1. Encapsulation

Encapsulation is necessary to prevent contamination of the analyser, and to make sure that the sample is contained and in good thermal contact with the furnace. One of the most annoying sights after a DSC run is to find a pan burst with the contents all over the furnace, or unexpected peaks and bumps in the DSC trace with no real cause or meaning. Such errors can be due to poor sample encapsulation, which is one of the most important areas to be considered in order to obtain good data. Most manufacturers provide a range of sample pans for different purposes, with a range of different sizes, pan materials, and associated temperature and pressure ranges. Consideration should be given to the following issues when choosing pans and encapsulating your samples.

A: Temperature range

Make sure that the pan is specified for the desired temperature range and will not melt during a scan; remember that aluminum must not be used above 600◦C. Pan materials such as gold (mp 1063◦C), platinum or alumina can be used at higher temperatures.

B: Pressure build-up and pan deformation

Pressure build-up in an inappropriately chosen pan is a frequent cause of difficulties. It is important to determine whether the sample needs to be run in a hermetically sealed pan or not; dry samples unlikely to evolve significant amounts of volatiles below decomposition do not need to be sealed. Yet if run in a hermetically sealed pan then pressure will build up inside the pan as it is heated. Many hermetically sealed aluminum pans cannot withstand high internal pressure and will deform (not necessarily visible) and result in potential artefacts in the trace as heat transfer to the sample changes. Ultimately sample leakage and bursting can occur, which usually results in contamination of the analyser. The best solution for such systems is to work with crimped pans that do not seal, or use lids with holes in. If not available then it may be best to pierce the pan lid before encapsulation so that pressure does not build up. Sometimes one hole will block with sample (particularly if a hole is made after loading the sample) and back pressure will force sample out of the pan giving more artefacts, so it is better to have more than one hole in the lid.

In the case of foods and other water-containing materials hermetic sealing is important since volatile loss will mask other transitions and will dry out the sample. It is then necessary to use a pan system capable of withstanding the anticipated pressure. The maximum operating temperature of a sealed pan may vary but may be below 100◦C. Some thicknesses of aluminum and type of pan offer slightly higher temperature ranges but consideration should be given to ‘O’ ring sealed stainless steel pans which can hold pressures above 20 bar. These have been designed for use with biological materials or for systems involving condensation reactions, e.g. phenolics.

In the case of samples taken to very high temperature or pressure, capsules which hold internal pressures up to 150 bar should be used. Disposable high-pressure capsules which avoid the need for cleaning are available. For hazard evaluation, gold-plated high-pressure pans should be used, since these should be inert towards the sample.

C: Reactions with the pan

Samples that react with a pan can cause serious damage to an analyser since they may also react with the furnace beneath. Solder pastes and inorganic salts are typical of the type of samples where care must be taken. If in doubt check it out separately from the analyser, and then choose a pan type which is inert. Sometimes the effect of catalysis is of interest and copper pans may be used to provide a catalytic effect. Aluminum pans are normally made of very high-purity metal to prevent unwanted catalytic effects.

D: Pan cleanliness

Most pans can be used as received but sometimes a batch may be found to be slightly contaminated, possibly with a trace of machine oils used in pan manufacture. If so, pans can be cleaned by volatilising off the oil. Heating to 300 ◦C should be more than adequate for this. If using a hot plate, do not heat a lot of pans all together or they may stick together. The use of clean pans is also important for very sensitive work with fast scan DSC.

E: Pans with very small holes

Some pan and lid systems have been developed with very small holes, typically about 50 um diameter that have found use with hydrate and solvate loss. They can be used with ordinary samples to relieve internal pressure, so long as the hole does not block, but are intended to improve resolution of mass loss from a sample. As volatiles do not tend to escape through such a small hole until the volatile partial pressure exceeds atmospheric pressure boiling point can be determined.

F: Liquid samples

Liquid samples must be placed in sealed pans of a type that can withstand any internal pressure build-up. Do not overfill the pan or contaminate the sealing surfaces, which will prevent sealing and cause leakage. When sealing a liquid, bring the dies together gently to avoid splashing the sample.

G: Sample contact

Samples need to be in good thermal contact with the pan. Liquids and powders, when pressed down, give good thermal contact, other samples should be cut with a flat surface that can be placed against the base of the pan. Avoid grinding materials unless you are sure it will not change their properties. If possible, films should not be layered to prevent multiple effects from the same transition, though this may be the only way to get enough samples. If so, take care to make sure that the films are pressed well together. Low-density samples provide poor heat transfer so should be compressed. Some crimping processes do this automatically; with others it may be of value to compress a sample between two pan bases or using a flat pan lid as an insert. Take care not to deform a pan and discard any pans that are obviously deformed before use. Sometimes pans of slightly thicker aluminum can give better heat transfer because they retain a flatter base.

H: Spillage

A frequent cause of contamination is from sample attaching to the outside of the pan. Check and remove any contaminant sticking to the outside of the pan, particularly the base of the pan. A soft brush is good for removing powders.

I: Use without pans

Occasionally, very stable samples may be run without pans, for example stable metals over a low temperature range. These benefit from increased thermal transfer to the sample. Thermally conductive pastes have also been used on occasion to try to improve thermal transfer, but such measures require great care, and the use of helium or other highly conductive purge gas may be a better alternative.

2. Temperature range

The starting temperature should be well below the beginning of the first transition that you want to measure in order to see it clearly following a period of flat baseline. This should take into account the period of the initial transient where the scan rate is not yet fully controlled and the baseline is not stable. With ambient DSC systems the starting temperature is often around 30◦C. The upper temperature should be below the decomposition temperature of the sample. Decomposing a material in a DSC normally gives rise to a very noisy drifting response and the evolved volatiles will contaminate the system. It is good practice to establish the decomposition temperature first using a TGA analyser if available.

3. Scan rate

Traditionally, the most common scan rate used by thermal analysts is 10◦C/min, but with commercially available instruments rates can be varied between 0.001 and 500◦C/min, often to significant advantage. Choice of scan rate can affect the following areas:

A: Sensitivity. The faster the scan rate the greater the sensitivity. If a sample with a small transition is received where great care is thought necessary, there is no point in scanning slowly since the transition is likely to be missed altogether. Whilst the energy of any given transition is fixed and should therefore be the same whatever the scan rate, the fact is that the faster the scan rate, the bigger the transition appears. Increasing scan rate increases sensitivity. The reason for this is that a DSC measures the flow of energy and during a fast scan the flow of energy increases, though over a shorter time period. A slow scan results in lower flows of energy over a longer time period, but since DSC data are usually shown with the x-axis as temperature, it simply looks like a transition is bigger at the faster rates; see Figures 1a and 1b. Hence an increased scan rate leads to an increase in sensitivity, so do not use slow rates for small difficult-to-find transitions, unless otherwise unavoidable.

Figure 1. The effect of increasing scan rate on indium. Part (a) (upper curve) is shown with the x-axis in time; part (b) (lower curve) is shown with the x-axis in temperature. The same energy flows faster in a shorter time period at the faster rates, giving larger peaks.

B: Temperature calibration. Calibrations need to be performed prior to analysis to ensure that the scan rate of use is calibrated. Different instruments use different approaches to calibration and manufacturer’s instructions should be followed.

C: Resolution. Because of thermal gradients across a sample the faster the scan rate the lower the resolution, and the slower the scan rate the sharper the resolution. Thermal gradients can be reduced by reducing sample size and improving thermal contact with the pan by good encapsulation.

D: Transition kinetics. Slow events such as a cure reaction may not complete if scanned quickly and may be displaced to a higher temperature where they can occur more rapidly. The kinetics of an event may need to be considered when choosing a scan rate.

E: Time of analysis. Speed of analysis is an issue in many businesses, and higher rates speed up throughput.

4. Sample size

The amount of sample used will vary according to the sample and application. In general, make sure that excessive amounts of sample are not added to the pan. Sample collapse when melting or softening can give rise to noise during or after the transition. Reducing sample size minimizes the chances of this happening. Often just a few milligrams or less is sufficient, typically 1–3 mg for pharmaceutical materials, though for very weak transitions and for accurate heat of fusion measurements more sample may be needed. Note that when considering accuracy of energy measurements, the accuracy of the balance should be taken into account. For most work a five-figure balance is needed, and six figures (capable of measuring to the microgram level) for more accurate heats of fusion. For polymers slightly larger samples of 10 mg are often used, but again this depends on the sample and the transition of interest. See the following chapters for practical discussion of individual application areas. As a general comment, if the sample is to be melted then choose a lower weight, though for accurate heat of fusion values typically 5mgis needed in order to measure the weight accurately, possibly higher if the balance accuracy is not very great. Choose a pan that can cope with sample size needed. Note that one of the major causes of contamination is also from using too large a sample size.

5. Purge gas

Purge gases are used to control the sample environment, purge volatiles from the system and prevent contamination. They reduce noise by preventing internal convection currents, prevent ice formation in sub-ambient systems, and can provide an active atmosphere. They are recommended for use and essential for some systems. The most common purge gas is nitrogen, which provides a generally inert atmosphere and prevents sample oxidation and is probably the default choice. Air and oxygen are sometimes used for oxidative tests such as OIT (oxidative induction time) measurements. Helium is used for work at very low temperatures where nitrogen and oxygen would condense, and is recommended for fast scan DSC studies. Other gases such as argon have been used and this can be helpful when operating at higher temperatures, typically above 600◦C. This gas has a lower thermal conductivity, which results in lower heat loses but at lower temperatures results in decreased resolution and sensitivity. Air, nitrogen and oxygen have similar thermal conductivities and instrument calibration is unaffected if they are switched. The use of helium or argon requires separate calibrations to be performed. For totally inert performance use high-purity gases together with copper or steel gas lines, and give time for the sample area to be fully purged of air before beginning a run.

Gas flows are controlled either by mass flow controllers or by using a pressure regulator and restrictor. If you want to check for gas flows, do not put the exit line into a beaker of water. Suck back can occur, destroying the furnace, and bubble formation causes noise. A soap bubble flow meter can be used if needed, but any device used to measure the normally low purge gas flows should not in itself cause a restriction.

6. Sub-ambient operation

A variety of cooling systems are available for most instruments and these are now mostly automatic in operation. Issues of filling liquid nitrogen systems and of availability mean that refrigerated coolers (intracoolers) are often preferred and some systems can operate at below −100◦C. Intracoolers are preferred for use with autosamplers since there is no risk of the coolant running out.

7. General practical points

Cleanliness is next to godliness as far as DSC use is concerned. Contamination will occur if a pan is placed on a dirty surface and then transferred to the furnace, so keep the analyser and the working area clean and tidy, and discard used pans to avoid mixing them up. DSC furnaces should be kept clean; refer to manufacturers instructions for the most appropriate cleaning method. Avoid abrasives and be aware of the dangers of using flammable solvents. Take note of specific instructions for any particular type of furnace.

• Do not heat aluminum pans above 600◦C.

• Do not operate furnaces in an oxidising atmosphere above their recommended limits.

• Do not use excessive force when cleaning a furnace.

• Discard used samples immediately after use before they become mixed up with fresh samples.

8. Preparing power compensation systems for use

With power compensation DSC analysers there are two small furnaces and some specific practices to be aware of.

Before starting and before turning on any sub-ambient accessories, inspect the furnaces to make sure that they are clean and the surrounding block is dry and free from condensation. Clean furnaces with a swab and suitable solvent, and complete this with a furnace burn-off in air at about 600◦C or above. For this purpose, the furnace should be opened to allow air access and volatiles to escape easily. Any pans should be removed. The furnaces are constructed of 95% platinum, 5% iridium, which will not oxidise within the temperature range of the analyser. Though organics will burn off, metal and inorganic content may burn in, so make sure that the furnaces are as clean as possible before a burn-off. The platinum furnace covers should also be clean; inspect the underside to make sure. The furnace covers should not be swapped about between sample and reference as they will have slightly different heat capacities which will affect instrument response run to run. They should fit easily, otherwise they may be incorrectly fitted. They can be reformed if needed using a reforming tool. Incorrectly fitting covers will lead to significant slope on the baseline. If a sample has leaked and stuck on the pan or a cover to the furnace, it is often helpful to heat the furnace to a point at which the sample softens and its adhesive properties are lost. The sample and pan can then be removed and the furnace cleaned.

The underside of the swing-away or rotating cover should be inspected from time to time to make sure that it is clean; particularly if sublimation is suspected. If the guard rings beside the furnaces are contaminated these can be removed and cleaned. They should be replaced with the slot facing the purge gas exit port, which is situated centrally behind the furnaces.

Turn on the purge gas and cooling system and allow them to stabilize; this applies to subambient or ambient systems where water circulation is used. Stable background temperatures and purge gas flow provide a controlled environment from which stable and reproducible performance can be obtained. Dry purge gases must be used for the instrument air shield if operating below ambient. If the system has been left for any period of time it is good to condition the analyser by heating to 400–500◦C before beginning your work.