GTP Risk Checker

Detailed Evaluation

Summary
Your Titration risk is high

Your main risks are:

The temperature change in your lab is moderate

Both the pH value and the volume of the titrant change with temperature, which can
affect your titration results.
Your system has not been professionally installed and no verification has been
performed. You can not be sure if the instrument is working properly.
You have not defined a periodic sensor calibration or test and therefore can not be sure
that the sensors are in good shape and measuring correctly
This can have a negative impact on result quality and reliability.
You have not defined a periodic titrant standardization, and therefore can not be sure
that the system is using the correct titrant concentration
The system using this incorrect value will have a negative impact on reliability of results.
When weighing in the sample, any error will immediately affect the titration result. Check
your weighing risk with the balance Risk Checker (Link IN NEUEM FENSTER öffnen:
http://www.good-weighing-practice.com/gwp/risk-check-2)
Your laboratory personell are professional and experienced; however, manual titrations
bear the risk of systematic and random errors.
Hand written results bear the risk of transcription errors
Depending on your environment, it's not assured that this kind of documentation fulfills
regulatory requirements.
Your laboratory staff has not been trained on the instrument, and untrained staff are
more prone to errors
This can have a direct impact on the results.

Take the measures indicated below to reduce your titration risk. METTLER TOLEDO is ready
to support you in your efforts with knowledge in Good Titration Practice and professional
instruments and service products. Contact your local METTLER TOLEDO titration expert or
visit www.mt.com/gtp
Question 1 Temperature
Question: How much does the temperature fluctuate in your lab?

Your Answer: between 2°C / 4°F and 4°C / 7°F

Your risk: medium

The temperature change in your lab is moderate. Both the pH value and the volume of
the titrant change with temperature, which can affect your titration results.

Background: The effects of temperature on the results

Temperature can have two different effects on a titration. The first becomes apparent
when performing an endpoint titration to a predefined pH value. The pH of the sample
depends on the degree of dissociation of the acids and bases present, which is
temperature dependent. A temperature change therefore gives rise to a real change in
the pH of the sample that cannot be accounted for without knowing the sample's exact
composition. A further effect concerns the actual measurement process and the slope of
the electrode calibration curve. The slope of the calibration curve is temperature
dependent so it is important to either perform the analysis at the same temperature at
which the electrode was adjusted, or to measure both temperatures and compensate for
the change in slope of the calibration curve. Fortunately, most modern instruments are
able to simultaneously measure the sample temperature and automatically compensate
for this error.

The second major effect that temperature has on the titration is related to the density, and
therefore the concentration of the titrant. For example, if the concentration of the titrant
was determined in the morning with a laboratory temperature of 20°C and samples are
titrated in the afternoon at a temperature of 30°C, this 10°C change in temperature can
cause a significant reduction in titrant concentration, affecting titration results. This is of
particular significance with non-aqueous titrants where the coefficient of thermal
expansion is much higher. Below is a table showing typical errors:

Error in concentration due to temperature change

Titrant % Error per °C change
0.1 M NaOH 0.027
1 M NaOH 0.036
0.1 M HCl 0.026
1 M HCl 0.029
1 component 5 mg/mL Karl Fischer Reagent 0.092
There are 3 possibilities to minimize this error. The best solution is to maintain constant
temperature in the laboratory. If the titrant concentration and samples are measured at the
same temperature then there is no error. Unfortunately, this is usually not feasible. The error
can be reduced by re-standardizing whenever there is a significant change in temperature.
The final possibility is to measure the temperature of the titrant and to correct for the change
in concentration with a correction factor. In the case of Karl Fischer reagent in the above
table, the correction factor is calculated as follows:

Factor, f = 1 + (Ttiter - Tsample) x 0.092/100

With a 10°C difference in temperature (Ttiter - Tsample = -10) this would result in a correction
factor of

f = 1 - 10 x 0.00092 = 0.9908

Multiplying raw results by this factor will account for the error.

Recommendations
If possible, place the instrument in a temperature-controlled environment. Air condition
prevents temperature shifts and therefore minimizes temperature effects on titration
results.
Do not expose the instrument to direct sunlight. Place it in a protected area.
Measure the temperature when performing pH measurements or End Point titrations.
The documentation indicates which results are comparable. METTLER TOLEDO offers
pH sensors with integrated temperature probes for this purpose.
Use certified pH buffers for sensor calibration and apply their temperature table.
METTLER TOLEDO titrators and pH meters automatically offer this possibility.
Perform titrant standardization when the temperature changes significantly. The titrant
standardization also eliminates changes in concentration due to temperature
fluctuations in the lab. For volumetric Karl Fischer titration standardization, twice a days
is recommended if the temperature in the lab fluctuates.
Measure the temperature of the sample solution and correct the titration results for the
titrant's volume change. This is easily done using a METTLER TOLEDO Excellence
titrator. The temperature can automatically be measured using a PT1000 temperature
probe, and the flexible calculation can calculate the correct concentration according to
the above formula.
Coulometric Karl Fischer titration generates the titrant directly in the solution by
applying a current, and is therefore temperature independent and doesn't require
standardization.
Question 2 Installation
Question: How was the titration system installed?

Your Answer: I installed it together with my colleagues

Your risk: high

Your system has not been professionally installed and no verification has been performed.
You can not be sure if the instrument is working properly.

Background: Instrument Installation

One goal of titration analyses is to get accurate and precise results quickly. Neglecting the
smallest thing can have an impact on the reliability and quality of the final result.
Professional installation is fundamental to every analytical system. Depending on the
requirements, the right level of measures needs to be applied. These individual
requirements encompass the following:

GLP: Quality of planning, performing, controlling and reporting of laboratory work

Certification: Quality of the instrument and of the measured values obtained by this
instrument

Validation: Quality of the analytical method and thus of the achieved results

Qualification: Quality monitoring over the lifetime of the instrument

GLP
International bodies have elaborated a set of rules concerning the analytical work in labs in
order to achieve a standard regulation recognized and accepted in all countries of the world.
These rules are commonly known as GLP rules (Good Laboratory Practice). GLP is a formal
framework for testing chemicals and consists of 10 specific rules. The goals of GLP are:

Quality assurance of test data

Mutual acceptance of data
Avoiding multiplication of tests

In particular, avoiding the multiplication of tests saves time and money. For instance, a
customer doesn't have to repeat the same test that was performed by the producer at the
end of the production process.

Certification of an automatic titrator

The certification is a check of the instrument in order to verify that the technical specifications
are fulfilled, or to verify if the actual specifications meet the required level.

Certification is only part of a list of measures to guarantee correct results. Other points that
have to be considered are system suitability tests, sensor calibration, titrant standardization,
and method validation. A complete certification of the titrator can be performed by testing
the sensor input and amplifier and the dosing accuracy (burette motor drive and glass
burette).
Validation of Titration Methods
While the goal of the analysis is to get correct results, very often the 'correct' result is not
necessarily the 'true' result. As a result, our goal is in fact to get the best possible result.
This means a result as accurate, precise and true as possible. To do this it's important to
critically investigate the factors affecting each of these things and minimize the negative
influences.

Qualification
Quality management requires the documentation of the performance over the lifetime of the
instrument, i.e., from the project phase through manufacturing, installation and operation
through disposal of the instrument.

All these steps are resumed in the comprehensive concept of Qualification:

Specification Qualification (SQ): Requirements, Functions, Design, HW/SW

Construction Qualification (CQ): Production control for each product
Design Qualification (DQ): Selection of correct instrument for intended use
Installation Qualification (IQ): Evidence of correct installation at customer's facility
Operational Qualification (OQ): Evidence and compliance to specifications, SOPs, initial
calibration, user training
Performance Qualification (PQ): Periodical calibration and certification
Maintenance Qualification (MQ): Definition of preventive maintenance and
calibration/certification intervals

Recommendations
Have the titration system installed by the instrument manufacturer's specialist.
Perform a General System Suitability Test, proving that the analysis is performed
according to the specifications.
Specially trained METTLER TOLEDO service engineers are able to perform the
calibration and certification of the titrator hardware with specific calibrated and certified
tools (CertiCase, Excellence Test Unit), ensuring traceability to international standards
The declaration of System Validation states that the titrator was developed and
manufactured in accordance with a strict quality management system.
IPac: service product that offers an initial equipment qualification including installation
and operational qualification (IQ/OQ) of the titrator at the installation site, and the
corresponding qualification documentation
EQPac: service product that offers complete qualification including full documentation of
the instrument's history.
Question 3 Sensor Performance Verification
Question: Have you defined the frequency of the sensor calibration and sensor test?

Your Answer: No, there is no sensor calibration period defined

Your risk: high

You have not defined a periodic sensor calibration or test and therefore can not be sure that
the sensors are in good shape and measuring correctly. This can have a negative impact on
result quality and reliability.

Background: Sensor Performance Verification

In pH endpoint titrations, critical factors affecting result accuracy are the slope and zero point
of the sensor. Both of these parameters are used to convert the raw mV signal from the
sensor into the pH of the sample solution using the Nernst equation:

, where

E = measured signal (in mV)

S = sensor slope = -2.3 RT/nF

pH0 = sensor zero point = E0/S

pH calibration curve

Since sensors are the actual 'measurement devices', they have a very large influence on the
result of any titration. Several factors contribute to their behaviour:

Response: If a sensor is sluggish because of old age or bad maintenance, the mV reading
detected by the sensor will lag behind the 'true' value. In general, this causes titrant
consumption to be too high, and the results will be falsely high.

Calibration: Sensor calibration is particularly important for endpoint titrations. Titrant

addition is stopped when a precisely defined measurement value is reached and the titrant
consumption up to this point is used for the calculation of the result. The accuracy of the
measurement value is directly related to the determined amount of the analyte content in the
sample. This also means that an accurate measurement value, possible because of a good
adjustment, gives an accurate content determination (applicable to pH and ion selective
electrodes only).

Every maintenance procedure performed on a sensor (cleaning, regeneration, etc.), requires

that the sensor be calibrated again.

Reference electrode: Another important factor is the reference electrode. It provides a

stable reference signal against which the measurement signal is determined, so it must be
inert and provide a stable potential.

Conditioning: The ion sensitive membrane of any pH or ion selective electrode needs to be
conditioned before the sensor can be used. If not, its response will be sluggish.

Additionally, when a pH sensor is used for extended periods in non-aqueous media, the
sensor needs to be conditioned before its next use to restore the hydration layer, which is
essential to a pH sensor's function.

Recommendations
Define a sensor calibration frequency to make sure that it's measuring correctly.
Depending on temperature fluctuations and measurement conditions, the sensor
calibration should initially be performed at least once per day. The results need to be
monitored, and when it's proven that the calibration results don't change, the frequency
of the calibration can be extended to every other day or eventually to once a week.
In stead of frequent calibrations, a sensor test can be performed to check if the sensor
is still functioning correctly without changing the calibration data. The sensor test also
checks the signal drift over one minute, which is an important indication of the
response time and signal stability.
Take the appropriate measures to automatically remind users to calibrate, preventing
use of a non-calibrated sensor. METTLER TOLEDO Excellence titrators offer the
functionality of monitoring sensors' life span as well as usable life. When a calibration
needs to be performed or the sensor needs replacing, the user is automatically
reminded. A sensor can even be blocked from use if the setting is chosen accordingly.
Define acceptance limits for the calibration results. All METTLER TOLEDO titrators
allow these limits to be set in the titration method. If the limit is exceeded, the user is
prompted and the calibration data is not saved.
When not in use, sensors should always be stored in electrolyte, never deionized water,
and in the majority of cases, not dry. The exception to this is with some ion-selective
electrodes, which can be stored dry.
Question 4 Titrant Performance Verification
Question: Have you tested and defined the frequency of the titrant standardization?

Your Answer: No, there is no titrant standardization period defined

Your risk: high

You have not defined a periodic titrant standardization, and therefore can not be sure that
the system is using the correct titrant concentration. The system using this incorrect value
will have a negative impact on reliability of results.

Background: Titrants
The concentration of a titrant needs to be known accurately to be able to determine the
content of analyte in the sample solution. If the titrant concentration is unknown or
inaccurate, there will inevitably be a degree of error. From the determined titrant
concentration and the nominal titrant concentration, a titer value is calculated. The titer is
the ratio of 'determined concentration / nominal concentration' and is generally close to 1.

For titer determinations, a primary standard is preferable. A primary standard is a substance

that reacts with the titrant in a known ratio and also is extremely pure, very stable, has no
hydration water, and a high molecular weight. All these properties ensure that an accurate
determination of the titrant concentration can be performed. Examples of primary standards
are tris-hydroxy amino methane (THAM) for acids, potassium hydrogen phthalate (KHP) for
bases, NaCl/KCl for argentometric titrations and potassium dichromate for redox reactions.
Further information about primary standards for specific titrations can be found in regulatory
documentation, e.g. NIST Standard Reference Materials (SRMs).

Titrants can deteriorate over time through various external influences like oxidation,
precipitation, carbon dioxide absorption or UV degradation. We can account for some of
these by using drying tubes filled with NaOH on carrier material (CO2 absorption in alkaline
titrants) or with brown bottles (light protection), but this isn't always possible and often also
not completely effective. The titrant deterioration changes its concentration, and the rate of
deterioration determines how long it can be used before re-standardizing. This is called the
“usable life' of the titrant. After the usable life has expired, a new titer determination should
be performed before running subsequent analyses In most titrators, the usable life of a
specific titrant can be defined and standardization intervals can be enforced before being
allowed to continue.

The lifespan of a titrant is the time after which a titrant should be replaced. The lifespan is
different for every type of titrant. For example, acids are more stable than alkaline titrants
and can have a life span of up to a year, compared to 6 months for a base. During the life
span, regular standardization should be performed to guarantee reliable results.

Recommendations
Define a standardization frequency to make sure that the system is using the correct
titrant concentration in its calculations. The frequency depends on the system, the
application and on the requirements. The ideal frequency for a specific system can be
evaluated by performing a daily standardization and monitoring the results. If the titer is
stable over a long period, a standardization frequency of every other day or even once
a week can be applied.
Define a titrant life span.
Take the appropriate measures to automatically remind users to perform a
standardization, preventing them from using a titrant with an exceeded usable life.
METTLER TOLEDO Excellence titrators offer the functionality of monitoring titrants.
When there is a scheduled standardization, or the titrant is scheduled to be replaced,
the user is automatically reminded. A titrant can also be blocked from use if the setting
is chosen accordingly.
Define acceptance limits for the titer determination. All METTLER TOLEDO titrators
allow these limits to be set in the standardization method. If the limit is exceeded, the
user is prompted and the new titer value will not be saved.
Prepare and store titrants with care. This is of particular importance with titrants that are
unstable over time. In the case of alkaline titrants such as sodium and potassium
hydroxide, it's important that they are prepared with water or solvent free of carbon
dioxide, and that they are protected from atmospheric exposure to carbon dioxide. The
latter be done by attaching a drying tube containing an absorbent (e.g. NaOH on a
granular carrier) to the titrant bottle. If no measures are taken, CO2 is absorbed into the
titrant and it reacts with the base to form carbonate. The titrant becomes a mixture of
bases and it becomes impossible to obtain correct results. An alternative is using
commercially available titrants in a 10 or 20 liter cubic container that are protected from
the atmosphere.
Titrants such as iodine, permanganate and dichromate are light sensitive and need to
be protected by storing them in brown glass bottles in the short term, and by storing
them in the dark in the long term. When it comes to handling, one of the more critical
reagents is Karl Fischer. It needs to be protected from light as well as from the ingress
of atmospheric humidity. This is done by attaching a drying tube containing silica gel or
molecular sieve to the titrant bottle.
Question 5 Application: Sample size
Question: How do you measure the sample size?

Your Answer: Sample is weighed with a balance

Your risk: medium

When weighing in the sample, any error will immediately affect the titration result. Check
your weighing risk with the balance Risk Checker (Link IN NEUEM FENSTER öffnen:
http://www.good-weighing-practice.com/gwp/risk-check-2)

Background: Sample Size

By far, the biggest source of random errors resulting in precision problems is sample
handling. These errors include inhomogeneity of the sample, sample storage problems,
incorrect sample size, weighing errors, and careless handling. Critical in most cases is the
sample size.

The sample should be large enough to ensure that it is representative, but it shouldn't be so
large that repeated burette fillings are necessary during the titration. The ideal titration
should give a titrant consumption of 30 to 80% of a single burette volume. If inhomogeneity
requires the sample to be larger than this, then a more concentrated titrant should be used.

On the other extreme, the sample should be large enough so that weighing or sample
measuring errors are kept to a minimum. Here, a suitable balance must be used to ensure
that the sample size exceeds the minimum weight of the balance. This minimum weight is
defined as the weight which when measured tenfold, results in a repeatability of less than a
certain pre-defined value; for example, United States Pharmacopoeia (USP) states a value
of less than 0.1%.

In the example of standardization of KF titrant with disodium tartrate dihydrate, a small

sample of around 50 mg is required due to its limited solubility in methanol. In this case, to
ensure a weighing uncertainty of less than 0.1% a balance is needed with a readability of at
least 0.01 mg.

In the case of liquid samples, the appropriate grade of glassware or volumetric apparatus
must be used and care must be taken to avoid handling errors such as parallax.

The sample volume that is used for any titration should be sufficient to cover the sensor's
active parts (junction and sensing membrane or metal ring). Normally, this volume should be
around 50 mL to be able to fit the stirrer, titration tubes and dosing tubes in with the probe.
For smaller sample sizes, special micro-titration beakers can be used.

If the sample is a liquid, it can be added with a pipette. Higher accuracy, precision and
reliability is achieved with automated sample aliquotation. Here, a liquid handling system is
controlled by the titrator. It aspirates the liquid sample and doses a predefined volume into
the titration beaker. This reduces the systematic error and minimizes random errors.

Be aware of electrostatic influences on powdered samples. If the titration beakers are

electrostatically charged, the powdered samples typically adhere to the walls of the beakers
and might not be accessible for titration, leading to incorrect results.

Recommendations
Make sure that you're using the needed balance resolution for your sample size. The
accuracy and precision of the sample size must be smaller than the expected accuracy
and precision of the titration result.
For solid samples, use a calibrated analytical balance with 0.1 mg or 0.01 mg
readability. Check your sample size measuring risk with the GWP Risk Check on
www.mt.com/gwp.
Reduce electrostatic charge of the sample beaker by letting plastic beakers stand alone
for a moment or by using glass beakers.
For liquid samples use a calibrated high quality Rainin® pipette.
For liquid samples that require high accuracy and precision, use an appropriate liquid
handling system. The METTLER TOLEDO Liquid Handler is fully controlled by the T90
or T70 Excellence Titrators, and doses sample aliquots with an accuracy of 5uL and a
precision of < 0.1% over a big volume range. It fulfils the ISO8655-3 standard.
Choose a sample size that consumes 30 to 80% of the burette's volume. The sample
size, titrant concentration or burette size can be changed to reach this goal. METTLER
TOLEDO Compact Karl Fischer titrators calculate the optimum sample size based on
the currently used burette, titrant and estimated water content.
Dilute the sample with an appropriate solvent volume to make sure the active parts of
the equipment are covered. If necessary, use micro-titration equipment.
Question 6 Application: Instrumentation
Question: What type of instrument are you using?

Your Answer: Manual titration with a glass burette

Your risk: high

Your laboratory personell are professional and experienced; however, manual titrations bear
the risk of systematic and random errors.

Background: Instrumentation
An autotitrator allows the automation of all operations involved in titration: titrant addition,
monitoring of the reaction (signal acquisition), recognition of the endpoint, data storage,
calculation and results storage.

Automated titrators follow a defined sequence of operations. This sequence is performed

and repeated several times until the endpoint or the equivalence point of the titration
reaction is reached. All important steps of the titration are performed with higher accuracy
and precision when using an automated titrator in stead of a manual burette and color
indicator or pH meter.

User dependency is also minimized as the dosing speed, end point detection and curve
evaluation are performed following mathematical procedures in a repeatable way.
Influences that impact the user's decisions on these parameters are excluded.

Sample changers further eliminate user dependency, as sample beaker replacement and
equipment cleaning and conditioning steps are automated and user independent. In
addition, the user's time is freed to perform other tasks.

Sample preparation can be a source of error. If it's performed automatically using the
appropriate titrator accessories, the repeatability is enhanced further.

Recommendations
An autotitrator significantly improves accuracy, repeatability and reliability of results.
Titrant dosing, end point recognition, calculations and data transfer are automatic and
user independent. All METTLER TOLEDO titrators offer these advantages over a
manual burette.
Using a sample changer lowers the risk of sample carryover and standardizes cleaning
procedures. METTLER TOLEDO offers a complete portfolio. The simple Rondolino
automated titration stand has 9 sample positions and allows fixed rinsing and
conditioning steps. The flexible Rondo line allows 12 to 60 samples with sample
volumes between 600 mL and 20 mL together with advanced, flexible rinsing and
conditioning procedures. Most samples are analyzed in a series using the Sample
Handler with 42 to 80 sample beakers for direct titration.
The appropriate accessories automatically perform sample preparation steps, further
reducing manual errors and user dependence. Sample heating, cooling, dilution,
stirring, homogenizing, aliquotation or lid handling can all be automated.
Question 7 Documentation
Question: How are the results documented?

Your Answer: Results are hand written

Your risk: high

Hand written results bear the risk of transcription errors. Depending on your environment,
it's not assured that this kind of documentation fulfills regulatory requirements.

Background: Documentation
A result is only as reliable as its documentation. GLP has clear requirements for
documentation to ensure results are reliable and comparable. This can be hand written into
a lab journal, automatically printed or electronically recorded.

GLP compliant printouts must be collected and stored. They include all relevant data and
don't bear the risk of transcription errors or accidental incompleteness as hand written
records would. If titration curves, measured value tables or used resources are needed, they
can also automatically be printed, and printouts can be signed if an approval sequence has
been defined.

If many samples are analyzed, it results in great amounts of data, and if the approval
sequence is complicated, electronic records are more favorable. All results are collected in a
safe database and always ready for inspection or further evaluation. Data can automatically
be sent to a superordinate system without possibilities of transcription errors.

Full electronic traceability, security and documentation is needed in order to fulfil the FDA 21
CFR part 11 regulation for electronic records for submission, inspection and archiving
purposes.

In general, the documentation needs to fulfil the laboratory's requirements. Too much
complicated documentation is a waste of effort, and too little can result in problems with
regulatory inspections or in a need to repeat analyses.

Recommendations
Check or define your laboratory's documentation requirements.
Based on these requirements, evaluate which is the most efficient way to fulfil them:
Manual writing, automatic paper printouts or electronic records
All METTLER TOLEDO titrators can be connected to a compact dot matrix printer for
basic GLP compliant print outs, and all methods can be printed out.
If measured value tables, titration curves or many sample data/results are printed, a
standard 8 ½ x 11” or DIN A4 printer can be used with METTLER TOLEDO titrators.
Excellence and Compact titrators even print out fields for paper signing procedures.
LabX light titration PC software collects all titration results, measured value tables and
curves in a database that is always accessible for inspection. Instrument control,
method development, sample series handling and many other features are additional
benefits of using LabX light titration PC software.
Full support of compliance with FDA 21 CFR part 11 is offered by LabX pro. User
management with different user rights, audit trail, histories and electronic signature
make sure the auditors find what they are looking for. Additionally, users save time by
not having to deal with documentation.
Question 8 Human Influence
Question: Has your laboratory staff been trained on the instrument?

Your Answer: No

Your risk: high

Your laboratory staff has not been trained on the instrument, and untrained staff are more
prone to errors. This can have a direct impact on the results.

Background: Human Influence

When a machine can perform a repeated task, it's normally more repeatable than a human
being as many influences are more constant. However, human beings still have to run the
laboratory, carry and organize samples, start analyses and make many decisions of which
the instruments are not capable. Lab technicians need to be trained to make the right
decisions and to operate the instruments correctly.

In general, METTLER TOLEDO titrators and analytical instruments are optimized for simple
and efficient operation. With the One Click® concept, the user performs very few operations.
If the tecnician can operate one One Click® instrument, he can also operate the other One
Click® instruments - Karl Fischer titrators, Density meters, Refractometers and Melting Point
instruments. When installing an instrument, the users still need to be trained to take best
advantage of the instrument. Well trained analysts help ensure that the instrument is being
used correctly. Periodically, new staff needs instrument training, also.

Recommendations
Make sure the analysts have been professionally trained on the instrument. The
instrument qualification (IPac or EQPac) includes a basic user training. For a
comprehensive training containing practical exercises, EduPac is available.
Have new staff trained on the instrument. Experienced technicians can pass on their
knowledge. METTLER TOLEDO's EduPac is also valuable in refreshing the knowledge
of existing co-workers or in training new staff.

Mettler-Toledo AG, Analytical

CH-8603 Schwerzenbach
Switzerland

Subject to technical changes

©10/2010 Mettler-Toledo AG
SBU Anachem