Calibration and Adjustment in Thermal Analysis

 
Slide 0: Calibration and Adjustment in Thermal Analysis

Ladies and Gentlemen

Welcome to our seminar on calibration and adjustment for thermo-analytical instruments.

Slide 1: Topics

In this seminar, I would like to explain why regular calibration and adjustment are important during the lifetime of your instrument.

• I will provide background information.
• I will describe the simple workflow for calibration and adjustment.
• I will show you the different adjustment possibilities.
• Finally, I would like to present a unique concept from METTLER TOLEDO: FlexCal®, the database for adjustment parameters.

Slide 2: Background information – Why Calibrate and Adjust Your Instrument?

Let’s talk about your motivation for calibration and adjustment.

Generally speaking, each instrument comes fully adjusted from the manufacturer.

So, you might ask yourself, “Why should I calibrate (and adjust) my instrument then?”

With time, every high precision instrument suffers from wear and tear.

This has an impact on its performance and, of course, on the reliability of the results.

Therefore, it is highly advisable to schedule regular calibration routines as part of a basic maintenance plan to uncover any deficiencies.

The goal of calibration and adjustment is to provide a measuring system that always delivers reproducible and accurate results.


Slide 3: Background information – Definitions: Trueness, Precision, Accuracy (1)

Before I move on to the actual calibration procedure, I need to clarify a few terms. One set of terms deals with statistics and another set with calibration and adjustment.

We need to be familiar with these terms first in order to understand what we are trying to achieve.

What are trueness, accuracy and precision? These three terms deal with the distribution of results obtained during a set of measurements.

Trueness is a parameter that describes the closeness of agreement between a mean value obtained from a large number of test results and an accepted reference value, also known as the “true” value.

Accuracy is a parameter that describes the closeness of agreement between a test result and an accepted reference value.

Precision is a parameter that describes the spread of data around a mean value.

If we look at the figure on the slide, the measured values should lie as close as possible to the true value and the standard deviation, which describes the precision, should be as small as possible.

The next slide illustrates this concept in the form of a bull’s eye.


Slide 4: Background information – Definitions: Trueness, Precision, Accuracy (2)

Only in picture A do we get accurate results: The spread of data is small and the data points are close to the center of the bull’s eye. This is what we want to achieve if we want trustworthy results.

The pictures B, C and D represent undesirable results: we certainly have a mix of random and non-random “hits”, but the results are definitely not acceptable.

In the next section, I will explain how we can achieve the best results as shown in picture A.


Slide 5: Background information – Definitions: Calibration and Adjustment

In the previous section, we talked about reliable results from a statistical point of view.

We want to be able to rely on the data that the measuring instrument has delivered. But how can we achieve this goal?

We must invest a little bit of time and effort to calibrate and perhaps adjust the instrument as part of a regularly scheduled maintenance plan.

But before we move on to the procedure, I need to define a few more terms: calibration, adjustment, tolerance limit and reference substance.

Calibration is the act of checking (by comparison with a reference substance) the accuracy of a measuring instrument. This means that you are making a comparison of a measurement result using a reference substance for which the “true” value of the measured property is known.

Adjustment is defined as modifying the specific instrument parameters so that the measurement results of the calibration performed afterward are within the tolerance limit.

The tolerance limit is defined as the specified outer limits for permitted deviations from a “true” value. The tolerance limit is the responsibility of the operator and depends on his requirements regarding accuracy.

Ideally, the measuring system should produce results with an error less than the tolerance limit defined by the operator.

For the calibration measurements you need calibration methods, and you also need standards, notably known as reference substances.

A reference substance is a substance that is suitable for the calibration measurement, and whose specific properties used for the calibration, such its melting point, enthalpy of fusion or modulus,  are well established and accepted.

A reference substance is deemed suitable if it is easy to handle, readily available and stable.

Certified reference material is a reference substance, one or more of whose property values are traceable to primary standards.

All reference substances are delivered with a document certifying certain properties of the material.

METTLER TOLEDO, for example, offers metals such as indium, zinc and aluminum.

These substances are certified for their purity, but not for physical quantities such as enthalpy.

A certified reference material comes with a certificate for the certification of the glass transition of a particular substance.

For regulated areas, certified reference material should be available and can be obtained from LGC, NIST or PTB.

However, if you need to certify your instrument at high temperatures, you may encounter some restrictions because no supplier can deliver such certified materials.

For example, no substances are certified for enthalpy at 1000 °C or above.


Slide 6: Background information – What Needs To Be Calibrated?

Depending on its type, each instrument has different parameters that need to be considered for calibration.

The temperature needs to be calibrated for all instruments. In addition, the specific quantity measured by the instrument has to be calibrated, for example, heat flow for DSC or displacement in TMA.


Slide 7: Background information – Factors That Have an Impact

Unfortunately, most of these quantities are affected by the experimental conditions. The techniques that require a crucible, for example, are subject to the properties of the crucible and these have an effect on the measurement result. Such properties are thermal conductivity, mass and size.

Other such experimental factors are the furnace atmosphere, the purge gas flow rate and the various sample holders used in each type of instrument.

Fortunately, these phenomenological effects can be eliminated by calibration, and the measured values for your sample will not be influenced by these external factors.

Let’s see how this is done.


Slide 8: Calibration Procedure

So, when is calibration called for?
Possible instances are:

• with a new instrument
• when a specified time period has elapsed
• when an instrument has suffered a shock or experienced vibrations which potentially may have put it out of calibration
• whenever observations appear questionable

The calibration and adjustment of any instrument consists of a two main steps: preparation and following a simple workflow. I will describe these steps in detail as we go along.


Slide 9: Preparation

The first part begins with

1. Preparation, which consists of defining the measurement combination and method.

• This includes all of the parameters of your experiment such as choosing the crucible, the atmosphere, the heating rate, etcetera.
  

The methods for calibration could either be supplied by the manufacturer or you can develop your own methods for your specific needs.

• You also need to define the tolerance limits for your analytical procedure.
• You should not forget to define the calibration interval.
• Select the right number and type of reference substances. I will discuss reference substances in the next section because it is very important.

Finally, should your analytical procedure change, you may have to adapt your calibration method accordingly.

Slide 10: Choosing the Reference Substances

Imagine that you have to calibrate your instrument over a broad temperature range. In this case, it is not sufficient to calibrate for just one temperature. You should choose several reference substances.

The diagram shows a theoretical error curve and a parabolic correction curve based on 3 reference substances over a wide temperature range.

The error range is minimized between 100 °C and 500 °C between the 3 reference substances. At temperatures much lower than 50 °C, the error curve diverges excessively. This can be seen on the upper left side of the picture with the blue curve.

Note that for 2 reference substances, the error range is much larger. This can be seen from the red line on the lower left side of the picture.

We can therefore draw the following conclusions:

The more reference substances the better.

The reference substances should cover the range of interest.

You should not extrapolate 50 °C above or below the outer tolerance limits that you have defined.

Slide 11: Making a Decision

After you have completed the preparation and have performed a calibration, the decision-making phase begins.

There are generally two routes.

• If the calibration values are OK, then you can measure your samples. This is of course assuming that the calibration interval has not expired.
• If the calibration values are not within the tolerance limits you have defined, then you need to adjust the instrument.
  

Slide 12: Adjustment Procedure – DSC

Assuming the results of the calibration procedure require that you adjust the instrument, then here, too, you can follow a simple adjustment procedure.

I’ve based this and the following examples on the parameters for a DSC. These are tau lag, temperature and sensor adjustment.

Some of you might not have encountered the term “Tau lag”. Therefore I want to briefly describe what it is without going into too much detail.

It is a time constant describing the behavior of the furnace and it guarantees that the apparent influence of the heating rate has no effect on the results.

Note that each instrument is already adjusted at the factory for tau lag. If really necessary, this can be re-adjusted by a specialist.

Let us focus on the large picture on the right side. This is an enlargement of the area marked blue on the left side of the slide.

The workflow for performing an adjustment is simple.

For each calibration type, temperature, and sensor calibration respectively, the procedure is the same.

1. Since the values do not lie within the defined tolerance limits, you have to adjust the instrument. To do this, you can of course use the calibration results.
2. As a final check, run the calibration method again to verify that the adjustment was successful.
3. If the adjustment was successful, you can continue with your sample measurements.
4. If the adjustment was not successful, you need to adjust the instrument again.
   

To summarize the procedure, one full loop consists of calibrating, adjusting, calibrating to check for correct adjustment, and measuring your sample.

The ideal situation is when the result obtained after an adjustment procedure is no longer influenced by experimental factors such as the heating rate, crucible or gas flow rate.


Slide 13: Tau Lag Adjustment

So far, we have covered a lot of theory about statistics, calibration and adjustment. For the next few minutes, I’d like to show you the difference between instruments that have been properly adjusted and those that are not well-adjusted as well as the effect this can have on your results.

The first example illustrates how the onset melting temperature of indium is apparently influenced by the heating rate in Picture A. It seems that by increasing the heating rate, the onset temperature shifts to a higher temperature, which is physically wrong.

The falsely recorded onset temperatures are corrected by a tau lag adjustment.

As a result, we see corrected curves in Picture B, and the onset melting temperatures now agree with the true value of 156.6 °C.

Slide 14: Temperature Adjustment – Dynamic
    
In the next two examples, I will talk about temperature adjustment for dynamic and isothermal measurements.

First, let’s discuss dynamic measurement.

In Picture A, we see that the onset melting temperatures and the peak temperatures for indium and tin are significantly shifted, depending on the heating rate.

The measured values for these substances differ greatly, by several degrees in each case!

After temperature adjustment, the values agree perfectly for each substance.

Slide 15: Temperature adjustment – Isothermal (1)

Not only do we have to consider temperature adjustment for dynamic measurement methods, this is also important for isothermal measurement methods.

Isothermal methods are used for kinetics studies, OIT measurements, and for sorption measurements.

How can we make sure that the temperature reading is correct under isothermal conditions?

For this purpose, we’ve used indium and tin as our reference substances. Each substance is used for one specific temperature instead of a temperature range, as is the case with dynamic temperature measurements.

For both substances, the melting onset temperature is incorrect in Picture A. For indium we have a difference of 0.3 °C and for tin even 2.0 °C.

After adjustment, the onset melting temperatures agree with the true values. You can see the difference in Picture B.


Slide 16: Sensor adjustment – Heat flow (DSC)

We’ve arrived at the third example – sensor calibration and adjustment. All instruments have a sensor, but not all measure the same properties. The next few slides show the different calibration and adjustment possibilities for several instruments with respect to the sensor.
In the case of the DSC, the most important sources of non-reproducibility are the heat transfer between the sensor and crucible, and between the crucible and the sample.

These influences are eliminated when the DSC is properly adjusted using certified reference substances, for example from LGC (UK), NIST (USA) or PTB (Germany).


Slide 17: Sensor Adjustment – Weight (TGA)

Thermogravimetric analysis is a technique that measures the change in weight of a sample as it is heated, cooled or held at a constant temperature.

The heart of the TGA is the ultra-sensitive METTLER TOLEDO balance cell.

Correct weight adjustment is of vital importance for this instrument.

You have the option of performing an internal calibration with two rings or an external calibration with certified weights.

Slide 18: Sensor Adjustment – Displacement (TMA)

Thermomechanical analysis is used to measure the dimensional changes of a sample as a function of temperature. One of the most frequently studied properties is the expansion coefficient.

Samples can shrink or increase in length when heated.

For the TMA, the ability to correctly measure displacement is therefore a key parameter that requires regular attention.

You can adjust the length using different certified gauge blocks.


Slide 19: Sensor Calibration: Force & Displacement (DMA)

Dynamic mechanical analysis is used to measure the mechanical and viscoelastic properties of materials as a function of temperature, time and frequency when they are subjected to periodic stress.

The desired modulus can be calculated from the measured force and displacement amplitudes, taking into account the specific geometry of the sample.

Therefore, the instrument needs to be calibrated for length and force so that the correct modulus can be determined.

This requires certified gauge blocks and a spring.

An extremely precise spindle for the z-position, capable of defining 1 μm (speaker: say “one micro-meter”) steps, allows for fully automatic length adjustment after the spindle has been adjusted with gauge blocks.

In addition, a certified spring is used to calibrate the piezoelectric sensor over the whole force range.


Slide 20: Calibration – Modulus Accuracy (DMA)

The modulus accuracy can be proven by measuring a crystal silicon bar in the 3-point bending clamp according the specified method. This procedure is of course only valid for silicon.

The details of the measurement method and the expected value range can be read from the certificate included with the special reference sample.

If the modulus falls within the specified range, you have verified the modulus accuracy.

You can order this set from METTLER TOLEDO.

Slide 21: A Unique Concept from METTLER TOLEDO: FlexCal®

The instrument should be able to provide values that are independent of the heating or cooling rate, crucible type or gas atmosphere in the furnace.

Each measuring cell used should therefore also have its individual set of calibration parameters, and not be influenced by other instruments or by changing the controlling unit.

To achieve this outstanding goal, METTLER TOLEDO has implemented the FlexCal® option in its STARe System, which includes the methods and the database to store and handle the necessary calibration parameters.

Each instrument linked to the system has its own ID related to the actual calibration parameters.

Each instrument-specific set of parameters describes primarily the relationship between temperature, heating rate, crucible, gases and sensor type for the standard setup.

With the database-supported calibration model, FlexCal®, thermo-analytical results, such as the onset of melting or heat of fusion, no longer depend on the heating rate, the type of crucibles and gas atmosphere selected.

For ease of use, a new total calibration method performs a complete calibration using one or more reference standards.

The database supports the GLP-consistent documentation of the actual calibration parameters.


Slide 22: Summary

As we have seen, instruments that are not properly adjusted will give inaccurate and inconsistent results.

Therefore, it is important to regularly perform a calibration and check it against the defined tolerance limits.

This is essential for every laboratory to help ensure the integrity of your results.

In addition, METTLER TOLEDO has well-trained service and sales engineers who can assist you with equipment qualification, calibration and adjustment, training and application advice as well as provide assistance for service and maintenance issues.


Slide 23: For More Information on Calibration and Adjustment
Finally, I would like to draw your attention to information about calibration and adjustment that you can download from the Internet.

METTLER TOLEDO publishes articles on thermal analysis and applications from different fields twice a year in UserCom, the well-known METTLER TOLEDO biannual technical customer magazine.

Back issues can be downloaded as PDFs from www.mt.com under Usercoms.


Slide 24: For More Information on Thermal Analysis
In addition, you can download information about webinars, application handbooks or information of a more general nature from the Internet addresses shown on this slide.

Slide 25: Thank You
This concludes my presentation on calibration and adjustment in thermal analysis.

Thank you very much for your interest and attention.