - Batch PH356 ⁴

Table 16: System precision (V _k in%) of the measured salaries of all agents according to 0 depending on the chromatographic conditions (variants 1 to 6); x _M - mean (%); VB - confidence interval of the mean for a probability of 95% and assumed t-distribution (%); V _k - coefficient of variation (%); V variant

Table 17: F-test results (f1 = 5, f2 = 5, 99%) and t-test (f = 10, 95%) for variants 2 to 6 in relation to the standard analysis method Table 8, p. 24) ; V variant

With With a probability of 95%, the mean values differ the series of measurements are not significantly different from each other, the calculated test quantities are below the table value. The calculated statistical t-values for desogestrel (Variant 2) and oxime J-867 (variant 5) are only slightly above the Table value, the deviation can be neglected as the for this Measurement series with variances of 0.36 are very small. For all value ranges could variance homogeneity with a probability of 99%, the analytical method is therefore sufficiently robust.

process stability

The Chromatography of the methanolic drug solutions used for the study the system precision were used immediately after preparation (starting value) and after six hours. During this period, they were in daylight in white glass vials, under exclusion of light in amber glass bottles, at room temperature and in the refrigerator stored at 2-8 ° C. The comparison of the peak areas the starting value with the areas, obtained by chromatography after six hours storage under the above conditions described, showed no significant differences.

selectivity

The selectivity The HPLC methods were proved by the proof that the solvents to be used Methanol, tetrahydrofuran and acetone under the indicated chromatographic Conditions clearly from the signal of the drug to be determined gave distinguishable signals (Table 18).

Table 18: Relative retention times (RRT) of the solvents methanol, tetrahydrofuran and acetone in relation to the retention times (Rt) of the active substances according to 0; HPLC methods used: see Table 8)

Of the Detection of the selective separation of synthesis by-products and degradation products of the active substances to be investigated within the analytical methods was for the following reasons not required: during the cleaning procedures (Chapter 0) were the active ingredients no Stress conditions, such as extreme pH changes, intense light influence or strong temperature fluctuations, subjected to a significant increase the degradation products could be excluded. The influence of these factors on the by-product profile of the substances was within the development work at Jenapharm GmbH & Co. KG examined in detail.

Investigations of cleaning solutions with the analytical methods described were carried out in the trace range, so that potential synthesis by-products or degradation products were below the detection limit. The impurity profiles ^{6 of} all steroid substances investigated within this work are known (Table 19 - Table 24).

Table 19: By-products and their content in estriol

Table 20: By-products and their content in estradiol valerate

Table 21: By-products and their content in dienogest

Table 22: By-products and their content in desogestrel

Table 23: By-products and their content in oxime J-867

Table 24: By-products and their content in carbamate J-956

The Suitability of the developed analytical methods for quantitative determination of the six active substances was in the concentration range 0.001 to 0.15 mg / ml. The detection and quantitative Limits of quantification as the most important criteria were far below the required acceptance limits for Cleaning procedures, independent of which criteria, 100 ppm or dose criterion, for the assessment come into use.

Recovery investigations

Final-rinse process

Experimental Procedure

• 10 mg of steroid substance (section 0), exactly weighed, were dissolved in methanol or acetone or tetrahydrofuran to 100.0 ml (concentration: 0.1 mg / ml). Using a 1 ml volumetric pipette, the following material surfaces were wetted so that approximately an area of 12-16 cm ^{2 was} contaminated:
• Borosilicate glass: Watch glass dishes DIN 12341 Duran ^® , supplier: Fisher Scientific GmbH, Schwerte, Article No .: 9263432
• Soda-lime glass: watch glass bowls DIN 12341 soda-lime-glass, supplier: Fisher Scientific GmbH, Schwerte, Article No .: 9263005
• Polyethylene: PE-LD cover, supplier: Fisher Scientific GmbH, Sword, article no .: 9033201
• Stainless steel: Evaporating dishes, supplier: Fisher Scientific GmbH, Sword, article no .: 9000460
• Silicone: Sealing Washers, Supplier: Fisher Scientific GmbH, Sword, article no .: 9209435
• Teflon: PTFE watch glass bowls, supplier: Fisher Scientific GmbH, Sword Article No .: 3424269

The contaminated materials were left at room temperature for about 15 hours stored, this was approximately full Evaporation of the solvents used.

The entire series of experiments was carried out with substance solutions of concentration 0.01 repeated mg / ml.

The Substance residues were with 10 ml of the appropriate solvent without mechanical contact the material surfaces rinsed, subsequently the content of the substance solutions was determined without further sample preparation using high-performance liquid chromatography.

Of the Process of wetting and subsequent rinsing was per substance, solvent, Material and concentration 3 times repeated, so that a sufficient ensured statistical coverage of the results obtained was.

In Table 25 is the maximum relative standard deviations of Recovery rates from three individual results, measured at all Materials at a concentration of 0.01 mg / ml, shown. The table already includes the statistical evaluation of the results obtained using the Swab method.

For silicone were relative standard deviations of the recovery rates of more than 5%, all other materials included 1 or 2%. Due to the very low substance concentrations in the solutions used standard deviations below 10% are acceptable.

Table 25: Maximum relative standard deviations (%) of recovery rates for all six active substances with final rinse and swab test; Materials used: borosilicate, soda lime glass, polyethylene, stainless steel, silicone, teflon

Recovery in dependence of equipment materials

The Calculation of the recovery rates of all active substances described in 0 on the materials borosilicate and soda lime glass, polyethylene, Stainless steel, silicone and Teflon were made using Equation 10 for contamination solutions of Concentration 0.1 mg / ml

Equation 10: Calculation of the recovery rates in%, WF recovery (%), GS - content of active substance per ml in the rinsing solution (mg) determined by HPLC, GA - known content of active substance in the contamination solution (weight: 0.1 mg resp 0.01 mg) WF = GS / GA · 100

51 : Recovery rates (%) of the final-rinse method in tetrahydrofuran for liquid chromatography of different materials; Conc: 0.1 mg / ml

6 : Recovery rates (%) of the active substances with final rinse method in methanol for spectroscopy of different materials, concentration: 0.1 mg / ml

7 : Recovery rates (%) of the drugs with final rinse method in acetone for spectroscopy different materials, concentration: 0.1 mg / ml

borosilicate and soda lime glass

Technical glasses can be divided into oxide, non-silicate and non-oxide glasses. Borosilicate glass and alkali-lime glass are among the silicate glasses, the silica glass and the lead glass to the oxide glasses. Among the non-oxide glasses include fluoride, sulfide and nitrate glasses. Borat glasses and phosphate glasses belong to the group of non-silicate glass silicate glasses, and these are the only references within this work, materials with an SiO ₂ structure, in whose lattices alkali and alkaline earth metals or, depending on their application, boric acid, transition metal oxides and rare earths are incorporated are.

As in crystals, the atoms of the glass form a three-dimensional one Network with the difference that the glass is not an ordered grid has. The orientation of the silicon-oxygen tetrahedra in the Glass structure is rigid compared to crystalline compounds Network structures variable and this feature results in a Number of characteristic properties.

Almost all glass plants and pipelines used in the chemical and pharmaceutical industries and a large proportion of laboratory equipment are made of borosilicate glass of the Pyrex type. This type of glass is composed of the ternary system Na ₂ OB ₂ O ₃ -SiO ₂ with about 80 mol% SiO ₂ , which was added to increase the water resistance and to achieve the smallest expansion coefficient Al ₂ O ₃ as a network former

Also Kalknatronsilikatgläser the composition Na ₂ O-SiO ₂ and CaO-SiO ₂ , known under the name window glass come after the addition of about 2-3% by weight of alumina used as laboratory glassware in pharmaceutical production.

at Both types of glass dominate the proportion of silicon dioxide with greater than 60 mol%.

at Use of methanol as flushing medium were in particular estradiol valerate and desogestrel observed significant "substance losses" (table), it was only a fraction of the surface in contact with the glass surface Found again steroid concentrations.

Table: Residual contamination (100% - Recovery rates in%) of estradiol valerate and desogestrel with final rinse method on glass, concentration: 0.1 mg / ml (methanol)

The special structural features of estradiol valerate and desogestrel, ie long alkyl chains or methyl and methylene groups, contribute to the non-polar character of these substances. In direct contact with SiO ₂ molecules, adsorption may occur through formation of oxygen-hydrogen bridges, whereby the geometry of the substance molecules and the free accessibility of terminal hydrogen atoms may play a role.

The adsorption of selected macromolecules, in particular amino acids, on glass surfaces has been described in the literature. Models have shown that the protein poly-L-lysine undergoes stable exothermic hydrogen bonding with measurable binding enthalpy to SiO ₂ surfaces. This adsorption effect is enhanced by the presence of polar solvents such as water or methanol due to the presence of polar solvents such as water or methanol due to hydration or solvation.

Mantus et al. investigated the adsorption behavior of polyamino acids of the general structure H ₂ N-CHR-CO - [- HN-CHR-CO-] _n -NH-CHR-COOH on planar glass substrates by means of static secondary ion mass spectrometry. The mass spectra of alkyl polyamino acids, including polyglycine (R = - H), polyalanine (R = - CH ₃ ), polyvaline (R = - CH (CH ₃ ) ₂ ) and polyleucine (R = - CH ₂ CH (CH ₃ ) ₂ ) showed an increase in the intensity of the main ion peak as the number of alkyl chains in the molecule increased. The peak intensity is in addition to the number of monomers in the macromolecule dependent on the mobility, that is, the interaction of the polymer with the surface. Thus, there was a much more intense adsorption of nonpolar higher alkylated polyamino acids on the glass surfaces. Estradiol valerate and desogestrel show similar behavior on plastic and stainless steel surfaces in terms of their adsorption affinity.

at Use of acetone as a cleaning medium could be studied for all Substances generally recovery rates of more than 90% determined on glass surfaces become.

Teflon

The recovery of the steroidal agents according to the final rinse procedure is generally at least 90%, so this material can be classified as harmless with regard to possible contamination after cleaning ( 51 . 6 . 7 ).

The polytetrafluoroethylene (Teflon) known for its chemical resistance [-CF ₂ -CF ₂ -CF ₂ -CF ₂ -] _n is prepared by gaseous tetrafluoroethylene by overpressure polymerization and further processed before use by special sintering methods. ⁱⁱⁱ

Teflon is used in pharmaceutical manufacturing plants predominantly as sealing and Valve material used, in part, also come piping whose Innenoberfäche coated with a Teflon film is used in production.

in the Compared to the reactor materials glass and stainless steel is in sealing materials the product-touching area by orders of magnitude smaller, however, these materials are particularly effective at high, e.g. Carcinogenic substances should not be disregarded. This Facts should be given special consideration if silicone materials are used instead of Teflon.

silicone

Due to their SiO ₂ structure, silicones have glass properties, which are attributed to their resistance to heat and cold and their chemical resistance. On the other hand, this material is characterized by elasticity and plasticity and a water-repellent effect, ie plastic properties. For this reason, silicones are used as sealants and insulating materials, inter alia, in the construction industry, in electrical engineering and in chemical and pharmaceutical plant engineering.

Chemical structure of silicones; R - alkyl radical

• 1. Reduktion von Siliziumdioxid zu Silizium mit Kohle bei Temperaturen über 1700°C (endotherme Reaktion)
  
• 2. Kupferkatalysierte Reaktion von Alkylhalogeniden und Silizium zu Alkylchlorsilanen (Müller-Rochow-Synthese)^iv
  
  
  Die Reaktion ist endotherm und läuft bei Temperaturen von ca. 300 °C ab.
• 3. Hydrolyse von Alkylchlorsilanen zu Silanolen
  

With the silicones, a new fully synthetic class of materials has been developed since there are no carbon-silicon bonds in natural organic systems. The synthesis of silicones occurs in three steps.

• 1. Reduction of silicon dioxide to silicon with carbon at temperatures above 1700 ° C (endothermic reaction)
  
• 2. Copper-Catalyzed Reaction of Alkyl Halides and Silicon to Alkylchlorosilanes (Müller-Rochow Synthesis) ^iv
  
  
  The reaction is endothermic and takes place at temperatures of about 300 ° C.
• 3. Hydrolysis of alkylchlorosilanes to silanols
  

The Silanols are not isolated by targeted control of the polycondensation step the formation of chain or ring-shaped silicones from silanediols takes place or of spatial structures with differently pronounced degree of crosslinking.

Silicone structures, 1-chain or ring-shaped, 2-three-dimensional crosslinking, R - alkyl radical

With increasing networking, the plastic properties vary, so within this class of liquids (Silicone oils), viscous materials (silicone resins) and elastomers (silicone rubber) occurrence.

Commercially available silicone products consist of a mixture of silicones with the addition of additives, e.g. UV stabilizers, plasticizers and heat-cold stabilizers. The elastic three-dimensional cross-linked silicones are from whole different strength and porosity thus make a critical Factor when used as part of chemical synthesis plants be used.

The Recoveries shown in the table were obtained by wetting and then rinse of commercial Silicone rubber gaskets section after single use determined.

Table 27: Recovery rates (%) of the drugs with final rinse method on silicone, solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy), concentration 0.1 mg / ml

Recoveries for active ingredients such as estradiol valerate and desogestrel were below 50% or below 10% determined.

is the porous one surface structure the sole cause of the silicone the 'disappearance' of Drugs? The substance solutions can partially penetrate into the intermolecular spaces, However, the question arises, why not in general at all steroid substances penetrated into the silicone surface.

silicones have pronounced hydrophobic properties. Estradiol valerate and desogestrel very non-polar substances, allowing an intense interaction between these steroid molecules and the silicone surface can be accepted.

stainless steel

The recovery of the steroids estriol, estradiol valerate, dienogest, desogestrel, oxime J-867 and carbamate J-956 on stainless steel surfaces using different concentrations of substance in the wetting is in 8th shown. Desogestrel, oxime J-867 and carbamate J-956 remain about 50-60% on the metal surface when using methanol as a strong polar solvent.

The Interaction of steroidal agents with stainless steel surfaces was examined more intensively.

8th shows recovery rates (%) of the drugs with final rinse method on stainless steel, solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy (section 0), concentration: 0.1 mg / ml and 0.01 mg / ml

polyethylene

When Polyethylenes are versatile thermoplastics referred to by means of low-pressure or high-pressure polymerization, catalytically controlled, made of ethylene. Dependent on from the reaction conditions Degree of crosslinking, molecular size and the Degree of crystallinity determined become, through these parameters become chemical-physical properties like hardness, Transparency, solubility, Elongation, melting range, etc. are decisive Are defined.

polyethylenes have the property when exposed to aliphatic hydrocarbons to swell, opposite polar solvents Like alcohols and ketones, they are considered to be largely at temperatures around 20 ° C Insensitive polyethylene is used as tubing and tubing material in the used widely in pharmaceutical production.

• – Eine visuelle Kontrolle auf Sauberkeit ist nicht möglich.
• – Bei Verwendung organischer Lösungsmittel bei der Reinigung kann es zum Quellen des Kunststoffmaterials kommen, Wirksubstanzen penetrieren in gelöster Form in die Plaste.
• – Die Reinigung bei höheren Temperaturen sollte vermieden werden, da Polyethylene nur bis 60°C in üblichen Lösungsmitteln als praktisch unlöslich gelten.
• – Schläuche und Rohre geringen Durchmessers können generell nur mittels Final-Rinse-Methode gereinigt werden.

The cleaning of plastic hose connections is considered particularly critical for the following reasons:

• - A visual check for cleanliness is not possible.
• - When using organic solvents during cleaning, it may swell the Kunststoffmate Rials come, active substances penetrate in dissolved form in the plastic.
• - Cleaning at higher temperatures should be avoided as polyethylenes are considered practically insoluble only up to 60 ° C in common solvents.
• - Small diameter tubing and tubing can generally only be cleaned using the final rinse method.

investigations within this work for the recovery of steroidal active substances after contact with polyethylene showed in particular for the nonpolar Substances desogestrel and estradiol valerate critical results

Table 28: Recovery rates of desogestrel and estradiol valerate (%) with final rinse method on polyethylene, solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy (section 0), concentration: 0.1 mg / ml and 0.01 mg / ml

For estriol, Dienogest, Oxim J-867 and carbamate J-956 were almost invariably Recovery rates of more than 90% found.

That lipophilic substances, adsorb on plastic surfaces, while hydrophilic molecules hardly interact with plastic, was von Krieglstein and co-workers examined empirically. solutions lipophilic substances of known concentration in phosphate buffer solutions, like Promazine and Promethazine, were in plastic containers within Shaken in a defined period of time, during the subsequent determination the substance content was only about one third of the initial concentration found. When using hydrophilic substances such as salicylic acid or Amidopyrine remained approximate 100% in the buffer solution.

Similar Studies have been published by Kakemi et al. carried out.

Plastic plates Polyethylene, polypropylene and others have become more popular in paraben phosphate buffer solutions Concentrated and concentrated at constant temperature over a kept for a specific period of time. In the study of methyl, Ethyl, n-propyl, n-butyl and benzylparaben could be detected be that the adsorption with a longer carbon chain continuously increased. For example, benzylparaben migrated approximately 60- to 70-fold Comparison to methyl and ethylparaben in the polyethylene surface.

propofol (2,6-Diisopropylphenol) (Ph. Eur. 4.05), dissolved in 5% glucose, as anesthetic intravenous applied. This solution was over longer time in plastic containers (Polypropylene and Polyvinylchloridcontainern) stored, wherein to interactions of the very lipophilic propofol with the container material came. Thus, after 30 days storage of the solution in PVC at room temperature a loss of active ingredient of about 30% was found.

polyethylenes (PE) are dependent from those chosen in the manufacture Reaction conditions a different degree of crystallinity on.

So are made by high pressure PE largely branched, have a low density, and the proportion of crystallinity is at about 40-50%.

at Polyethylenes from low pressure process, the proportion of crystallinity increases significantly to 60-80%, this mostly unbranched products have a higher density.

Steroid molecules interact Priority with the crystalline areas of the polyethylene and enrich get in there.

Recovery in dependence the degree of contamination

Of the Comparison of recovery rates with solutions of different concentration were obtained for shows Borosilicate glass, soda lime glass, polyethylene and Teflon not worth mentioning Differences. Exceptions are silicone and stainless steel.

at Use of silicone has generally been significantly lower Amount of steroid in the rinse solvent found when using solutions wetted at a lower concentration. This typical saturation phenomenon was not further investigated within this work. differences After cleaning of stainless steel surfaces could in particular methanolic solutions be observed with varying concentration, being almost exclusively at lower substance concentration lower recovery rates were determined.

9 : Recovery rates (%) for estriol with final rinse method of various materials (section 0), solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy, concentration of contamination solutions: 0.1 mg / ml and 0.01 mg / ml

10 : Recovery rates (%) for oxime J-867 with final rinse method of various materials (section 0), solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy (section 0), concentration of contaminating solutions: 0.1 mg / ml and 0.01 mg / ml

Swab method

At the Swab method, also referred to as wipe test, with the help of textile materials, e.g. Cellulose, a contaminated area more defined Size cleaned. The material, previously with water or an organic solvent wets then becomes subjected to an extraction. Unlike the final rinse procedure the swab method only cleans a certain area of the plant, subsequently An extrapolation is made to the total area of the plant. The application The Swab procedure requires that the equipment be easily accessible and accessible the sampling of previously defined areas must be representative. Both conditions are generally not present in chemical synthesis plants. Should the Swab method still be used because detected was that by rinsing the plant only a fraction of the remaining residual substance is detected, the equipment is dismantled. The items should be cleaned, then the residual contamination must be determined by swab test. This costly Process that occasionally unavoidable is should be avoided.

In The recoveries were the same material surfaces both for the Final rinse procedure as well contaminated the Swab method (Section 0).

Within These investigations were influenced by material forms such as curvatures characteristic of Pipelines, and edges, i. optimal or less easy to clean surfaces not considered.

Experimental Procedure

• – Genaue Einwaage von 10 mg Steroidsubstanz
• – Verwendete Lösungsmittel: Methanol für Spektroskopie, Aceton für Spektroskopie, Tetrahydrofuran für Flüssigchromatografie
• – Verwendung der Materialien Borosilikatglas, Kalknatronglas, Polyethylen, Edelstahl, Silikon, Teflon
• – Gleichmäßige Benetzung der Oberflächen mit Lösungen der Konzentrationen 0,1 mg/ ml und 0,01 mg/ ml
• – Aufbewahrung der benetzten Materialien bei Raumtemperatur über ca. 15 Stunden

The preparation of the materials to be cleaned was carried out according to the final rinse procedure.

• - Exact initial weight of 10 mg steroid substance
• - Solvents used: methanol for spectroscopy, acetone for spectroscopy, tetrahydrofuran for liquid chromatography
• - Use of materials borosilicate glass, soda lime glass, polyethylene, stainless steel, silicone, Teflon
• - Uniform wetting of the surfaces with solutions of the concentrations 0.1 mg / ml and 0.01 mg / ml
• - Storage of wetted materials at room temperature for about 15 hours

Medical cotton wadding was used for the Swab tests: Cellona cotton wool, 4 cm × 3m, supplier: Rauscher & Co. GmbH, Lössnitztal, article no .: 10690 Before using the textile material, a special cleaning procedure was carried out:
The cotton wool was cut into pieces about 3 cm long and in a Soxhlet apparatus with about 300 ml of chloroform for spectroscopy (quality Uvasol ^® , manufacturer: Merck KgaA, Darmstadt, purity (GC ⁹ ) ≥ 99.0%) over a period of extracted for about 60 minutes. The cotton pieces were removed with tweezers and stored for 15 minutes in a drying oven at about 120 ° C for drying. Until use as a swab material, the cotton swabs were stored in tightly closed glass bottles.

to Purification of the wetted with active substance material surfaces was a cotton swab with 2 ml of the solvent moistened with the solutions previously and contaminated the materials. The sampling took place according to sample.

One Cotton ball was subsequently immediately in a with 10 ml of solvent, i.e. Methanol, acetone or tetrahydrofuran filled beaker and sonicated at room temperature for about 3 minutes, the resulting solution was used for HPLC content determination, the calculation of the Residual contamination took place according to equation 10.

Of the Process of wetting and the subsequent wipe test were pro Substance, solvent, material and concentration carried out 3 times, so that a sufficient statistical coverage of the received Results are given (Table 25).

recovery dependent on of equipment materials

It could be shown, with a few exceptions, that the Swab method the final rinse test clearly inferior. The recovery rates of the active ingredients were significantly lower than those using the Swab test Results obtained using the final rinse test.

2 : Recovery rates (%) of the active substances with Swab method in tetrahydrofuran for liquid chromatography of different materials, concentration: 0.1 mg / ml

123 : Recovery rates (%) of the active ingredients with Swab method in methanol for spectroscopy of different materials, concentration: 0.1 mg / ml

13 : Recovery rates (%) of the active ingredients with Swab method in acetone for spectroscopy of different materials, concentration: 0.1 mg / ml

Swab method on glass surfaces

The recoveries of the active substances using the Swab method for borosilicate glass and soda lime glass was about 90% or lower.

Surprised was that only when using methanol as flushing solvent of borosilicate glass in Compare to soda-lime glass without exception about 10% more steroid substance was found again.

According to H. Scholze, the density of a glass depends on its composition, that is, on the molecular weight of the cations in the glass network. If ions of different sizes are incorporated into the silica network, the lattice widening or shrinking occurs. The incorporation of boron into the SiO ₂ network leads to a higher density, combined with a strong polarization of the oxygen atoms. The SiO ₄ tetrahedra and BO ₃ pyramids linked together in the borosilicate glass include more or less small voids.

Sodium ions occupy more space due to their higher ionic radius and lead to a loosening of the network, this effect is even more evident in the incorporation of Ca ²⁺ ions (Table). The ionic radii in the glass always depend on the coordination of the cation in the oxygen polyhedron, thereby creating a disordered, random glass network linkage. Table shows characteristic properties of sodium, calcium and boron ions in the glass matrix.

Table 29: Coordination numbers and ionic radii of sodium, boron and calcium ions in the glass matrix

The expansion of the SiO ₂ network by alkali metal or alkaline earth metal ions, as is the case with soda-lime glasses, could be the reason why in the Swab process, active substances could penetrate into the cavities to a greater extent as a result of mechanical stress on the surfaces. However, in contrast to methanol, when using the solvents acetone and tetrahydrofuran no significant differences were found in the comparison of soda lime and borosilicate glass.

The fact that when using methanol as a rinse solvent about 10% more substance, regardless of their structure, on borosilicate surfaces in comparison to soda lime glass sticks can be attributed to the different water content in the glasses. Water is partially dissociated in the SiO ₂ network in the form of OH groups, which may also be bonded via hydrogen bonds. On the 'reactive' glass surfaces silanol groups are preferably formed, which can interact with the methanolic solvation shell of the active substances. The much higher water content of borosilicate glasses (about 400 ppm) and the associated more hydrophilic character could be the reason for the different adsorption behavior compared to soda lime glasses (about 50-100 ppm).

Also after applying the final rinse method with methanol became lower Recovery rates on borosilicate glass determined, but only for the non-polar active substances estradiol valerate and desogestrel.

When Particularly critical here is the almost 90% retention of desogestrel on Borosilikatoberflächen after to evaluate the final rinse test with methanol (concentration: 0.01 mg / ml).

capacity Final rinse procedure versus swab method

The The six active substances described were the recovery rates dependent on from the solvent used sampling method, Swab (SW) and Final Rinse Test (FR), being surfaces with solutions the concentration 0.01 mg / ml were wetted.

14 : Recovery rates (%) of estriol with final rinse and swab method of different materials, solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy, concentration: 0.01 mg / ml

15 : Recovery rates (%) of estradiol valerate with final rinse and swab method of different materials, solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy (section 0), concentration: 0.01 mg / ml

16 : Recovery rates (%) of dienogest with final rinse and swab method of different materials, solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy (section 0), concentration: 0.01 mg / ml

17 : Recovery rates (%) of desogestrel with final rinse and swab method of different materials, solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy (section 0), concentration: 0.01 mg / ml

18 : Recovery rates (%) of oxime J-867 with final rinse and swab method of different materials, solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy, concentration: 0.01 mg / ml

19 : Recovery rates (%) of carbamate J-956 with final rinse and swab method of different materials, solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy, concentration: 0.01 mg / ml. Except for some exceptions , much higher recovery rates with the final rinse procedure compared to the swab test also determined in lower concentration ranges.

exceptions:

at The use of methanol proved to be the Swab test in particular Desogestrel on borosilicate glass and polyethylene and on oxime J-867 and carbamate J-956 on stainless steel as the more powerful process, although only recovery rates of about 80% (Oxim J-867, carbamate J-956) and about 40 and 65% (desogestrel), respectively.

Independent of used sampling methods and solvents became less than 55% of the nonpolar substances desogestrel and estradiol valerate found on silicone, on this material when used of acetone and methanol the recovery rates for these both active ingredients even less than 25%.

The maximum recoveries obtained using the Swab method are generally less than 90%, compared to the final rinse method lower by more than 10%. This phenomenon could even at extreme smooth, little porous Teflon surfaces to be watched.

• – Mit Hilfe der lösungsmittelgetränkten Watte wurde nur ein Bruchteil des Wirkstoffs abgelöst, da bei dieser Methode wesentlich weniger Lösungsmittel (2 ml) im Vergleich zum Rinse-Verfahren (10 ml) verwendet wurde.
• – Die Extraktion der Wattestücke erfolgte nicht quantitativ, das heißt, die durch den Wischtest aufgenommene Substanz wurde nicht vollständig wieder herausgelöst.

The following causes can be considered:

• - With the aid of the solvent-impregnated cotton, only a fraction of the active ingredient was peeled off, since this method used significantly less solvent (2 ml) compared to the rinse method (10 ml).
• - The extraction of cotton pieces was not quantitative, that is, the substance absorbed by the wipe test was not completely dissolved out again.

Combination of final rinse procedure and swab method

Out the fact that the recovery rates determined using the Swab test generally lower in comparison to the final rinse procedure, The question then arose as to whether the remaining substance remains after completion of the Swab tests from the surfaces by means of a final rinse procedure be rinsed quantitatively could. For this attempt became arbitrary Carbamate J-956 selected as model substance. The swab procedure was on all previously described materials with methanol as solvent repeated. Subsequently The already cleaned with cotton wool surfaces were again with methanol rinsed.

The obtained solutions from both purification methods were analyzed for their steroid content by HPLC examined. To avoid random Errors were repeated twice throughout the procedure, for recovery rates Swab and final rinse test were relative standard deviations certainly.

Table 26: Relative standard deviations RSD (%) of the recovery rates of carbamate J-956 in methanol for spectroscopy (sum of swab and final rinse procedures)

4 : Recovery rates (%) of carbamate J-956 in methanol for spectroscopy with Swab method and subsequent final rinse test, concentration: 0.1 mg / ml

With the Swab method alone the active ingredient could be recovered between 70% (stainless steel) and 90% (Teflon). By subsequent rinsing with methanol were of glass, polyethylene, silicone and For stainless steel, extremely low recovery was achieved with Swab (70%), but compared to the other materials a high recovery (25%) was achieved by the Rinse method.

In Table 27 is the exclusive one recovered by the rinse method and the as Total calculated salaries from swab and rinse procedures for the Substance Carbamate J-956 shown.

Table 27: Recovery rates (%) of carbamate J-956 in methanol with final rinse test (0) compared to Swab method and subsequent final rinse test, concentration: 0.1 mg / ml

The third column (Table 27) raises the question of whether the Swab test the test substance penetrates into the material surface and therefore during the subsequent final rinse procedure can not be eliminated. The cause of the difference between the recovery rates measured by Swab or final rinse procedure also in the insufficient extraction of the cotton swab after the swab test lie. Therefore, the following tests were carried out.

Interactions Drugs - Swab Materials

For the investigation was prepared medical cotton wool.

The Cotton swabs were successively mixed with 1 ml of the substance solution Concentration 0.01 mg / ml soaked and subsequently About 15 hours at room temperature left in the air. Subsequently was the cotton wool with 10 ml of solvent, Methanol, acetone, or THF and about 3 minutes at room temperature Ultrasound treated. The supernumerary solution was used without further preparation for HPLC content determination. The process of wetting and subsequent extraction was per Substance and solvent Repeated 3 times, so that the results found as statistical sufficiently secured estimated can be (Table 28).

Table 28: Recovery rates (%) of the active substances after swave extraction, solvents: tetrahydrofuran for liquid chromatography, acetone and methanol for spectroscopy (chapter 0), concentration: 0.01 mg / ml; relative standard deviations for 3 repeat measurements (%) in brackets

All six substances could be recovered almost quantitatively from the material used The extraction was carried out according to the Swab procedure.

The Swab method could be confirmed as optimal. It has been proven that irreversible interactions between the active ingredients and the textile material can be excluded and that the examined test molecules are sufficiently stable under the described extraction conditions.

Influence of physical molecular properties on the adsorption

Polarity of steroid molecules

Of the Final rinse and swab test with their material-related diverging Results suggested the question of what influence the physical Properties of the active ingredients could have on the adsorption. each chemical compound represents a system of positive and negative Charges dar. Due to the deformability of the electron shell, the so-called polarization, also have molecules that are outward are electroneutral, a permanent dipole moment μ that defines is as a product of the distance of the charge centers l and Charge value e (Equation 11).

Equation 11: Definition of the dipole moment μ = e · l

The chemical structure of molecules, especially the presence of functional groups, is crucial for the Size of the permanent Polarity, which is described by the dipole moment, for each substance represents a substance constant.

For saturated hydrocarbons the dipole moment can be due to the high symmetry of the carbon atom neglected These substances are non-polar.

The Total polarization of a molecule with permanent dipole moment in an electric field results from the sum of the permanent and induced dipole moments, the from electron shift, shift polarization of atomic nuclei and orientation polarization. According to the equation by Clausius-Mosotti is valid for the total polarization of a molecule Equation 12.

Equation 12: calculation of the molar polarization P according to Clausius-Mosotti; ε _r - relative dielectric constant, M - molar mass, p - density

By the measurement of the relative dielectric constant by means of a Plate capacitor can over the change the capacity the molar polarization of a molecule as the sum of orientation and shift polarization are measured.

When using high-frequency electromagnetic fields, ie when exposed to light in the frequency range 4.3 · 10 ¹⁴ to 8.5 · 10 ¹⁴ Hz, dipoles can no longer align, it is measured instead of the relative dielectric constant, the refractive index n. According to Maxwell, between the dielectric constant σ and the refractive index n, there is the relationship according to Equation 13, so that with Equation 14 the molar refraction can be calculated.

Equation 13: calculation of the dielectric constant D according to Maxwell; n - refractive index ε = n 2

Equation 14: calculation of the molar refraction R according to Lorenz-Lorentz; ε ₀ - electric field constant, N _L - Avogadro constant, α _e - electron polarizability

The determination of molar refraction of steroids in solution proved problematic. In order to avoid solvent-dipole effects as far as possible, the selection of a non-polar solvent was necessary, in the literature, inter alia, alkanes, benzene and 1,4-dioxane are recommended. The solubilities of the active compounds in n-hexane ¹⁰ and 1,4-dioxane were in particular for estriol (n-hexane: 0.009%, 1,4-dioxane: 0.38%) and dienogest (n-hexane: 0.006%; 4-dioxane: 1.8%) so low that no experimental determination of molar refraction was made.

The permanent dipole moment leaves approximately calculate as the sum of the binding moments, that of a functional Group or a heteroatom can be delivered in the molecule (Table 29). However, it should be noted that the dipole moment by a Amount and direction defined, ie vectorial, size is set opposite directed binding moments can completely compensate.

The Bonding moment model has decisive disadvantages in the empirical Calculation of complex molecules in particular. So every bond goes independent with a constant value from the molecular environment into the calculation, adjacent polar bonds can affect each other so that the resulting binding moment gets smaller. Furthermore have unbalanced lying lone pairs a very large influence on the total binding moment, further are inductive and mesomeric Effects in dependence from the conformation of the molecule to take into account.

Table 29: Bonding moments (in D) of some functional groups

With The theoretical model MM3 was used for the substances estriol, estradiolvalerate, Desogestrel, dienogest, oxime J-867 and carbamate J-956 the dipole moments determined semi-empirically (Table 30). The dipole moment becomes thereby described as vector starting from the geometric center of the molecule, wherein in the calculation energies of bond lengths and angles, energies due spatial Orientation, torsional energies of bonds, Van der Waals interactions etc. enter.

The Semi-empirically determined dipole moments correlate with the retention behavior of the active ingredients in reversed-phase chromatography (Table 30). The Active substances adsorb to 'reversed-phase surfaces' the stronger, depending they are more unpolar. The retention decreases in the order aliphatic - ether, aldehydes, Ketones - Amines - Alcohols, Phenols - Carboxylic acids from. Water is the weakest mobile phase and can not handle the non-polar long alkyl chains the stationary one Phase. "It So is a relatively high proportion of organic solvent required in the mobile phase to nonpolar probes from the reverse phase surface to replace again.

Table 30: Dipole moments (D) calculated with MM3 for selected drugs; Composition of the mobile phase in RP chromatography and retention times

It The result is thus the following series with increasing polarity: desogestrel <estradiol valerate <carbamate J-956 <oxime J-867 <dienogest <estriol

It there is a direct relationship between the polarity of a molecule and the adsorption behavior on different materials.

The Retention of nonpolar substances, e.g. Desogestrel, on surfaces is compared to polar substances, e.g. Estriol, much stronger.

Table 31: Comparison of recovery rates (%) of desogestrel and estriol from different materials; Final rinse method in methanol for spectroscopy (section 0), concentration: 0.1 mg / ml

Structure of plant surfaces

Next the chemical and physical properties of the active ingredients the surface structure of the Equipment materials another moment in the consideration of adsorption represents.

The surface a solid deviates greatly from the regular spatial Arrangement of the atoms within the crystal lattice. On reason the disturbed Symmetry and the non-saturated Valences on the surface The properties of the surface electrons differ from those of the Electrons in the grid.

comes a foreign molecule is in contact with an "ideal" crystal surface give it energy, but probably no binding for energetic reasons with the surface received. Has a surface against it irregular errors On, it can interact with multiple atoms simultaneously and to hold the molecule on the surface come.

species from flaws e.g. Steps between shallow atomic layers, corners and edges on the ground incomplete Be layers or so-called dislocations.

The surface of a crystal increases continuously with the number of impurities and thus becomes more accessible for the Adsorption of gaseous or liquid Particles.

An exact characterization of the pure surface of a substrate, ie the substance on which the adsorption takes place, is only conditionally possible. With the help of the kinetic gas theory, it can be estimated that a freshly prepared surface in the atmosphere under normal pressure is covered with gas molecules within a fraction of a second. That is why it is only possible under ultrahigh vacuum conditions, pure To examine crystal surfaces.

With Using spectroscopic methods, e.g. the photoemission spectroscopy, can Analyzes of the chemical composition of the surfaces or the binding of adsorbed molecules to the substrate with X-rays or high-energy UV radiation carried out become.

So could be detected by UV photoelectron spectra of isolated and metallic palladium surfaces adsorbed benzene molecules It is shown that the energy of the Π-electrons differs for pyridine these differences could not be determined. The energy differences can be determined by the geometric arrangement of the molecules on the surface of palladium explain benzene molecules lie flat on the surface, the interaction with the Pd atoms occurs via the Π-electron system. Pyridine molecules are against it vertically, whereby over the lone pair of electrons of the nitrogen form a σ bond to the metal atoms is formed.

With It is the help of low energy electron diffraction (LEED) technology also possible, Statements about the surface structure or for the order of adsorbed layers by means of electron diffraction hold true.

at the application of field emission microscopy (FEM) and field ionization microscopy (FIM) on metallic surfaces Before and after the adsorption of foreign atoms, it is possible to use the different patterns of fluorescence on binding types and positions close.

physisorption and chemisorption on surfaces

Substances can be bound in different ways to a material surface, according to the type of binding a distinction physisorption (physical adsorption) and chemisorption (chemical adsorption). 5 : Physisorption (P) and chemisorption (C) on surfaces

Approaches a molecule a surface, it interacts with the surface atoms and a smaller one Amount of potential energy is released. In this as physisorption As indicated, the attachment to the crystal interface occurs via van der Waals interactions.

A further approach is with a greater release of energy connected, wherein between the approaching molecule and the Substrate real chemical bonds arise or within the Molecule bond breaks or deformations can occur. This process is called chemisorption. Both for physisorption as well as for Chemisorption can Adsorption enthalpies by measuring the temperature increase of a Sample of known heat capacity experimentally be determined, but so far only for gaseous substances is (Table 32).

Measured adsorption enthalpies of approx. - 20 to - 80 kJ mol ^-1 indicate physisorption.

Table 32: Chemisorption enthalpies (kJ · mol ^-1 ) on metal surfaces for some gases

adsorption processes, independently whether by van der Waals interactions or by training chemical covalent bonds are, except for minor Exceptions always associated with an energy release.

on the other hand must for the desorption, that is for the supersede of adsorbates from the surface always an activation energy can be applied that at least the Have amount of enthalpy of physisorption or chemisorption got to.

The transfer These relationships on adsorbates used in solution with surfaces in Interaction and where solvation and desolvation processes run at the same time, is only conditionally possible.

• 1. Solvatisierte Substanzmoleküle kommen mit der Oberfläche in Berührung.
• 2. Infolge dieser Wechselwirkung kann es zur Zerstörung der Solvathülle kommen, wenn die Adsorptionsenergie größer ist als die Solvatationsenergie. Das Lösungsmittel verdampft von der Oberfläche, dieser Prozess verläuft bei Raumtemperatur nicht vollständig.
• 3. Das Ablösen der Substanzmoleküle durch Zugabe von Lösungsmittel erfolgt problemlos, wenn die Adsorbate ihre Solvathülle ganz oder zum Teil noch besitzen. Bei freien Molekülen, die eine Chemisorption mit den Oberflächenatomen durch Ausbildung z.B. kovalenter Bindungen eingegangen sind, ist ein Aktivierungspotential erforderlich, um die Bindungen zu brechen und die Solvatationshüllen wieder aufzubauen.
• 4. Nur solvatisierte Moleküle können die Grenzfläche verlassen.

The complex nature of adsorption and desorption processes strongly consists in:

• 1. Solvated substance molecules come into contact with the surface.
• 2. As a result of this interaction, the solvation shell can be destroyed if the adsorption energy is greater than the solvation energy. The solvent evaporates from the surface, this process is not complete at room temperature.
• 3. The detachment of the substance molecules by the addition of solvent takes place without any problems if the adsorbates still have all or part of their solvation shell. For free molecules that have undergone chemisorption with the surface atoms through the formation of, for example, covalent bonds, an activation potential is required to break the bonds and rebuild the solvation shells.
• 4. Only solvated molecules can leave the interface.

at This simplified consideration became the surface structures and properties the substrates are not taken into consideration as a very decisive factor: Adsorption and desorption of molecules in solution.

adsorption on stainless steel surfaces

investigations for the adsorption of organic molecules on metallic surfaces already described several times in the literature. β-lactoglobulin, a protein the u.a. in the acidification process the milk is produced, has been modified by chemical modification so that due to the induced charge changes the adsorption behavior on the stainless steel surface completely changed has been.

Free carboxyl groups were chemically bound with glycinamide: - COOH → - CONHCH ₂ CONH ₂

Amino groups of the molecule were acetylated: - NH ₂ → - NHCOCH ₃

It could be proved by the humiliation of the positive Charge after acetylation of the amino groups compared the modified molecules to untreated β-lactoglobulin were not adsorbed on the stainless steel particles. Exactly the opposite was the case with the glycinamide-treated sample. As a cause for this Adsorption differences were electrostatic interactions, that is Attraction. rejection processes discussed with the negatively charged steel surface.

stainless steel alloys form a thin disordered under normal environmental conditions Chromium (III) oxide-hydroxide passive layer, which for the non-corrosive character the surfaces responsible for.

By Investigations by means of IR spectroscopy on amorphous chromium (III) oxide-hydroxide colloids, which are modeled on the chromium oxide passive layers could be used, depending on specific binding mechanisms be detected by the structure of the adsorbates. Was investigated the adsorption behavior of sulfates, orthophosphates and glucuronic acid in aqueous Chromium (III) oxide-hydroxide colloidal solutions. Sulfates could only be weak Interaction with the chromium (III) ions of the surface detected although Cr (III) ions are known to be very good complexing agents.

The situation is different for phosphates and the glucuronic acid CHO- (CHOH) ₄ -COOH. Both molecules form strong coordination compounds with Cr (III) and are deprotonated. This lowers the pH at the colloid surface so that new anions can be deposited.

experimental Investigations on defined stainless steel surfaces

to Investigation of the adsorption behavior of steroidal drugs Recovery studies were done on stainless steel with defined composition and known surface properties carried out.

There are a variety of steels of different composition, accordingly
In addition to alloyed and unalloyed steels, DIN EN 10020 distinguishes between basic, quality and stainless steels le. For the production of equipment for the pharmaceutical industry only alloyed stainless steels due to their corrosion resistance and thermodynamic properties into consideration.

alloyed Contains steel besides carbon other elements, e.g. Chrome, manganese, nickel, Cobalt, molybdenum etc. These alloy metals influence the production process Crystal lattice form of steel and one distinguishes in austenitic steels with cubic-face-centered Grids and ferritic steels with cubic body-centered grating.

The Studies within this work were done with austenitic Stainless steel materials (Table 33, Table 34).

Table 33: Chromium-nickel steel alloy, containing copper, Chemical composition in

Table 34: Chromium-nickel steel alloy, containing titanium, Chemical composition in

Both Chrome-nickel steels (Table 33, Table 34) are typical for the preparation of pharmaceutical Equipment used materials, the titanium-containing steel on Reason of his higher corrosion resistance across from acids is much more expensive. In the following text both alloys simplified referred to as copper steel or titanium steel.

The Roughness of a stainless steel surface Can be processed by methods such as mechanical, chemical and electrochemical polishing are influenced. Mechanical polishing done by grinding the metal surface with wood, glass, plastics or similar Materials using polishing agents, e.g. Metal oxides, Silicon carbide or kieselguhr. In chemical polishing, which barely applied in practice, the surface becomes with acids treated.

The smoothest stainless steel surfaces are obtained when the to be smoothed surface as an anode, and using acid mixtures As catholyte edges and tips are solved on the surface. This method is called electrochemical polishing. For all surface treatment methods becomes a geometrically ideal surface However, this ideal state is never reached. It always stays a regular or irregular deviation from a smooth surface shape, which is called roughness.

The roughness of a surface can be described by several characteristics, such as the roughness Rt, the mean roughness R _a ¹⁴ and the mean roughness R _M.

If a reference route for based on the surface inspection can be laid the sizes to the definition a surface calculated according to equation 15, equation 16 and equation 17 become.

Equation 15: Calculation of the roughness depth Rt R t = y B - y G = y OF max - y OF min

Equation 16: Calculation of mean roughness R _a

Equation 17: Calculation of mean roughness R _M

The Characteristics for the roughness a stainless steel surface can determined with the help of stylus instruments be worn, with tactile surfaces worn mechanically become. Furthermore optical, pneumatic and electrical processes are used, to get information about the macroscopic roughness of a surface in the micrometer range too receive.

at lower surface roughness provide microscopy techniques, e.g. the scanning electron microscopy, additional Statements about the nature of the stainless steel surfaces, such as the presence of the smallest Foreign particles or defects in the nanometer range.

Within This work was made of stainless steel materials by Johannes Ohlendorf GmbH Stahl, Gütersloh, used, their surface properties with the help of a Perthometer in cooperation with the manufacturer were determined. In this measurement method is against the surface to be tested placed a capacitor provided with a capacitor. The surface presents the second capacitor, so depending on the roughness the distance between the two capacitors varies, associated with a measurable change the capacity.

For the investigations, 0.8 mm thick stainless steel metal plates of size 5.6 × 5.6 cm were used, the surfaces of which were mechanically ground or electropolished on one side. The measured average roughnesses R _{a of} the individual plates (Table 35) are expected to have a relatively large variance after mechanical surface treatment.

Table 35: Mean roughnesses R _a (μm) of stainless steel surfaces; Mean values in parentheses

Solutions from Estriol, estradiol valerate, dienogest, desogestrel, oxime J-867 and carbamate J-956 in acetone and methanol at a concentration of 0.01 mg / ml were prepared. After wetting the metal surfaces, like described in section 0, the plates were about 15 hours at Stored at room temperature.

Subsequently were the substance residues with 10 ml of the appropriate solvent without mechanical contact the surfaces rinsed.

Of the Content of the substance solutions was determined by HPLC. The process of wetting and then rinsing off was for statistical protection per substance and solvent, each with 3 steel plates the same surface finish performed. The calculated relative standard deviations were independent for three repeat tests Solvent, Active ingredient and surface treatment generally less than 1%.

Mechanically polished stainless steel surfaces

21 : Recovery rates (%) of the active substances on copper steel as a function of the roughness of the surface (μm), final rinse method in methanol for spectroscopy, concentration: 0.01 mg / ml

22 : Recovery rates (%) of the active substances on titanium steel as a function of the roughness of the surface (μm), final rinse method in methanol for spectroscopy, concentration: 0.01 mg / ml

23 : Recovery rates (%) of the active substances on copper steel as a function of the roughness of the surface (μm), final rinse procedure in acetone for spectroscopy Concentration: 0.01 mg / ml

24 : Recovery rates (%) of the active substances on titanium steel as a function of the roughness of the surface (μm), final rinse procedure in acetone for spectroscopy concentration: 0.01 mg / ml

methanol

When methanol was used as the rinse solvent, recoveries of more than 90% were found for both the estrogens estriol and estradiol valerate and the progestin dienogest, regardless of the alloy composition and the surface finish of the steel plates. The decrease in recovery rates depending on the increase in the roughness of the surfaces is negligibly small with a maximum of 6%
The situation is different with the substances desogestrel, oxime J-867 and carbamate J-956, the residual contamination of these substances on steel surfaces of coarse surface quality is essential. Desogestrel, oxime J-867 and carbamate J-956 showed almost identical behavior, especially on the titanium-containing chromium-nickel steel plates, with average roughnesses of about 0.8-1.2 microns, about 20-25% of the substance adhere.

On Copper steel was about 10-25% The substances desogestrel, oxime J-867 and carbamate J-956 on rougher surfaces (0.8-1.2 microns) retained.

acetone

estriol, Estradiol valerate and dienogest are with recovery rates of more than 90% on both stainless steel alloys as uncritical.

The Use of titanium steel with roughness of more than 0.5 μm as plant material should be avoided in the preparation of oxime J-867 and carbamate J-956 become). For both substances differ in terms of their molecular structure only an extra Ethylcarbamoyl group on the oxime nitrogen of the phenyl ring (11-position) differ, became independent from the solvent used a stronger one Adsorption behavior observed on titanium-containing chromium-nickel steel, about 20-25% The substance remained on the surface with a roughness of 0.6-1.2 microns adhere.

25 : Recovery rates (%) of oxime J-867 and carbamate J-956 on titanium and copper steel as a function of the surface roughness (μm), final rinse in acetone or methanol for spectroscopy (Section 0), concentration: 0.01 mg / ml

electrochemical polished stainless steel surfaces

stainless steel materials with particularly smooth surfaces, which can only be realized by means of electrochemical methods can, are opposite mechanically ground surfaces in general to be preferred.

For all examined Substances with the exception of desogestrel were recoverable both with the use of copper-containing and titanium-containing steels more as 90%, independent of which, which rinse solvent was used.

For estriol, Estradiol valerate and dienogest is the use of mechanical ground and electropolished stainless steel surfaces equally possible, this Substances have little interaction with the ones used Materials, resulting in recoveries of more than 92% reflects.

critical See it for the substances oxime J-867 and carbamate J-956 (Table 36). at comparable average roughness of the steel plates of about 0.8 microns showed When using methanol a significantly higher adsorption, if instead the electropolished, mechanically ground metal plates were used. When using acetone, the recovery rates are on copper steel independently from the surface treatment comparable (about 90 or 97%). Both substances remained however about 25% adhere to titanium-containing mechanically ground steel surfaces, if acetone as a rinse solvent has been used.

Table 36: Recovery rates (%) of oxime J-867 and carbamate J-956 on titanium and copper steel (roughness ≈ 0.8 μm), final rinse in methanol or acetone for spectroscopy. Concentration: 0.01 mg / ml

The present results are different than the investigations of Villafranca and Zambrano.

These tested profiles of chromium-nickel steel plates in dependence from the respective surface refinement with the aim of improving the cleaning and sterilization process in the pharmaceutical industry. They could show that with the help of the electropolishing a round, smooth surface structure produced, so tips and edges were removed. investigations for the retention of sucrose on differently finished surfaces carried out with the result that lesser on mechanically treated stainless steel surfaces Roughness about three times the amount of substance remained stuck as on electropolished surfaces, although for the latter a higher one mean roughness was measured (Table 37).

Table 37: Adsorption of sucrose on stainless steel as a function of average roughness (μm)

Small mean roughnesses are not necessarily associated with a smooth surface texture. Mechanical grinding of stainless steels up to Ra values of approx. 0.2 μm always leads to corners, points and burrs which make cleaning after substance contamination more difficult compared to electrochemically polished surfaces. The associated higher adsorption was found especially for substances with large spatial extent, so Oxim J-867 and carbamate J-956 (Table 36).

in the By contrast, the recovery behavior of desogestrel gives way on both stainless steel alloys of the other examined Substances off.

Recovery of desogestrel

methanol

desogestrel remains on electrochemical when using methanol as a rinse solvent polished stainless steel surfaces, especially on titanium steel, adhere much more strongly than on mechanical sanded plates. The difference is with low roughness (~ 0.2 μm) about 40% for Titanium steel and about 20% for copper steel.

Smooth burr-free surfaces associated with a higher retention of the substances, this apparent contradiction indicates a chemisorption of the substance concerned on the stainless steel surface. Remarkably, on electropolished surfaces with lower average roughness a larger desogestrel level was retained when using methanol than on areas with higher R _a values.

26 : Recovery rates (%) of desogestrel on titanium and copper steel as a function of surface treatment, final rinse procedure in methanol for spectroscopy, concentration: 0.01 mg / ml On rough, grate-containing surfaces, desogestrel crystals can mechanically accumulate in nooks and crannies (They were also investigated.). On the relevant plate (5.6 × 5.6 cm) are still about 2.2 micrograms of substance.

On electrochemically polished stainless steel plates (5,6 × 5,6 cm) with very smooth surface structure -), on the other hand, have no crystals fixed on this surface liable a uniformly distributed Film of approx. 4.4 μg Substance caused by chemical interaction with the metal surface.

• – Desogestrel auf Titanstahl (mechanisch poliert) (Ra ~ 0,8 μm), Final-Rinse-Verfahren in Aceton für Spektroskopie (Abschnitt 0), Konzentration: 0,01 mg/ ml, 2 – Nichtkontaminierte Titanstahlplatte (mechanisch poliert); Aufnahme mit System-Mikroskop BX52 der Olympus Optical Co. GmbH im Auflicht mit Nutzung der Polarisationsoption,
• – Desogestrel auf Titanstahl (elektrochemisch poliert) (Ra ~ 0,2 μm), Final-Rinse-Verfahren in Methanol für Spektroskopie (Abschnitt 4.2), Konzentration: 0,01 mg/ ml, 2 – Nichtkontaminierte Titanstahlplatte (elektrochemisch poliert); Aufnahme mit System-Mikroskop BX52 der Olympus Optical Co. GmbH im Auflicht mit Nutzung der Polarisationsoption

It was further investigated.

• - Desogestrel on titanium steel (mechanically polished) (Ra ~ 0.8 μm), final rinse method in acetone for spectroscopy (section 0), concentration: 0.01 mg / ml, 2 - non-contaminated titanium steel plate (mechanically polished); Recording with microscope system BX52 of the Olympus Optical Co. GmbH in reflected light with use of the polarization option,
• - Desogestrel on titanium steel (electrochemically polished) (Ra ~ 0.2 μm), final rinse method in methanol for spectroscopy (section 4.2), concentration: 0.01 mg / ml, 2 - non-contaminated titanium steel plate (electrochemically polished); Image taken with system microscope BX52 from Olympus Optical Co. GmbH in reflected light using the polarization option

desogestrel has a C-H-acid Ething group as a special structural feature. It can be assumed that in the presence of polar solvents, e.g. Methanol, a coordination compound with Cr (III) received becomes. This effect will most likely by the presence of an adjacent hydroxyl group in the desogestrel forced. When using acetone as a solvent could be similar Behavior can not be observed, that is, there was no desogestive film on the metal surface.

The IR spectra of adsorbed D-glucuronic acid and a 0.25M D-glucuronic acid aqueous solution both showed known bands in the range 950-1150 cm ^-1 (CO bands). In addition, different IR bands were found for the dissolved or adsorbed carboxylic acid.

It Proof has been furnished that glucuronate anion has a Binding to the Cr (III) ions of the chromium (III) oxide hydroxide colloids took place.

acetone

at electrochemically ground stainless steel surfaces has the average roughness the surfaces the height the recovery of desogestrel has no significant influence, if as a solvent Acetone was used.

As expected on mechanically polished steel plates with lower roughness found a lower desogestrel residual contamination.

As described in Oxim J-867 and carbamate J-956, this adsorption phenomenon on the nature of the contaminated surface, the presence of dead spaces like corners and edges, be traced back.

27 : Recovery rates (%) of desogestrel on mechanically ground or electrochemically polished titanium and copper steel surfaces, final rinse in acetone for spectroscopy, concentration: 0.01 mg / ml.

contamination of active ingredients through plant materials

So far was the subject of investigation, as it was by retention of pharmaceutical active substances on plant surfaces for cross contamination can come with the product, which then in the apparatus one secondary Synthesis step or a forming operation is subjected.

In spite of validated cleaning process is always on, albeit minimal, Contamination risk exist. To absolutely exclude this, will be particularly highly effective pharmaceutical substances in so-called Single-purpose systems made. But even with the use of such "dedicated" systems remains the Question open to what extent the plant materials such as stainless steel or glass have an influence the quality to exercise the active ingredients capital.

can it due to mechanical movement in apparatus or containers to Abrasion of metal particles, there is a risk of "leaching" of materials due to rinsing processes with protic and aprotic solvents? Studies on the heavy metal content of steroid substances were carried out, to largely exclude qualitative risks.

Heavy metal content in steroids

Voltammetric content determination

Of the Content of iron and other existing in stainless steel alloys Heavy metals in steroidal drugs were determined as model substance Oxim J-867 was selected, an amorphous substance with a particle size smaller than 30 μm.

All Salary determinations were made voltammetrically and were in collaboration with the institute for Pharmacy of FSU Jena performed.

The Voltammetry is an electrochemical analytical method Determination of substances that oxidize anodically or cathodically let reduce. In this method, current-potential curves on electrodes in dependence absorbed by the concentration of the analytes. Within the voltammetric Method is used between potentiometric and polarographic Differentiated methods. In potentiometry the measurement takes place the potential change at the working electrode at constant current. The measurement of the limiting current by recording individual current-voltage curves in dependence the concentration of the analyte is called polarography, this method was used in the following studies. By means of special enrichment procedures, e.g. inverse voltammetry, Substances in the lowest concentrations could be determined. In the process, the metals or metal complexes to be determined became first electrolytically deposited at the electrode, then followed the electrochemical separation, which is associated with a measurable concentration-dependent current flow.

For the voltammetric Heavy metal was initially a wet digestion of Active substance sample with concentrated sulfuric acid and hydrogen peroxide at 300 ° C within 45 min. The resulting solution was washed with sodium hydroxide solution adjusted to a pH of 2-3 and diluted with water.

Determination of iron:

About 10 ml of the solution obtained from the wet digestion process was adjusted to pH 5 with citrate buffer. The measurement was carried out in the range between 50 mV and - 350 mV, whereby iron (III) was cathodically reduced to iron (II). The quantification of iron was carried out by Aufstockmethode.

Determination of copper:

The digestion solution was treated with ethylenediaminetetraacetic acid. After cathodic reduction of the copper complex to Cu ⁰ with an enrichment voltage of - 400 mV followed by the anodic reoxidation of Cu ⁰ at - 400 to 100 mV and the complex formation. This process was associated with a, proportional to the copper concentration, current flow. The quantification of copper was also carried out by means of an upsampling method.

Determination of nickel and cobalt:

nickel and cobalt were used as dimethylglyoxime chelate complexes with an enrichment stress from - 400 mV adsorbed on the cathode. The at the subsequent cathodic Reduction at - 800 mV to - 1400 mV measured current flow was the nickel and cobalt concentration proportional, the quantification was carried out by Aufstockmethode.

All Measurements were carried out on a voltammetry apparatus "746 VA Trace Analyzer" from the company "Deutsche Metrohm GmbH & Co. KG ", Filderstadt.

Heavy metal content in Oxime J-867

Heavy metal intake through mechanical stress

Approximately 70 g of a lot of Oxim J-867 (section 0) were shaken continuously in a 1 liter stainless steel lidded cask (Table 40) over a longer period of time. The relatively small amount of substance with a fine particle size and the comparatively large stainless steel surface (about 9 dm ² ) were chosen to accelerate contamination and to intensify the interaction between substance and container. The shaking process was continuous over about 7.5 days; Within this period of time, substance samples were taken at regular intervals and tested for their iron content. A duplicate determination was made.

Of the measured iron content of the substance samples (Table 38) as a function of by the time was negligible small, the blank value was 1 ppm, a function of the duration of the Shaking process could can not be detected.

Table 38: Iron content (ppm) in oxime J-867 after continuous movement in stainless steel containers

Heavy metal intake in dependence from the storage period

takes a pharmaceutical active substance during storage over a prolonged period particles from the container material on? The following investigation focused exclusively the question whether in substances that are defined within a Period in stainless steel barrels (Table 40) were stored, iron could be detected. The chemical composition of the steel containers used corresponded to that Chrome-nickel steel alloy (Table 33) used as material for chemical synthesis equipment is used. The determination of Heavy metals copper, nickel and cobalt were also made as they are as ingredients or trace elements in the steel alloys available.

template of the same substances stored in brown glass containers (Table 41), have been confirmed of the blank value.

Oxime J-867 - Samples were stored for extended periods at room temperature and protected from light in both container materials (Table 39). Subsequently, these samples are in terms of their Heavy metal content, as described in Section 0, has been tested voltammetrically.

Table 39: Storage conditions and storage time of oxime J-867 - Samples

Table 40: Chemical composition of stainless steel containers (%), container volume: approx. 1 l,

Table 41: Chemical composition of brown glass bottles (%), volume: approx. 20 ml,

Of the measured iron content was negligible in all samples small.

The in stainless steels Existing elements of copper, nickel and cobalt could be found in all tested substances are practically not detected.

Table 42: Content of iron, copper, nickel and cobalt (ppm) in oxime J-867 samples after storage in stainless steel containers or amber glass containers

Information about the Interaction of active pharmaceutical ingredients and medicinal products with materials in which chemical and biological syntheses, cleaning and shaping operations, tabletting, and coating processes and other procedures take place are of great importance. Avoidance or reduction of possibility Cross-contamination to a safe minimum is an essential one Part of quality assurance in the pharmaceutical industry.

Analytical methods were developed with which steroidal drugs in the trace range up to 10-100 ng / ml could be detected, well below the required acceptance limits. The correctness, reproducibility, linearity and robustness of these liquid chromatographic methods were demonstrated within a validation program. Due to the fact that pharmaceutical agents are highly pure substances, it has not been necessary to develop selective methods with respect to synthesis by-products and degradation products. Based on the characteristic impurity profiles, it has been shown that at active substance concentrations of 0.1 mg / ml to 0.01 mg / ml potential by-products are below the analytical grade Detection limit occurred and thus were no longer detected. The selection of the model substances was carried out according to characteristic structural features for this substance class, so that generalized statements can be made for other estrogens and gestagens.

Further Investigations for the recovery of steroid drugs on materials used in pharmaceutical manufacturing come, like glass, stainless steel alloys and plastics, so silicone, Teflon and polyethylene.

To Approx. 1300 recovery attempts could be made regarding the availability of Plant materials and solvents general statements are formulated. There were interesting ones Options, the extent of interactions between drug molecules, solvents and predict reactor materials or their surfaces or for each substance to find the best combination (Table 43).

Table 43: Optimum combination of plant materials, solvents and substances to minimize cross-contamination after cleaning

The Results of the recovery studies show that non-polar drugs, e.g. Desogestrel, because of interaction of these substances with the material surfaces are preferably to produce in Einzweckanlagen to sub-mixtures to avoid with secondary products. Is this from the pharmaceutical manufacturer from economic squeeze not realizable, is mandatory in the first consecutive batch of a other active substance with a validated method of analysis on the Pre-product to be tested. With this method, the predecessor substance must as an impurity selectively with the appropriate sensitivity be detected.

The Both sampling processes were Final Rinse Procedure and Swab Test compared in terms of their efficiency.

In of pharmaceutical drug manufacture is the final rinse procedure for evaluation the residual contamination, the Swab method should be avoided. The found using the Swab method compared to the final rinse test Lower recoveries did not reflect any representative results Image of the residual substance residues on the material surfaces.

The Fact that by reapplying the final rinse test after the Swab operation not achieved by rinsing once Results have been achieved to conclude that the mechanical stress of the to be cleaned surfaces accompanying Swab test for penetration of the residual substance in Defects and deeper intermediate spaces of the materials led.

Out The general application of the final rinse procedure also results a high economic benefit. This results on the one hand from the avoidance of disassembly of the Plant parts and subsequent Re-composition after cleaning. Furthermore deleted when using the final rinse method the elaborate for the swab test characteristic extractive sample preparation for analytical Purity test.

At the Use of plastics as hose or composite materials is Teflon clearly favor the retention of all of them without exception investigated drugs were negligible with less than 10% small. The use of silicone as a product-contacting material must be considered extraordinary be assessed critically, especially non-polar substances such as desogestrel and estradiol valerate make a strong interaction with this material. Whether silicones only up to a certain saturation Take up substances probable, however, was not investigated in this study.

polyethylene behave opposite nonpolar molecules similar to Silicone, such interactions are already several times in the literature described and considered particularly critical. So penetrate e.g. Active ingredients from infusion and injection solutions in polyethylene surfaces and to lead to a not to be overlooked Concentration reduction. Therefore, polyethylene should be considered as product-touching Material not used in multi-purpose pharmaceutical plants come.

It it could be shown that non-polar molecules adsorb more strongly to glass surfaces, if polar solvents such as methanol solvation and thus the adsorption by training stable exothermic hydrogen bonds to the silanol groups accelerate the 'glass grid'

The Studies on the retention behavior of nonpolar substances Borosilicate and soda lime glass, in particular when using polar solvent showed significant differences. Borosilicate glasses have due to their higher water content, in comparison to soda-lime glasses is this about twenty times, more terminal silanol groups, with the solvated molecules interact. Nonpolar molecules adsorb more strongly on borosilicate glass surfaces The reason should be soda lime glasses as reactor materials clearly favored. Nevertheless, as a result of important excellent properties, such as optimal water resistance and low expansion coefficients, almost exclusively borosilicate glass materials used in the pharmaceutical industry.

stainless steel alloys with smooth little rough surfaces are to be given priority as reactor material in general. however could be shown within this work that mechanically ground materials with average roughness up to about 0.4 um in place of the essential more expensive electrochemically polished stainless steel surfaces Reason for the low interaction between active substances and steel materials can be used.

The Type of surface finishing up to a defined roughness thus has no influence on the quality of pharmaceutical products resulting from this knowledge High cost savings in the construction of new synthesis plants is essential.

substances due to their molecular structure in the presence of polar solvents act as 'anions' can (Desogestrel), become very smooth, with the help of electrochemical Methods generated surfaces Adsorbed, it could the formation of an amorphous surface film be detected. For active ingredients with this structural feature should only be the most cost-effective mechanically ground reactor materials with average roughness up to approx. 0.5 μm be used.

Aauch was the influence of stainless steel materials on the quality of active substances examined with a model substance. Heavy metals from the metal matrix were in the substance samples only in negligible concentrations detected near the blank value. For everyone at Jenapharm GmbH manufactured active ingredients applies according to company specifications the limit for Heavy metals of maximum 0,002% (= 20 ppm), tested after Pharm. Eur. Method 2.4.8, method C for Dienogest, desogestrel, estradiol valerate and estriol and according to USP method <231> II for oxime J-867 and carbamate J-956. This value was also reached almost in no sample. Thus, in addition to the production, storage and transport of steroidal active substances in stainless steel containers safely.

The Ensuring that through manufacturing and cleaning processes, storage, Transport and influence of the staff does not affect the quality of the products has been solved according to the invention demonstrated how cross-contamination is avoided or at one for humans acceptable level reduced can be. This requirement arises in the pharmaceutical active ingredient and finished goods manufacture one of the most important components of quality control represents.

Claims (1)

Method for reducing cross-contamination between active pharmaceutical ingredients and ver impurities in production plants, in particular for steroidal active substances, characterized in that by appropriate investigation methods steroidal agents in a range up to 10-100 ng / ml are determined and that at a steroidal active substance concentration of 0.1 mg / ml to 0.01 mg / ml of analytically negligible by-products occur.