1

Analytical & Chemical

Composition Measurement

North Lindsey College Compiled by Dr S. Koudis, September 2013 Updated by Mrs A. Trenholme, September 2014

2

LIST OF CONTENTS

Chapter 1: Sampling for the Process industry Page

1.1 The need for sampling 4 1.2 Sampling system design considerations 6 1.3 Considerations for sampling systems 7 1.4 Existing methods of sampling 8 1.5 New methods and equipment 9 1.6 Types of sample conditioners 14 1.7 Gas sample conditioning devices 15 1.8 Examples of sampling systems 19

Chapter 2: Gas analysis

2.1 Block diagram for a typical gas analyser 25 2.2 Magnetic Wind analyser 27 2.3 Magnetodynamic Oxygen analyser 32 2.4 Zirconium Oxide potentiometric probe 34 2.5 Single beam IR analyser 37

Chapter 3: Density measurement

3.1 Type 1 - Static pressure operated instruments 39 3.2 Type 2 – Weighing tube instrument 40 3.3 Type 3 - Buoyancy type instruments 42 3.4 Type 4 – Absorption of radiation instruments 43 3.5 Type 5 – The PAAR oscillation tube meter 44

Chapter 4: Humidity measurement

4.1 Applications of humidity measurement 47 4.2 What is humidity? 47 4.3 Thermoelectric Dewpoint instrument 49 4.4 Wet and Dry bulb hygrometers 50 4.5 The Gregory balanced temperature hygrometer 52 4.6 Hair hygrometers 53 4.7 Capacitance hygrometers 54 4.8 Infra-Red instruments 55 4.9 Electrical conductivity instrument 56 4.10 Piezoelectric humidity instrument 57

Chapter 5: Viscosity measurement

5.1 Types of viscosity 58 5.2 Glass viscometers 61 5.3 Rotational viscometers 62 5.4 Falling body viscometers 62 5.5 Vibrational viscometers 63 5.6 Arbitrary viscosity determination 64 5.7 Continuous viscosity measurement 67

3

Chapter 6: Electrical conductivity measurement

6.1 Applications of conductivity measurement 71 6.2 Measurement of electrical conductivity 71 6.3 Conductivity cells 74 6.4 Instruments for conventional a.c. measurement 76 6.5 Choice of measurement parameters 77

Chapter 7: Electrochemical methods

7.1 Classification of Electrochemical methods 78 7.2 The pH and its measurement 80 7.3 The Calomel reference electrode 82 7.4 The Glass measuring electrode 83 7.5 Electrical circuits for use with Glass electrodes, the pH meter 84 7.6 Alkaline and Asymmetrical errors 86 7.7 Industrial pH systems with Glass electrodes 87 7.8 Redox (Eh) potential and its measurement 88

Chapter 8: Chromatography

8.1 Classification of chromatographic techniques 90 8.2 Basic principles of Chromatography 91 8.3 Gas chromatography, sampling system, columns, carrier gas 96 8.4 Component detection, TCD, FID, ECD 99 8.5 Operation and data handling of a chromatograph 102

References 103

4

Chapter 1: Sampling for the Process Industry

1.1 The need for sampling

Throughout the years, people have been trying to find better methods to sample process liquids and gases. While sampling has been done for years, only recently have manufacturers begun to truly evaluate the needs of customers and develop better sampling systems to meet today’s more sophisticated sampling requirements.

There are many reasons sampling is done in a process piping system. Sampling of

processes is routinely done to check

(i) product quality

(ii) product purity

(iii) safety of operation and effluent monitoring

(iv) efficiency checks and process optimisation

(v) costing of raw materials

(vi) a feedback signal in control loops

Sampling is typically a major concern for most plants since product samples must be

representative and accurate. Environmental and safety concerns related to sample

taking of toxic or hazardous materials are also having a larger impact on the

development of sampling devices.

In fact without the use of continuous analysis, many of the chemical processes

currently in use would be inoperable. In a less quality and cost intensive era, grab

samples may be used. In a grab sample a specific quantity of process fluid or solid

is extracted, transported to the laboratory and analysed by classical chemical

analysis. The results of this analysis can be used to adjust the processing in order to

make the product more desirable. Today, grab samples do not allow a product to be

made having a high specification and at the right price.

5

The block diagram of a generalised sampling system is shown below.

The probe must be made of a material that will not react with or contaminate the

process fluid or solid. It must be sited so that it will, under operating conditions,

extract a representative sample. Representative means that the small amount of

sample extracted will be exactly the same as (at any particular instant) the vast bulk

of the material still in process stream.

The sample conditioning can be one or many processing stages required to convert

the raw sample into a form acceptable to the analyser. Conditioning may include the

following:

(i) drying to remove water vapour

(ii) pressure control

(iii) flow control

(iv) filtration (removal of unwanted solids)

(v) scrubbing (removal of unwanted or toxic fractions of the sample which are

not relevant to the parameter being measured)

(vi) entrainment separation (most analysers can operate on only one phase of

substance, liquid, vapour, gas or solid – any excess phases must be

separated, such as in a liquid analyser gas bubbles would lead to errors).

The measuring element receives the conditioned sample and converts it into another

energy form. It is essential that the characteristic of the measuring element is such

that a repeatable relationship exists between the variable being measured and the

signal produced. Some examples of measuring elements with their inputs and

outputs are given below:

6

pH sensors – emf (mV)

density sensors – ∆F (frequency change)

conductivity sensors – ∆R (resistance change)

The difficulty with sample disposal depends upon the nature of the conditioned

sample. If the substance is basically harmless then drain off or vent to atmosphere.

If however the sample contains toxic or explosive components, then a safe method

of “decontamination” must be found. With difficult samples it is useful if a low

pressure region of the process can be found where the “spent sample” may be

returned to the bulk of the process material.

The transmitter will take the signal from the measuring element and convert it into a

form more suitable for transmission or computation. If the transmitter output has to

send the signal over large distances then its output will usually be an analogue 4-

20mA signal or a digital series communication protocol. The transmitter may also

include some form of linearization and scaling between its input and outputs. The

display/recorder/controller will accept the signal from the transmitter and use that

data to control the process or convert the signal into “engineering units” for display to

process operators.

1.2 Sampling system design considerations

A major problem in any analyser control loop is due to lags in the system, i.e. the

time that elapses between the change of composition occurring in the process and

the controller being aware of it. The lags are usually transfer lags i.e. due to the

capacitance and resistance of the pipes and conditioning equipment in between the

probe and the analyser.

When designing a sampling system consideration must be given to reduce lags.

This is achieved by keeping the volume of pipes, filters, etc to a minimum. In the

case of pipes reducing the volume must mean a decrease in diameter, but a

decreased diameter inevitably means increased resistance to flow which will counter

the reason for decreasing the diameter. It may be appreciated that the physical

sizing of sampling systems is always a compromise between acceptable resistance

and capacitance. Also smaller filters have less filtering surface area and hence tend

to “choke” more quickly. In general smaller conditioning units have shorter lives

between maintenance.

Perhaps the most dramatic reductions in lags is achieved in gas systems by

designing for “minimum system capacity”. System capacity is a measure of the

mass of gas in the sampling system waiting to go through the analyser.

7

In analysers classed as being “in-line” the measuring element (probe) is mounted in

the process main (or vessel) and if the measuring element can tolerate the “raw”

main conditions reliably, then in-line analysers will significantly increase the speed of

response.

High speed sample loops also reduce the system capacity thus reducing system

lags. They are employed when the analyser must be sited remotely from the

process main.

1.3 Considerations for sampling systems

When considering process sampling, several critical questions must be answered to determine what type of sampling device will be required and the best method for taking the sample.

Some of the basic questions to answer are:

(i) Why is the media being sampled? (process verification, bacteria counts, quality assurance)

(ii) How often is the media being sampled?

(iii) What type of media is being sampled? (powder, slurry, liquid, gas)

(iv) What are the properties of the sample? (corrosive, hazardous, flammable, carcinogenic)

(v) Where in the process is the sample being taken?

(vi) What is the viscosity of the sample and are solids present?

(vii) Does the sample crystallize?

(viii) Will the sample be taken from a pipeline or vessel?

8

1.4 Existing methods of sampling

Existing sampling methods range from opening a valve and allowing the liquid

sample to flow into a bucket or jar, to more sophisticated sampling systems providing

representative and safe samples from both pipelines and vessels or reactors. Over

the last ten years, the sophistication of these sampling methods has increased

dramatically due to the requirements of the Occupational Safety and Health

Administration (OSHA) and the Environmental Protection Agency (EPA) as well as

plant environmental and safety concerns.

In the past, sampling methods have traditionally been engineered internally by

companies using existing process valves on the market (ball, globe, or plug) and

adapting them to their process systems and sampling needs. The most basic of

these systems is to provide a “tee” in a process pipe or vessel with a small valve on

the “tee” that would be used to drop the sample from the line or tank. A bucket or jar

would be held under the valve opening to catch and hold the sample being taken.

Sometimes a glove box is built around the valve to protect operators while they are

taking the sample. In some applications, separate bypass loops are built for taking

samples. In these cases, there are typically several isolation valves piped into the

system to isolate the sample in the bypass loop and allow the sample to be dropped

into a sample container.

Some batch processes utilizing reactions would require samples to be taken from the

reactor vessel during the process. Samples from these systems were typically taken

by opening the hatch at the top of the reactor and dipping a cup attached to a long

stick into the vessel to withdraw a sample.

In other cases, samples are taken by creating a re-circulation loop with the product

being pumped out of the reactor, through a piping system including a standard valve

for taking the sample and then back into the reactor. This is very expensive and

usually requires that you use two nozzles on top of the reactor.

One additional method for sampling from a reactor or tank is to take a sample

through a traditional ram type, y-pattern sample valve, or tank bottom ball valve at

the bottom of the reactor. This type of valve is mounted flush with the bottom of the

reactor and the valve is opened to allow the sample to flow from the bottom of the

reactor to a sample container. This method of sampling is efficient but cannot

provide a representative sample since the sample is always taken from the bottom of

the vessel or reactor.

9

1.5 New methods and equipment

The latest sample valve designs try to incorporate safety, environmental protection,

and reliability while allowing the user to take representative process samples. These

designs are continuing to evolve as industry push sampling system manufacturers to

improve their existing systems.

Current sample valve designs can be easily categorized from the area in the process

where the samples are being taken.

The general categories are:

(i) in-line sampling and

(ii) reactor or tank sampling.

In-line sampling systems allow process samples to be taken directly from a pipe

into a sample container. This ensures that a representative sample is taken from the

process stream and eliminates the need for a separate bypass to be installed into

the piping system to allow sample taking. The newest in-line sampling devices also

ensure that samples are not exposed to the atmosphere. This feature ensures the

safety of the operator from hazardous samples, while also maintaining the integrity of

the sample since the atmosphere cannot contaminate the sample.

These in-line systems have several different methods of retrieving the sample from

the pipeline. The basic in-line sampling valve either mounts between two standard

flanges or can be mounted directly by welding into the side of the pipe.

Side mounted sampling devices are typically more compact and use a small ball or

globe valve to isolate the sample container from the process pipeline. They will either

have an orifice directly on the side of the pipeline connected to the valve or will utilize

a probe sticking out into the pipeline connected to the isolation valve. These devices

typically have small orifices and are best used for clean liquid samples.

The sample container is usually a glass jar with a septum top (self-sealing

diaphragm). The sample is transferred into the jar through a pair of needles

puncturing the septum. One needle is to vent the air in the jar while the other needle

allows the sample to fill the jar. The vent is typically piped to a scrubber system for

hazardous processes.

The septum bottles typically provide the best containment of the sample since there

is no chance that the sample can be exposed to the atmosphere. One drawback of

using septum bottles is that the needles are typically a small diameter and can block

when sampling liquids containing solid particles or liquids that tend to crystallize.

Wafer or flanged in-line systems that mount between two flanges in a piping system

typically use a rising stem design with a spindle that seats at the bottom of the valve.

The spindle is lifted either by a multi-turn hand wheel or a quarter turn lever. Once

10

the spindle is lifted the sample can be dropped into a sample container. These

devices provide the most representative samples since there are no dead legs in the

valve and the sample comes directly from the process pipeline.

These devices typically have larger orifices and can handle liquids containing solid

particles or crystallizing liquids. These devices have two ports for sample taking. A

smaller port vents the air from the bottle and the larger port allows the sample to

enter the bottle. The vent port is usually piped to a scrubber system when the

samplers are being used in hazardous processes.

Multi-turn designs allow accurate metering of the sample into the container but are

not fail-safe since the operator could potentially leave the sample valve open and

overflow the sample container. Lever operated designs are usually safer since they

incorporate a spring loaded handle which automatically closes the valve when the

operator releases the handle.

There are two types of sample containers used:

(i) wafer and

(ii) flanged style sampling devices.

The most common container is a glass bottle threaded into the bottom of the sample

valve. This bottle holds the sample when it is dropped from the pipeline. The bottle is

not the best method of containing the sample since they must be unscrewed from

the sample device and capped. This exposes the sample to atmosphere and

exposes the operator to the sample. Using a glove box around the sample device as

shown in figure 1 can minimize the exposure of the operator to the sample. This

allows the bottle to be capped in the glove box without exposing the operator to any

fumes from the sample.

11

Figure 1: Sample container

Another method for transferring the sample from a flanged or wafer type in-line

sample valve is to use a syringe (figure 2) to remove the sample from the pipeline.

The syringe allows the sample to be totally contained and transported safely back to

the lab. This method ensures the safety of the operator and the integrity of the

sample. The syringe does have a drawback in that the orifices are smaller in the

syringe so that it can only handle clean liquids. The syringe type sampling systems

also require cleaning after each sample to prevent contamination of the next sample

to be taken.

There are many other sampling applications for which new in-line equipment is being

developed. There are aseptic sampling systems, which are used in the food

processing industry. These in-line systems incorporate equipment that sterilize the

sample valve before each sample is taken. These systems are typically used in

applications where samples are taken for bacteria counts to document the purity of

the processed liquids.

12

Figure 2: Sapro Syringe

Reactor sampling systems are becoming more popular especially in

pharmaceutical and chemical plants. These systems usually incorporate a PTFE

lined or alloy dip tube mounted to the top of the reactor. The dip tube allows access

into the reactor for the suction hose as well as the introduction of additional

chemicals to the reactor. The dip tube is placed at a level in the reactor where the

client wants to sample from.

On top of the dip tube goes the reactor sampling system. This system usually

consists of an isolation valve on the dip tube, a combination ball check/sight glass,

and a transfer device to take the sample from the sight glass into a bottle or other

container. The simplest systems use a vacuum to draw the sample up the suction

hose and into the sampling system. More sophisticated systems will use a re-

circulation pump attached to the sampling system that re-circulates the sample

before it is taken so that more accurate and representative samples can be taken.

These systems with re-circulation pumps can typically also be used to monitor pH in

the process while the sample is being re-circulated (see figure 3).

13

The reactor sampling systems are typically very modular and can be customized to

the specific requirements. Most systems will also allow for additional access ports to

the sampling system. These additional access ports can be used to clean the

sampling system or purge the system with nitrogen after each sample is taken.

14

1.6 Types of sample conditioners

Sample conditioners can be separated into three categories:

(i) Refrigerant type using compressors and evaporators like a household refrigerator to cool a sample.

(ii) Peltier elements using the Seebeck effect to cool a sample with a semiconductor plate and heat dissipation.

(iii) Various forms of catch-pot, silica gel filters, scrubbers and the like.

The refrigerant type will always be with us for a number of reasons. It is more

efficient. Although no form of refrigerant system can be seen as particularly efficient in normal engineering terms, this type of cooler requires the least power input for a given cooling effect. They can be built to produce very high levels of cooling, making them ideal for stationary installations and similar use. The construction principles are well-known, making maintenance less of a problem. There are, nevertheless, some important drawbacks to the use of this type of equipment. The refrigerant used can become a problem for ecological reasons. The traditional R12 coolant that was used in refrigerators for years is out of popularity due to its contribution to the greenhouse effect. This change can only be applauded but has caused a number of problems in the refrigerant branch. Using propane as a coolant in electrical equipment can only be viewed with horror. Not only is it a highly flammable gas, but its contribution to the greenhouse effect is possibly greater than that of the earlier refrigerant. There are now other refrigerants on the market, but the price and efficiency have both undoubtedly suffered from the change.

The next disadvantage is weight. Anyone who has carried a household refrigerator up a few flights of steps will need no extra explanation to this point! A motor, compressor and evaporator have a certain weight, regardless of what you do. Related to this is the problem of bulk. The evaporator requires a certain area and the heat exchanger must be physically separated from the cooling unit to increase efficiency. There is no point in trying to cool your own waste heat energy!

Such a refrigeration unit has moving parts and, as we all know, anything that moves is subject to wear. Motors and compressors will wear out in time and require replacement, making routine maintenance essential.

The heat exchanger of a refrigerator is a fragile unit. It consists of thin pipes or similar containing the hot fluid try to give up its heat energy. These can be easily damaged, leading to loss of refrigerant and, naturally, dysfunction.

Shaking the refrigerant in transport will lead to vapour locks which must be left to settle before the unit is used. Using the unit immediately after transport will lead to premature failure of the pump and other parts.

Why anyone would use a refrigerant type when there is a Peltier element available. The answer is, of course, power and efficiency. A Peltier element requires an immense amount of current to produce the cooling effect needed and must generally be cascaded to produce any useful cooling at all. It simply has the advantage of relatively low cost, light weight and no moving parts. It can be instantly used after

15

transport without any fear of failure and will cool down more rapidly than the refrigerant type.

The last type, the catch-pot and the like also have a place. They are perfectly

acceptable where there is no fear of loss of sample due to solubility and have the great advantage of low cost and negligible maintenance, except for regular emptying!

This is naturally not everything that is required for sample conditioning, but cooling of the sample is the first stage. The next step is to remove the condensate quickly and efficiently from the unit before any effect on the gas becomes unavoidable. The most common method is the use of a pump or similar to remove the water from the lowest point of the condensing chamber. Careful use of geometry can ensure that there is little contact between gas and condensate and, hence, minimal interaction.

1.7 Gas sample conditioning devices

Gas analysers require sample gas which is:

Dry Free from dust At correct pressure At correct flow rate

Correct conditioning of the sample gas will ensure:

Accurate measurement High reliability and availability of the analyser system Reduced maintenance requirements Reduced cost of maintenance

Sample dryers and humidifiers

Dryers remove water from sample gas streams and can achieve a resultant dew-point of <-30oC with flow rates of a few ml/mins to 10-20 l/min.

Loss of water soluble gases such as HCl, SOx and NOx is negligible giving more accurate measurements. Humidifiers bring the humidity of a gas stream to that in the ambient atmosphere and find uses in medicine, gas analysis and fuel cell technology. Process humidifiers can humidify gas at flow rates of hundreds of litres a minute to nearly 100% RH at elevated temperatures.

16

Coalescing filters

Coalescing filtration is achieved by forcing the gas stream through the filter element. The flow path shown in figure 4 reflects a coalescing application. Small aerosol particles are forced together as they pass through the fine inner layer of the filter element. The larger drops created in this process then begin to fall out of the gas stream due to their mass.

Figure 4: Coalescing filter

Features include:

Coalesces liquid acid mists with excellent removal of SO2 mists Removes water droplets Large dirt-holding capacity Minimal maintenance Easily replaceable, low cost filter element Corrosion resistant Compact Simple installation

Eductors (non-mechanical pumps) Eductors provide a non-mechanical means of pumping or mixing a sample stream. An eductor is a device in which the energy of one stream is used to pump another.

17

Figure 5: Eductor

The pumping gas, typically compressed air, develops a vacuum at the suction port

that causes the sample in the suction chamber to be entrained with the stream and

be delivered through the exhaust port.

Features include:

No moving parts Minimal maintenance Compact design No electrical parts, suitable for use with flammable gases Corrosion resistant

Inertial bypass filters

If used upstream it will greatly increase the life and performance of coalescing filters.

These filters can be installed to reduce the response time of analysers by forming a

fast by-pass loop. Options include heating, to prevent condensation.

Features include:

Continuous filtration No moving parts Transports large flow, reduces response time Cleanable components Compact design No electricity required Low pressure drop

18

Figure 6: Inertial bypass filter

Ammonia scrubber Ammonia scrubbers are designed to remove ammonia from a gas stream and protect gas analysers from clogging due to the formation of ammonium salts.

Ammonia is sometimes added to stack gases to reduce the nitrogen oxides content of the gases by conversion of them to nitrogen and water (DeNOx process). Typically, an excess of ammonia is added and will be present in the gas samples. This ammonia residue will readily react with other components such as sulphur dioxide in the sample to form ammonium salts. This salt is relatively low-boiling, so when the sample cools down it precipitates out as a solid, clogging sampling line analysers and other processing equipment. The scrubbing principle is based upon phosphoric acid. The scrubber requires occasional regeneration; the period depends on flow rate and ammonia concentration.

Features include:

Lengthens analyser life Reduces maintenance Easy replacement of scrubbing media No moving parts Heated to prevent condensation

19

Figure 7: Ammonia scrubber

1.8 Examples of sampling systems

The need for representative samples plays a critical role in ensuring product

verification. Yet sampling directly from the process often includes the risks of

exposure to the operator as well as contamination and pollution to the environment.

The DOPAK® sampling method reduces such risks with its patented design and

simple method of operation.

Figure 8: The DOPAK sampler

20

The DOPAK® sampler (figure 8) solves the problem of taking samples of toxic,

dangerous and volatile substances as the operator is well shielded from contact with

the product being sampled. Local spillage can be avoided and volatile substances

will not escape into the atmosphere.

In its simplest form, the sampler consists of a valve, needle assembly and protective

sleeve. The needle assembly is attached to the valve and houses a pair of needles.

One needle accepts sample from the process: the other acts as a means to vent.

Figure 9: Steps of operation

DOPAK operation

Step 1

The sampler is installed directly into the process via a connection on top the valve. A

sample is drawn from the process and arrives at atmospheric pressure in the sample

container. This container consists of a glass bottle with a threaded, open-top cap.

Between the bottle and cap a self-sealing diaphragm (septum) is placed.

Step 2

The bottle, with cap and septum, is inserted into the sleeve until the needles

extending from the needle assembly pierce the septum.

Step 3

Once in position, the operator opens the valve, allowing product to flow as shown in

figure 9 (2).

21

Step 4

When the required amount has been taken, the operator closes valve and the bottle

is pulled out of the sleeve. The septum is of such material that when pulled from the

needles, it reseals automatically. The bottle is then taken to the process laboratory

for evaluation. The sample fluid can be extracted from the container via a syringe

thereby reducing or eliminating exposure of laboratory personnel to the sample fluid.

Then the evaluation is complete, the entire contents can be disposed of in the still

sealed container.

Needle assemblies

The DOPAK® Sampler design centres on the needle assembly. It consists of a

housing in which two needles are mounted. In standard practice, the longer needle is

used to vent out the air displaced by the incoming fluid, which is sometimes mixed

with gas or vapour. Two venting models are available. The VTA design (vent to air)

allows the air from the bottle to be vented directly to the atmosphere. The VTO

model (vent to outlet) provides a fitting connection to the vent needle. In this way the

vent needle can lead to a waste system, wet collector, scrubber, activated carbon

container, or back to the process if the process pressure allows. The DOPAK®

sampling system does not allow for a high-pressure build up in the container or a

total blockage of the vent outlet. If this happens, the septum would open up

somewhat resembling a safety valve. Because of this, placing valves in the vent

outlet should be avoided. If required, a check valve could be installed in the vent

outlet. The purpose is to prevent any unwanted backflow of the waste system when

the sampler is not in use.

One-handle operation

DOPAK® sampling systems are still based on the basic DOPAK® principle of filling a

sample into a sealed sample container. This can be either a bottle, sealed with cap

and septum or a cylinder. Whenever a DOPAK® sampling system requires more

than one valve, it is considered whether operation technically can be done by one

handle. In DOPAK®’s opinion, one-handle operation ensures greater safety,

avoiding sequence-mistakes in operating more than one valve during the sampling

operation. In most cases, even an interlock-feature could be added to ensure access

to the system and/or to guarantee the correct sequence of various steps in the

sampling operation.

System 23 and System 32

The group of systems has been divided into two ranges, System 23 and System 32.

All System 23s have been designed with a predefined volume sample chamber.

22

System 32s generally lacks a sample chamber. They are more geared towards the

use of a standard DOPAK® Process sampler with an additional valve that is

operated in conjunction with the main valve. Combinations of a System 23 and a

System 32 are feasible. Both use more than one valve. In the case of the System 23,

the valves are internally coupled through the sample chamber. The System 32 is

different in that valves are externally coupled.

The System 23 has been designed to enable sampling of predefined quantities of

fluids independent of the process pressure or installation. The system can be used

regardless if operated under vacuum or high pressure. The system 23 (shown in

figure 10) is built with a sampling chamber that can be shut off simultaneously at

both ends with two three-way valves. Both valves are connected in such way that

they are operated by one handle. With continuous flow through the sample chamber

i.e. system purge, a fresh sample is assured. This cleans both the process piping to

the sampler especially that of higher viscosity fluids when an inert purge is not able

to completely remove 100% of the fluid contained in the sample chamber.

Figure 10: System 23 with a sampling chamber

With Dopak’s principle the sample fluid will end up at atmospheric pressure inside

the sample container, regardless of process pressure or vacuum condition. With the

System 23, the sample fluid can be circulated through it at a high pressure or under

a vacuum. The inert purge then removes the sample from the System 23 at a pre-set

pressure. The volume of the sample chamber decides the volume of the sample.

23

This volume has to be specified when ordering the system. The factory has chosen a

range of standard volumes. Any other size can be supplied on request.

Special models

Figure 11 shows System 23 with a cooling jacket around the sample chamber. The

jacket is suitable for water-cooling and constructed form the same materials the

System 23. It is possible to leave the sample in the sample chamber to let it cool off

to ambient before it is transferred into the sample container. However, the purpose of

this design is to lower the temperature of the sample when captured in the sample

chamber quicker before it is transferred into the sample container. Some fluids like

hydrogen cyanide (HCN), require welded or flanged connections through-out the

process installation. The System 23 has to be adapted to meet this requirement. The

operation is no different from the standard system.

24

Figure 11: System 23 with cooling jacket

All process connections to the valves are welded. The sample chamber is split in two

parts and welded to the valves. It is flanged together to enable access to interior

parts of the valve for maintenance. The application of a split-body sample chamber

can come in handy when the sample material could inadvertently solidify and the

sample chamber must be cleaned regularly.

25

Chapter 2: Gas analysis

2.1 Block diagram for a typical gas analyser

The ability to analyse one or more components of a gas sample depends on the

availability of suitable detectors which are responsive to the components of interest

in the sample and which can be applied over the required concentration range.

A block diagram of the components of a typical gas analyser is shown below.

The sample is taken into the instrument either as a continuous stream or in discrete

parts of the stream and is adjusted as necessary in the sampling unit to the

temperature, pressure and flow-rate requirements of the remainder of the system.

Any treatment of the sample, for example separation of the sample into its

components, removal of interfering components or reaction with an auxiliary gas is

carried out in the processing unit. Finally the sample is passed to the detector. The

signal from the detector is amplified if necessary and processed to display or

record the concentration of the components of interests in the sample.

In many gas analysers the time lag between sampling and analysis is reduced to a

minimum by taking a continuous stream of sample at a relatively high flow rate and

arranging for only a small portion to enter the analyser, the remainder being

bypassed to waste or returned to the process.

26

The need for oxygen analysers occurs in various industrial situations, e.g. flue

(exhaust) gases and Oxygen dissolved in various solutions.

There are a number of instruments available for oxygen analysis:

(i) Magnetic wind Oxygen analyser

(ii) Magnetodynamic Oxygen analyser

(iii) Zirconium Oxide analyser

Paramagnetism and Diamagnetism

All materials show some effect when placed in a magnetic field. In most materials

the effect is very small and intensive magnetic fields are required to make the effect

measurable.

Substances which are magnetised in the opposite direction to that of the applied

magnetic field are called diamagnetics.

Substances which are magnetised in the same direction as the applied field are

called paramagnetics. It is the atomic structure of substances which determines

whether it is diamagnetic or paramagnetic.

When suspended in a magnetic field a rod of paramagnetic substance tends to move

so that its length is along the direction of the magnetic field. Paramagnetic

substances are iron, steel, nickel, cobalt, etc.

When a rod of diamagnetic materials such as Bismuth is placed in a magnetic field it

tends to position its length at right angles with the magnetic field. The degree of

magnetisation of a substance is dependent on the intensity of the magnetic field.

Volume susceptibility of the material = Intensity of magnetisation

Intensity of the magnetic field

The susceptibility of diamagnetic substances is independent of temperature.

The susceptibility of paramagnetic substances decreases with increase of

temperature.

Field Field

S N N S N N

Diamagnetic

material

Magnet Paramagnetic

material

Magnet

27

Most substances are diamagnetic and the value of the susceptibility is usually small.

Paramagnetics have larger values.

Oxygen, unlike any other gas is paramagnetic.

The susceptibility of a paramagnetic material is positive whilst that of a diamagnetic

material is negative.

2.2 Magnetic Wind analyser

This type of analyser depends on the fact that oxygen as a paramagnetic substance

tends to move from the weaker to the stronger part of a magnetic field and that the

paramagnetism of Oxygen decreases as the temperature is raised.

volume susceptibility = c therefore

density T -

volume susceptibility = c x density

T -

where T = absolute temperature

c, = constants

but for a gas the density is proportional to ⁄

therefore the volume susceptibility = c

T2 - T

Note: The volume susceptibility of O2 is inversely proportional of T.

28

Principle of operation of the magnetic wind instrument

The measuring cell consists of a circular annulus with a horizontal bypass tube on

the outside of which are wound two identical Platinum heating coils. These two coils

form two arms of a Wheatstone bridge circuit, the bridge being completed by two

external resistances (R). The coils are heated by means of the bridge current

supplied by a dc source of about 12V.

The winding on the left is placed between the poles of a very powerful magnet.

When a gas sample containing Oxygen enters the cell, the Oxygen tends to flow into

the bypass tube. Here it is heated so that its magnetic susceptibility is reduced. The

heated gas is pushed along the cross-tube by other cold gas entering at the left.

This gas flow cools the filaments, the left coil more than the right, and so changes

their resistance.

The change in resistance unbalances the Wheatstone bridge and the produced emf

is measured to give a signal which is proportional to the Oxygen content in the

sample being analysed.

Pt wires

12Vdc

Bypass tube

29

Errors associated with the Magnetic Wind analyser

1) The instrument is temperature sensitive. An increase in temperature causes

a decrease in the out-of-balance emf of about 1% per Kelvin. This can be

automatically compensated by a resistance thermometer placed in the gas

stream near the cell, which automatically adjusts the potential.

2) The calibration depends on the pressure of the gas in the cell. If the gas

pressure of a sample is higher than the pressure employed in calibration then

the resultant % Oxygen will be high (more Oxygen is present at higher

pressure for a particular volume).

If the pressure of operation is not the same as the calibration pressure a

correction factor must be applied. Automatic corrections can be achieved by

causing an appropriate change in the temperature of the measuring cell.

3) Another error arises from the fact that the analyser basically depends on the

thermal conductivity of the gas passing through the cross-tube. Any change

in the composition of the gas mixed with the oxygen changes the thermal

balance and so gives an error signal. This is known as the carrier-gas

effect.

To a first approximation the out of balance emf is given by: e = KC0

where e = out of balance emf

C0 = the Oxygen concentration

K = factor which varies with the composition of the carrier gas

The value of K for a mixture can be calculated by summing the partial products.

K (

)

Where CA, CB are the percentage concentrations of components A and B and KA and

KB are the corresponding values of K.

Typical values of K are given below:

Gas K Gas K

Ammonia 2.21 Nitrogen 1.00 Argon 0.59 Nitric Oxide 0.94 Carbon Dioxide 1.54 Nitrous Oxide 1.53 Carbon Monoxide 1.01 Oxygen 0.87 Chlorine 1.52 Sulphur Dioxide 1.96 Helium 0.59 Water Vapour 1.14 Hydrogen 1.11 Propane 3.50

30

Example

A sample of flue gas contains N2, CO2, CO, C3H8 (Propane), H2O vapour and O2. The

Oxygen analyser used to measure the % O2 has been calibrated using a similar

carrier gas i.e.

N2 50% changed to 55% CO2 10% changed to 6% CO 5% changed to 9% C3H8 1% changed to 3% H2O 23% changed to 16% O2 11% changed to 11%

Calculate the % error if the carrier gas composition changed as shown above.

Carrier gas composition

Calibration New Value N2 50 55 CO2 10 6 CO 5 9 C3H8 1 3 H2O 23 16 89 89

Carrier gas % (calibration value)

N2 50 50 x 100 = 56.17 89

0.56

CO2 10 10 x 100 = 11.23 89

0.11

CO 5

5 x 100 = 5.61 89

0.06

C3H8 1 1 x 100 =1.12 89

0.01

H2O 23 23 x 100 = 25.84 0.26 89 89 100

K = 0.56 x 1.00 + 0.11 x 1.54 + 0.06 x 1.01 + 0.01 x 3.5 + 0.26 x 1.14 =

100

K = 0.56 + 0.169 + 0.060 + 0.035 + 0.296 = 1.12 = 0.0112

100 100

e = K C0 = 0.0112 x 0.11 = 0.123V = 1.23 mV

Carrier Gas

% concentration

carrier gas

components

31

Carrier gas % (new value)

N2 55 55 x 100 = 89

61.79 0.62 0.07 0.1 0.03 0.18

CO2 6

6 x 100 = 89

6.74

CO 9

9 x 100 = 89

10.11

C3H8 3

3 x 100 = 89

3.37

H2O 16 89

16 x 100 = 89

17.97 100

K” = 0.62 x 1.00 + 0.07 x 1.54 + 0.1 x 1.01 + 0.03 x 3.5 + 0.18 x 1.14 =

100

K” = 0.62 + 0.1078 + 0.101 + 0.105 + 0.205 = 1.139 = 0.01139

100 100

e” = K” C0 = 0.11 x 0.01139 = 0.1253V = 1.25 mV

% error = e” – e x 100 = 1.25 – 1.23 x 100 = 1.6%

e” 1.23

% error = new value – calibration value x 100

new value

%

concentration

carrier gas

components

32

2.3 Magnetodynamic Oxygen analyser (Pauling cell) ‘Dumb-Bell’

The detecting element consists of a short bar with a sphere at each end forming a

dumb-bell shape. It is fabricated from fused silica, the spheres filled with nitrogen

gas and then sealed. The dumb-bell operates in a symmetrical non-uniform

magnetic field, and is located in the analysis cell on a rare metal suspension strip.

When the gas being analysed does not contain Oxygen, the dumb-bell, due to the

diamagnetic characteristic of the filling, moves away from the most intense part of

the magnetic field. Light rays from a lamp are reflected from a small mirror, attached

to the suspension strip, and fall on a split photo-cell. At the electrical zero position,

the light received by each half of the photo-cell is the same.

When the sample contains Oxygen, the dumb-bell spheres are pushed further out of

the magnetic field because of the paramagnetic property of the Oxygen. This

movement is sensed by the photo-cell which changes the signal to the balanced dc

amplifier. The amplifier output, besides being used to operate a readout meter, is

fed back to a coil wound on the dumb-bell, the plane of the coil being at right angles

to the axis of suspension. Thus the torque, due to the oxygen content of the sample,

is balanced by a reaction torque created by the feedback current.

33

The magnetic properties of Oxygen vary with temperature (the paramagnetism of

Oxygen decreases as the temperature is raised) and to avoid errors due to

temperature changes, the analysis cell is sometimes operated in a thermostatically

controlled compartment. Alternatively, a temperature compensator is included in the

measuring circuit. The pressure of gas in the analysis cell must be constant, and

correction to the instrument reading must be made should this pressure vary from

the specified value.

Installation Requirements

(i) firm, vibration free mounting

(ii) sampling system to be absolutely free from oil and grease

(iii) stabilised electrical power supplies as appropriate

(iv) all the equipment must be installed in a safe area

Calibration is done by comparison with gas mixtures of known Oxygen content.

34

2.4 Zirconium Oxide potentiometric probe

At temperatures above 350°C, Zirconium Oxide (Zr undergoes the following

chemical reaction:

Zr (solid) Zr4(+) (solid) + (gas) + 4e(-) . . . equation 1

oxidised state reduced state

How far the reaction moves to the right depends on:

(a) the temperature and

(b) the pressure of Oxygen present in the atmosphere around the hot Zirconium

Oxide.

A pure Zirconium (Zirconium Oxide) tube maintained at high temperature will

develop a potential between its surfaces. This potential is a function of the partial

pressures of Oxygen P1 & P2 which is in contact with the surfaces of the tube.

This is the principle involved in the Oxygen analyser shown below.

Alumina

tube

Inner electrode

Outer electrode

Zirconia tube

Alumina tube

Reference Gas @partial pressure P2

To high Impedance

potentiometer

Measured gas @partial pressure P1

35

The potential developed is given by the Nernst equation:

(

) equation 2

where R = gas constant (8.314 Joules)

T = temperature in °K

= number of electrons involved in the reaction

F = Faraday constant (96500 Coulombs)

The potential difference between the surfaces of the Zirconia tube can be measured

by platinum electrodes in contact with two surfaces as shown in previously. This will

give a measure of the ratio of the partial pressures of the Oxygen inside and outside

the probe.

The NERNST equation then becomes:

equation 3

where P1 and P2 are the external and internal partial pressures of ions

respectively.

If dry instrument air (20.9% Oxygen) is fed into the probe, the partial pressure of

Oxygen inside the tube may be regarded as constant, therefore P2 = 0.209.

equation 4

The probe must be used at temperatures in excess of 600°C and the reading must

be corrected for temperature, e.g. using a thermocouple. Some probes have a

temperature controlled heating element to maintain temperature constant.

Standard instruments have ranges of Oxygen concentration of 20.9 to 0.1% and can

measure Oxygen with accuracy of better than ± 10% of the reading.

To calibrate the potentiometer it is necessary to substitute values of the minimum

and maximum P1 (see equation 4) and then calculate the calibration emfs. These

emfs are then input to the potentiometer from a voltage source to represent the

respective O2% readings.

36

Example

Calculate the calibration emfs required to use a ZrO2 Oxygen probe to measure the

O2% concentration in the range, 0.01% to 0.001%. Assume the temperature is

750°C and dry instrument air is fed into the probe.

Given that: R = gas constant (8.314 Joules)

T = temperature in °K

= number of electrons involved in the reaction

F = Faraday constant (96500 Coulombs)

(

When the pressure of O2, P1 is 0.01% 0.0001

When the pressure of O2, P1 is 0.001% 0.00001

The calibration emf values are therefore:

for 0.01% O2 is 168 mV

for 0.001% O2 is 218 mV

NOTE:

When operating at a new temperature

37

2.5 Single beam IR analyser

The Servomex 2500 series is a multi-wavelength process analyser suitable for

monitoring up to three components in a sample. It can be used for stack gas

analysis. The analyser is controlled using an on-board microprocessor and can be

operated using a simple control panel which is mounted on the analyser.

The gas or liquid sample to be analysed must be passed continuously through the

analyser’s sample cell. This analyser is designed for continuous 24 hour per day

operation and should not normally switched off. Some versions of this type of

analyser can be used in hazardous areas.

The analyser consists of two cast end assemblies with hinged opening covers

connected by a rigid mounting beam, or chassis. The sample cell is mounted

between the two end assemblies, as shown above, and is removable for cleaning.

The source end contains the soft UV or IR source which generates a broad beam of

energy. The chopper box, situated inside the source end assembly, contains filters to

select the appropriate wavelength and a rotating chopper wheel which

splits/alternates the beam into; one for reference and one for the sample.

The IR or UV visible beam passes through the sample cell, some of its energy is

absorbed by the sample components. The remaining energy falls on the detector in

the detector end and is compared to a reference beam. The difference in energy

between the sample and reference beams is an indication of the concentration of the

components in the sample.

38

Chapter 3: Density measurement

Density is the mass per unit volume of a material at temperature T.

Density = Mass at temperature T (kg/m3) Volume The temperature must be specified as density is temperature dependent (i.e. the volume changes). Relative density is the ratio of the density of the sample compared to the density of

water (e.g. 1 gram/cm3 at same temperature T). Relative density is a ratio and has no units. Relative density = Density of sample (no units) Density of water at temperature T Measurement of density The measurement of density is based on the relationship Pressure = density x head

P = x g h If the head is constant then pressure is proportional to density. There are a number of types of instrument using this principle. These will be considered as Type 1 instruments.

39

3.1 Type 1 – Static pressure operated instrument Two techniques are available:

(a) Used where liquid level is constant and open to the atmosphere. This technique measures the pressure in the liquid a known distance below the surface. The air flow is regulated at a chosen small constant value. Gas pressure builds up in the stand pipe until it is equal to that due to the head of liquid above the bottom of the stand pipe.

Since g h is constant the pressure is proportional to density. The pressure can

be recorded or used to control a system to maintain the density of the liquid constant.

Pressure gauge

Stand pipe

Flow meter

Compressed air supply

Flow control valve

Fluid

40

(b) This density measurement system can be used where liquid level is not constant or cannot be maintained at atmospheric pressure.

Two stand pipes are immersed in the liquid so that the lower end of one is h

below that of the other. The shorter stand pipe is set below the minimum level of the liquid. Small constant flow rates are established at the flow meters. The difference in pressure build up between the two stand pipes is equal to

the pressure of the column of liquid (P = g h x density). Since h is constant

the pressure indicated by a differential pressure gauge is proportional to density.

3.2 Type 2 - Weighing tube instrument Since density is proportional to the mass of a substance in a given volume then the mass of a fixed volume of fluid can be used as a measure of density (V = m x ρ). In practice the mass is measured as weight (assuming g to be constant 9.81 m/s2). This is the principle of the weighing tube instrument illustrated below.

Flowmeters

Compressed air or gas

Differential pressure

gauge

Fluid

41

The process liquid flows through the weight tube and is counter balanced by the counter weight. When the weight of the tube increases due to an increase in the density of the process liquid the weight tube turns clockwise raising the flapper valve and allowing more air to escape from the nozzle. The resulting decrease in pressure above the diaphragm in the relay allows the valve mechanism to rise and increases the pressure in the feedback bellows and output connections. The feedback bellows pressure increases until sufficient force is generated to restore the original balance. The weight tube loop is effectively continuously balanced and barely moves more than a few seconds of arc. Because the bellows area and all the pivot lengths are fixed, the pressure increase at the output is directly proportional to the increase in density of the process liquid and is a measure of the density of liquid flowing in the tube. Calibration is achieved by hanging weights on the end of the empty weight tube loop which represent specific liquid densities.

Fluid in

Fluid out

Weigh tube

Pivot

Pivot Range adjuster

Hook for calibration pan

Pivot

Flapper

Nozzle

Nozzle control valve

Valve mechanism

(Equaliser) Regulated air supply

(1.3 bar)

Output to measuring

device or controller

(0.2 – 1.0 bar)

Silicone liquid

damping dashpot

Feedback bellows

Counter balance

Calibration balance pan and

weights

42

3.3 Type 3 - Buoyancy types (a) Totally immersed displacer type The upthrust on the totally submerged displacer or plummet is counter

balanced by the weight of the plummet and the weight of the calibrating chain. The plummet weight is such that at the middle of the density range it will support half the weight of the calibrating chain with the other half supported by the attachments with the measuring chamber.

The position of the plummet changes with the change of upthrust which takes

place with change in density of the liquid. Different plummet positions alter the inductance between the windings of a

differential transformer, altering its output (-5mV to +5 mV). For different density ranges a different chain-plummet assembly is used.

Displacer or plummet

Ferro-magnetic core

Primary

Measuring chamber

Liquid flow

Calibrating chain

Thermocouple or RTD for

temperature compensation

Inductance pick-up

(output to recorder)

Secondary

43

(b) Recording hydrometer type (free floating partially immersed) The control of liquid level and flow rate are important. The position of the plummet changes with the change of density and varies the voltage output of the differential transformer.

3.4 Type 4 - Absorption of radiation This is used to measure the density of materials flowing in a pipe. The pipe must be full. The technique is used on slurries and solutions in process plants.

Free floating plummet

weighted at the bottom

Inductance pick-up

Secondary

Primary

Liquid out

Liquid in

Lead shield

Detector

Source CS-137

Shutter

to measuring circuit

44

A -ray source (Caesium-137) emits radiation to pass through the flowing material. If the detector measures the intensities of radiation before and after the material flow,

Io and I, then density is given by

= I ln ( Io)

D ( I ) Where: D = pipe diameter

= mass absorption coefficient of radiation of the material flowing through the pipe This technique can measure densities and specific gravities very accurately.

3.5 Type 5 - The PAAR oscillating tube density meter The principle of this instrument is based on the change in the natural frequency of oscillation of a hollow U-tube oscillator when filled with liquids or gases of different densities. The liquid or gas increases the mass of the U-tube by an amount equal to the volume of the tube (internally) multiplied by the density of the liquid or gas. It is this increase in mass of the oscillator which changes its frequency of oscillation. Since volume remains constant, density is proportional to mass which is related to the frequency of oscillation. When the U-tube is oscillating in a plane perpendicular to the plane of the tube, with the open end of the U-tube firmly anchored, then the frequency of oscillation f is given by: f = 1 c

2 m + V

where m = mass of the U-tube

= density of liquid or gas V = volume of U-tube c = elasticity constant of the tube Therefore T (period of oscillation) is given by:

T = 2 m + V since f = 1

c T

T2 = (2)2 ( m + V) = 42m + 42V

c c c

45

If 42m = B and 42V = A c c

then T2 = B + A Therefore equation 1

and for two liquids or gases of different densities 1 and 2 and periods T1 and T2. equation 2 Since A and B contain the elasticity coefficient of the tube and its mass and volume, they are apparatus constants which can be determined by making period measurements on two standards of different but accurately known densities and use the values in equations (1) and (2) to calculate A and B.

= T2 - B

A

1 - 2 = 1 (T12 - T2

2)

A

Sample in

Sample out

Energy input

oscillator

Energy output

oscillator

Sample Flow +

Pressure are fixed

46

Calibration The most common calibration fluids used are ‘dry air’ and pure ‘de-ionised water’. Their densities at the particular temperature of interest can be found from tables. The ‘dry air’ is passed through the oscillating tube and their periodic time (Tair ) is measured and recorded by the output unit. The pure water is then passed through the oscillating tube and its periodic time (Tpure water) is measured by the output unit and recorded. These two values together with their accurate densities at the operating temperature, are fed into the data processor which prints out the values of A and B. These are then fed into the control potentiometers. Example Given that for a particular sample tube of pyrex glass

Water (25 C) 1 = 0.99704 g/cm3 T1 = 6.4281 sec

Air (25 C) 2 = 0.00116 g/cm3 T2 = 4.7573 sec

Calculate A and B and use these to calculate the density at 25 C of an alcohol

sample which has an oscillation period of 6.2953 sec.

1 - 2 = 1 (T12 - T2

2)

A

0.99704-0.00116 = 1 (6.4281x10-6)2 – (4.7573x10-6)2 therefore A=1.87x10-11 A

1 = T12 - B

A

0.99704 = (6.4281x10-6)2 – B therefore B= 2.267x10-11 1.87x10-11 Given T for alcohol is 6.2953µsec then its density is

= T2 - B = (6.2953x10-6)2 – 2.267x10-11 = 0.9070g/cm3

A 1.87x10-11

47

Chapter 4: Humidity measurement

4.1 Applications of humidity measurement 1 Paper stores - dry paper becomes statically charged by draught and/or

vibration. Discharge causes a fire hazard. Other effects of drying or condensation due to the wrong humidity cause a change in the size of the paper and curling of edges.

2 Inflammable powder producing factories and stores - static builds up in clouds

of inflammable powder and air unless the humidity is high enough to prevent this. Massive explosions may result in sugar factories, flour milling factories and coal mines.

3 Distillation processes - processes involving the evaporation of organic liquids

followed by cooling to produce condensation of the distillate vapours leads to water condensation and therefore contamination of the product. The humidity must therefore be kept below a certain value in order that the dewpoint remains below the condensation temperature of the distillate. This is important in the oil and petroleum fractions.

4 Workforce comfort - to work without distress the relative humidity of the

workplace must be optimised. Air conditioning systems include humidity measurement.

5 Natural gas - humidity or dewpoint measurements are made to check the

moisture content of gas supplies.

4.2 What is humidity? Humidity is caused by the presence of water vapour in the air (or other gaseous environments). The water vapour gets into the atmosphere mainly by natural evaporation of rivers, lakes, ponds, seas. Evaporation occurs at a temperature well below the boiling point of water. Water molecules with the highest kinetic energies are moving fast enough to break through the surface skin of the liquid water. Evaporation at low temperatures is slow. If the temperature of the water increases the kinetic energy of the molecules increases and a higher proportion of the water molecules has sufficient energy to break into the atmosphere. Once the water molecules become vapour they behave like gases and exert a pressure called the vapour pressure. Each gas present in a mixture exerts a pressure proportional to its percentage by volume in the mixture and is equal to that pressure that it would exert if it alone occupied the total space available. This pressure is called its partial pressure.

48

Example A sample of air consists of: 79% N

2 20% O2 0.9% inert gases 0.03% CO2 0.025% H2O 0.045% other gases

Assuming that the total pressure of the sample of air is 1bar at 20C, the partial pressures of the gases present are: 0.79 bar of N2 (divide by 100) 0.20 bar of O2 0.009 bar of inert gases 0.0003 bar of CO2 0.00025 bar of H2O 0.00045 bar of other gases

Sum of partial pressures = Total pressure Unlike gases, which can have any percentage composition, there is a limit to the vapour pressure of water which can be achieved in air or other gas samples at a particular temperature. This maximum is the saturation vapour pressure at a given

temperature. If air contains water vapour at a particular temperature and the water vapour is at the saturation vapour pressure, any slight cooling will cause some of the vapour to condense and form dew. The temperature at which dew deposits from a moist air or gas sample is called the dewpoint of the air or gas. Relative humidity is a fraction, decimal or percentage, which compares the amount

of water present with the amount of water required to saturate the air or gas at the same temperature. Relative humidity = Mass of water vapour present in a given volume of gas Mass of water vapour necessary to saturate the same volume of gas at the same temperature Many substances have the property of absorbing or releasing moisture very quickly to an extent dependent on the partial pressure of water vapour in the atmosphere. These are known as hygroscopic substances. As the moisture content changes, its properties such as resistance, conductance or capacitance also change. This permits electrical measurements of humidity to be made.

49

Measurement of humidity is based upon the principles already discussed. Types of humidity measuring instruments fall into one of the following categories:

(i) Dewpoint instruments (ii) Wet and dry bulb hygrometers (iii) Hair Hygrometers (iv) Electrical methods

4.3Thermoelectric dewpoint instrument This instrument is based on the principle that a clear mirror reflects a beam of light in a predictable manner whereas a mirror with dew on its surface disperses light in an irregular manner. A photo sensitive resistor is placed in such a position that the detection of scattered light due to the formation of dew on a cooled mirror will be detected immediately and can be used to stop the refrigeration unit (cooler) attached to the mirror. The mirror warms up and the dew clears so that light is no longer scattered onto the photosensitive resistor and the refrigeration unit starts to cool the mirror once again.

The temperature of the mirror quickly stabilises within 0.25C of the dewpoint and is monitored by a thermistor which operates a meter or controller. Such instruments are widely used to control drying or humidifying systems. The thermo-electric cooling unit is of the Peltier type which operates on the principle that when a current is applied to a junction of dissimilar metals its flow in one direction causes cooling and in the other direction causes heating.

Servo control

amplifier

Gas In

Light bulb

Collimating slots

Signal input

Stabilised

supply

Photosensitive resistor

mirror Gas out

Thermo-electric

(Peltier) cooler

o/p to temperature

recorder or humidity

controller

thermistor

Current output

50

4.4 Wet and Dry Bulb Hygrometers Two temperature indicators are used. These may be mercury in glass, or gas or liquid expansion, vapour pressure, bimetallic spiral, electrical resistance thermometers or thermocouples. One indicator has a cotton sleeve which is dipped into a water reservoir. This indicator is known as the wet bulb. The other indicator simply measures the ambient temperature and is known as the dry bulb.

The difference in output of these indicators measures the extent of cooling of the wet bulb due to evaporation. The lower the humidity of the sample the greater is the cooling of the wet bulb and the greater is the difference in the output of the two indicators. The output from the two indicators can be used to give dual temperature readings to give an indication of relative humidity. In simple systems tables are given to relate the temperatures to relative humidity. The temperature difference can be used to give a direct reading of relative humidity (see table below).

Wet bulb

output

Dry bulb

output

Gas out

Water inlet

Water outlet

Gas sample

(filtered in)

51

Dry

Bulb

Depression of Wet Bulb C

C 1 2 3 4 5 6 7 8 9 10 11 12

0 82 64 48 32 14

1 83 68 52 36 21

2 84 68 54 40 25 10

3 84 69 54 43 29 16

4 85 70 56 42 33 20

5 86 72 58 45 32 24

6 86 73 60 47 35 23

7 87 74 61 49 37 26 14

8 87 75 63 51 40 29 18

9 88 74 64 53 42 31 21 11

10 88 77 65 54 44 34 24 14

11 88 77 66 56 46 36 26 17

12 89 78 68 57 48 38 29 20

13 89 79 69 59 49 40 31 23

14 90 79 70 60 51 42 33 25 17 9

15 90 80 71 61 52 44 36 27 20 12

16 90 81 71 62 54 46 37 30 22 15

17 90 81 72 64 55 47 39 32 24 17 10

18 91 82 73 65 56 49 41 34 27 20 13 7

19 91 82 74 65 58 50 43 35 29 22 15 9

20 91 83 74 66 59 51 44 37 30 24 18 12

22 92 83 76 68 61 54 47 40 34 28 22 16

24 92 84 77 69 62 56 49 43 37 31 26 20

26 92 85 78 71 64 58 51 46 40 34 29 24

28 93 85 79 72 65 59 53 48 42 37 32 27

30 93 86 79 73 67 61 55 50 44 39 34 30

32 93 86 80 74 68 62 57 52 46 42 37 32

34 93 87 81 75 69 64 58 53 48 44 39 35

36 94 87 81 76 70 65 69 55 50 45 41 37

38 94 88 82 76 71 66 61 56 51 47 43 39

40 94 88 82 77 72 67 62 57 53 48 44 40

Wet and Dry bulb hygrometer

To find the relative humidity it is

necessary to interpolate between

the column ‘Dry Bulb C’ on the left-

hand side of the table and the

‘Depression of Wet Bulb C’ across

the top of table.

The ‘Depression of Wet Bulb C’ is

the difference between the Dry Bulb

reading and the Wet Bulb reading.

Example:

Dry Bulb Reading 15C

Wet Bulb Reading 11C

Depression of Wet Bulb 4C’

Relative humidity 61%

The following points should be

noted in maintaining Hygrometers:

1 The instrument should be

hung in the shade in such a

position as to allow free

circulation of air around

the Mercury bulbs, but at the

same time, care should be

taken to avoid draughts.

2 Distilled or clean rain water

only, should be used in the

cistern so as to avoid the

wick becoming clogged with

impurities left by evaporation.

3 The wick should be changed

regularly, in order to ensure

a constant flow of moisture to

the wet bulb.

52

4.5 The Gregory balanced temperature hygrometer This measures the temperature difference using a differential thermocouple, one junction wet and the other dry. Electrical windings round the wet bulb are supplied with enough power to maintain the two junctions at the same temperature. The power input is then proportional to the difference in temperature of the two junctions and is used as a measure of relative humidity. At constant room temperature the power required is directly proportional to relative humidity. With all wet and dry bulb systems sample flow-rate is important. It should not be less that 10 ft3/s and should be maintained at a constant value.

differential

thermocouple

heating coil wound around the Wet bulb

dc supply

Range

adjustment

R

R

G

R

G

Dry Wet

dc supply

53

4.6 Hair Hygrometers

Many fibres such as hair, wood fibre, paper and silk change in length with change in humidity. Most hair fibres increase in length with increase in humidity and become shorter

when humidity decreases. A change from 0100% in relative humidity causes about 2.5% change in length of human hair. Because the change is so small, the longer the hair the better. The small changes necessitate an amplification system. The usual method of amplification is a lever system operating a lightweight pointer relative to a scale. One particular model has a mirror attached to the lever which reflects a light spot to a translucent scale and acts as a zero weight optical lever to give greater

amplification and is accurate to 1%. Most hair hygrometers are only accurate to

3% and respond very slowly to humidity changes. Human hair changes its length with variation in temperature (an increase in

temperature of 1C decreases its length by 0.4%). Temperature change can therefore cause significant errors. Calibrations therefore need to be carried out

regularly. Hair becomes brittle at about 65C and suffers from some permanent

deformation below -9C thus limiting the temperature at which it can be used.

Sample out

Pulley

Tensioning spring

Mechanical zero

Sample in

54

4.7 Capacitance Hygrometers Aluminium Oxide is a hygroscopic substance and its behaviour as dielectric changes when moisture changes occur in its surroundings. The capacitance therefore changes due to dielectric changes.

C is the capacitance (F) A is the area of overlap of the two plates (m2) εr is the relative static permittivity (sometimes called the dielectric constant) of the material between the plates (for a vacuum, εr = 1); ε0 is the electric constant (ε0 ≈ 8.854×10−12 F m–1); and d is the distance between the plates (m) The capacitance change of a sensor containing Aluminium Oxide as a dielectric can be used to give a read-out of either relative humidity or dewpoint.

The symmetrical bridge above measures the unknown capacitor Cx by comparison to a standard capacitor Cs.

hygroscopic dielectric

Aluminium Oxide layer

(10µm)

Aluminium base plate

thin film of gold (porous

to gas sample)

to capacitance

bridge circuit

55

4.8 Infra-Red instruments Infra-red techniques are used for moisture measurements based on gaseous, liquid and solid samples. For gas and liquid samples the moisture content is indicated by the amount of infra-red radiation absorbed by the moisture. For gas samples the path length (the distance that light travels through a sample in an analytical cell) must be very long in order to get a significant level of absorption whilst for liquids the path length can be much shorter. Infra-red instruments can be used on-line and are therefore useful in control systems. There are two major classes of devices:

(i) single beam and (ii) double beam.

A double beam spectrophotometer compares the light intensity between two light paths, one path containing a reference sample and the other the test sample. A single beam spectrophotometer measures the relative light intensity of the beam before and after a test sample is inserted. Although comparison measurements from double beam instruments are easier and more stable, single beam instruments can have a larger dynamic range and are optically simpler and more compact. In single beam instruments calibration is achieved by placing known sample in the radiation path and then removing it for measuring purposes.

Single beam instrument In double beam instruments one beam is used for measuring the humidity of the sample whilst the other contains a sealed reference cell.

56

Double beam instrument For moisture measurements of solids reflectance methods are used. As a result this technique gives a measurement of the surface moisture content. The method is suitable for continuous measurement of humidity of material transported on a conveyor belt. The material must, however, be a reasonably good reflecting one.

4.9 Electrical Conductivity Moisture can produce a significant increase in electrical conductivity of a material. Electrical resistance measurements are generally made as an indication of moisture content of timber and plaster. This is usually achieved using a pair of sharp pointed probes which are pushed into the material. The probes are then connected to a Wheatstone bridge. With granular materials on-line measurements can be made by using electrodes which are plates or rollers but these suffer from erosion.

Assume R1 = R2 = R3 = R4 = 120Ω and Vs = 6V dc The Wheatstone bridge is balanced and Vo= 0V When R3=150Ω (ΔR=30Ω) the bridge becomes unbalanced and there is an output voltage Vo=375mV. Vo = Vs ∆R = 6 x 30Ω = 0.375V 4 R 4 x 120

57

4.10 Piezoelectric humidity instrument The oscillation frequency of a quartz crystal coated with hygroscopic material changes with the weight of water absorbed. Very small changes in frequency can be measured. In practice two Quartz crystals are used and wet and dry gas is passed over them alternatively. The frequency of oscillation is about 9MHz. The frequency of the crystal exposed to the wet gas will be lowered and that of the dry gas will be increased. The difference is used as an indication of moisture content. The response time of this type of instrument is better than other instruments. The instrument gives good stability. Regular calibrations are necessary. It is a complex and expensive instrument. Block diagram of Piezoelectric humidity instrument

Oscillator

Oscillator

Mixer F1-F2

Amplifier Clip Diode RC

circuit Meter

Sample gas (Wet)

Reference gas (Dry)

Vent

Vent

Crystals

F1

F2

ΔF

58

Chapter 5: Viscosity measurement

Viscosity measurement and control are of particular importance to the chemical, food

and petroleum processing industries.

5.1 Types of Viscosity

(i) Absolute Viscosity – this is the viscosity that is measured by a system

that is not under the influence of gravity for obtaining the measurement.

F = Force required to maintain a difference of velocity between two planes in

a liquid (N).

= Absolute or Dynamic Viscosity of the fluid. This is constant at a particular

temperature (NS/m2)

A = Area of the plane being moved (m2)

d = Distance between the planes (m)

V = Velocity of the moving plane (m/s)

(

for a particular temperature.

units used CENTIPOISES 1cP = 10-2 P = 10-3 NS/m2

Force F (N) Velocity V (m/s)

Fluid

between

planes

MOVING PLANE

Area A (m2)

Distance d (m)

FIXED PLANE

59

(ii) Kinematic Viscosity – this is the viscosity that is measured by a system

that uses gravity to obtain the measurement. Kinematic viscosity takes

into account the density of the fluid.

Kinematic Viscosity = Dynamic or Absolute Viscosity at t °C

Density at t °C

Units used CENTISTOKES 1cSt = 10-2 St = 10-6 m2/s

(iii) Apparent Viscosity – this is the viscosity of a non-Newtonian liquid. It is

the viscosity that is measured at a single shear rate or single point. It is

expressed in units of POISE.

Viscosity and Temperature

The viscosity of liquids generally decreases rapidly with increase in temperature. All

measuring instruments must therefore be maintained at constant temperature.

The viscosity of gases increases with increase in temperature but not to such a

dramatic extent.

Newtonian fluids

Newton assured that all materials have, at a given temperature, a viscosity that is

independent of the shear rate (shearing that the liquid experiences when present

between two planes). Newton suggested that twice the force will move the fluid

twice as fast. Fluids that behave like this are called Newtonian fluids.

Examples are:

Water

Most Mineral Oils

Gasoline

Kerosene

Most Salt Solutions in Water

60

Non-Newtonian fluids

It was found out later that Newtonian behaviour is only one of several behavioural

characteristics of fluids. Other fluids generally termed as non-Newtonian are sub-

categorised as pseudoplastic, dilatant, plastic, thixotropic and rheopectic.

Pseudoplastic Plastic Thixoplastic Rheopectic Dilatant

printers ink chewing gum silica gel

gypsum in water peanut butter

paper pulp tar most paints candy compounds

glue

molasses

lard

fruit juice

concentrates

When the shear rate in a non-Newtonian fluid is varied, the viscosity will change.

Thus, although any design of viscometer can be used for measuring Newtonian

fluids, this is not the case with non-Newtonian fluids. The speed of a rotational

viscometer for example will have its own effect on the measured viscosity. This

results on what is called apparent or measured viscosity.

61

5.2 Glass Viscometers

They are generally used for the determination of the kinematic viscosity of a fluid in

absolute units. In these a sample of fluid is drawn up a capillary tube by vacuum (or

forced up by pressure) and then allowed to fall under its own weight. A record is kept

of the time taken by the fluid level as it passes from the upper to the lower calibrated

mark on the capillary limb of a U-tube viscometer (see figure below).

The kinematic viscosity is then calculated from

Kinematic viscosity = C. t – B t Where C = calibration constant for the instrument t = time in seconds for the flow between marks B = a constant determined by the capillary diameter and design of the instrument For calibration purposes distilled water can be used as a standard fluid having a kinematic viscosity of 1.0068 cSt at 68ºF (20ºC). The quantities C and B effectively take into account the effects of minor parameters affecting the true (theoretical) rate of flow through the viscometer.

62

5.3 Rotational Viscometers Many types of rotational viscometer have been developed both for laboratory work and commercial use. They are particularly useful in that they can also be employed to measure the instantaneous viscosity of Non-Newtonian fluids at various rates of shear. For commercial use, many rotational viscometers are produced in portable form. A basic form is shown below.

The rotating member in this case is a disc, driven through a torque spring by a

synchronous electric motor. The viscous drag on the rotating disc causes an angular

deflection of the torque spring proportional to the viscosity. Actual spring deflection is

indicated by a pointer and scale, the scale being calibrated to read viscosity directly.

Certain positional errors may be involved, e.g. depth of immersion of the disc and

also angle of immersion. These can be eliminated by employing a measuring

cylinder in which the disc rotates, rather than being dipped directly into the fluid.

These and other limitations have largely been overcome in later designs.

5.4 Falling Body Viscometers

The principle employed in this class of instrument is that the velocity of a body falling

freely through a liquid under the action of gravity is proportional to the fluidity of the

liquid and hence its viscosity.

The general relationship can be expressed as:

V = A φ ( ρs - ρ )

where V = the fall velocity

φ= the fluidity of the liquid

63

ρs = the density of the solid body ρ = the density of the liquid A = constant depending on the position and dimensions of the container and the falling body. The usual from of apparatus is a sphere for the body, falling vertically down a container (falling sphere viscometer). Other shapes of body may be employed; a piston falling through a coaxial cylinder, a box or strip falling down between shearing blocks. Rolling body viscometers are primarily laboratory instruments although a number of designs have been developed for commercial applications.

5.5 Vibrational Viscometers The operating principle involved in a vibrational viscometer is that if an electrically driven vibrating element is immersed in a fluid a damping effect is produced proportional to the viscosity of the fluid. Further, if the element is driven at its natural resonant frequency the damping effect will cause the resonant frequency to decrease, and the amplitude of vibrating to be attenuated. In practice, since the change in resonant frequency is likely to be small, it is more convenient to express the damping effect in terms of change (increase) in resonant resistance which is a larger quantity and easily measured. Viscosity can then be expressed directly as follows: Dynamic Viscosity (η) = ∆Rn 1__ k π fr ρ Where ∆Rn = change in resonant electrical resistance k= constant depending on the physical and electrical properties of the transducer fr = resonant frequency ρ = mass density of the liquid An alternative approach is the measurement of amplitude attenuation of a flat strip resonator vibrating in a viscous fluid.

64

5.6 Arbitrary Viscosity determination Arbitrary viscosity measurements are conducted with short-tube efflux viscometers, the viscosity value being quoted directly in terms of “time” (seconds) for the discharge of a given volume of fluid through an orifice. Three such systems are in general use, all being the same in principle but differing mainly in the volume of flow measures and the orifice dimensions. These systems are: UK – Redwood No 1 or No 2 measuring viscosity directly in seconds. US – Saybolt Universal and Saybolt Furol, measuring viscosity directly in seconds. Continental Europe – Engler instrument. The viscosity with this instrument is

expressed in degrees (ºE) instead of time, representing the ratio of time for 200cm³ of fluid to flow from the viscometer to the time for 200cm³ of water to flow at 20ºC. The Redwood No 1 instrument comprises a cup fitted with an agate jet through which fluid can flow. The cup is surrounded by a water bath so that the fluid can be kept at a constant temperature. A graduated flask is mounted below the jet to collect the fluid as shown below.

Test procedure is to heat the fluid uniformly in a separate container to just above the test temperature, then pour it into a cup through a fine screen mesh. The fluid in the cup is stirred until it has reached the specific test temperature required e.g. 70ºF, 140ºF or 200ºF. A plug is then removed allowing the fluid to flow from the cup

Stirrer handle

Thermometers

Copper bath

Agate jet

Heating

tube

Stirrer

Ball

valve

Oil cup

65

through the jet. The time is recorded for 50ml of fluid to pass through the jet and collected in the graduated receiving vessel. This time (seconds) is quoted as the Redwood viscosity of the fluid at the specific test temperature. There are two forms of this instrument, as mentioned earlier, differing only in the size of the jet used and minor detail. The Redwood No 1 instrument is used for oils and other fluids having viscosity not exceeding 2000 seconds at the test temperature. The Redwood No 2 instrument has a larger jet with a flow rate ten times that of the No 1 instrument. It is normally used only for fluids having a viscosity greater than 1000 seconds at the test temperature on the No 1 scale. The Saybolt viscometers differ only in detail design and are essentially similar in principle. The Saybolt Universal range (SUS) covers approximately the same range as the Redwood No 1. The Saybolt Furol range is approximately similar to that of the Redwood No 2. Standard temperatures for measurement differ however. Saybolt Universal viscosities are determined at 100ºF and 210ºF. Saybolt Furol viscosities are determined at 122ºF. Engler viscosity is determined as a ratio. This ratio which is quoted as Engler degrees can easily be rendered as equivalent seconds by virtue of the fact that the time of flow for 200cm³ of water in this instrument is of the order of 50 to 52 seconds at 20ºC. The “equalivent” seconds value of Engler degrees is thus given closely by multiplying Engler degrees by 50. The kinematic viscosity can be found at follows: Kinematic viscosity = C t – B (see glass viscometers) t where B and C are constants. Typical values of these constants for the three standard instruments are: Instrument B C

Redwood 185 x 10ˉ5 280 x 10ˉ2 Saybolt 194 x 10ˉ5 236 x 10ˉ2 Engler 405 x 10ˉ5 158 x 10ˉ2 There is no exact way to convert arbitrary viscosity values to kinematic viscosity. The equalivents given in Table below will however, be accurate enough for most practical purposes.

66

67

5.7 Continuous Viscosity measurement

Process viscometers (plant viscometers) are designed to be permanently installed in

a pipework or fluid transport system to provide continuous viscosity measurement

rather than individual measurement by sampling.

Group 1

Rotational type viscometers are generally preferred but all the basic types are

capable of being adapted for continuous operation. Ultimate choice may depend on

the fluid involved, process temperature, pressure, accuracy and range. For

continuous measurement a bypass sampling technique is commonly employed,

although this can set its particular problems of ensuring that the sample measured is

fully representative of the main flow, particularly as regards fluid temperature and

thermal equilibrium. Rotational viscometers developed for continuous process

measurement can be of quite robust construction.

68

Group 2

Vibrational type viscometers offer some advantages over rotational types for

continuous process applications, particularly because the amplitude of oscillations of

the immersed element only needs to be very small and so the element is

substantially “static”. It is however likely to be seriously affected by any shocks, for

this reason Reed type viscometers are largely excluded for process work.

Group 3

Capillary or Efflux viscometers can be used for sampling although the sampling rate

is generally too low for automatic control. A typical process viscometer unit is shown

below.

This viscometer is designed for precise sampling and temperature control. The

former is a necessary feature for capillary measurement provided in this instance by

a precision gear pump driven by a synchronous motor. It is suitable both for rugged

construction with flameproof construction for hazardous areas and a wide range of

viscosity measurement since different capillary tubes can be accommodated if

necessary.

Group 4

Falling body viscometers suitable for process sampling may employ a piston (falling

body), periodically raised in a cylinder and allowed to fall. Raising the piston

automatically draws a quantity of fluid into the cylinder filling the volume below the

piston. When tripped the piston falls under the action of gravity displacing the fluid

volume (see diagram below).

69

Micro switches are used for timing, i.e. one micro switch is activated when the piston

is tripped in its raised position, and a second when the piston reached the lowest

point of its travel. The time interval between the micro switch operation is a direct

measure of the viscosity of the fluid.

In practice it is convenient to use a circular chart pen recorder continuously driven by

a synchronous motor, with a second motor traversing the pen across the chart at

constant speed. This second motor is switched by the micro-switches. Hence the

amplitude of the pen movement is a direct measure of time, and so the chart can be

calibrated to indicate viscosity directly proportional to pen movement.

70

71

Chapter 6: Electrical Conductivity

6.1 Applications of conductivity measurement

(i) Water of the highest purity is used for industrial purposes. Conductivity provides the most reliable measurement of water purity. The conductivity of pure water is highly temperature dependent due to the increase in the dissociation of water molecules with temperature.

(ii) The purity of water used in the steam (water circuit) of power stations is particularly important for the prevention of corrosion. Measurement of conductivity would detect any unwanted impurities in the water circuit.

(iii) Conductivity measurements when combined with a separation technique prove to be extremely sensitive and versatile detectors of chemical concentration.

(iv) A technique used to measure the concentration of sulphur dioxide in air is based on the measurement of the change in the conductivity of a reagent before and after it has absorbed sulphur dioxide. The principle of the measurement is to absorb the sulphur dioxide in peroxide solution, thus forming sulphuric acid which increases the electric conductivity of the absorbing reagent.

(v) A rapid continuous measurement of the salt in crude oil before and after desalting is based on the measurement of conductivity of a solution to which a known quantity of crude oil has been added.

6.2 Measurement of electrical conductivity Conductance (G) is the ability of a portion of electrolyte to carry current. Conductance is the reciprocal of resistance.

G = 1 R Unit of conductance is the Siemens (S). The resistivity (ρ) of a substance if the resistance measured between two opposite faces of a unit cube of material. For conductors is quoted in µΩmm or Ωm. Conductivity (σ) is the reciprocal of resistivity and is usually quoted in S/cm.

σ = 1 ρ

Liquids which conduct electricity are called electrolytes and the concentration or

purity of these is often determined by conductivity measurement.

72

In practice the conductance between the two electrodes of a cell is measured and then the conductivity is calculated by applying a factor called the cell constant. The

cell constant (α) varies directly with the distance between the electrodes and inversely with the area of the electrodes. Conductivity of solutions Liquid electrolytes are conductors by virtue of the amount of the ions they contain. The electrical conductivity of solutions will vary with pH and pOH since these are related to the hydrogen ion concentration and hydroxyl ion concentration. Strong acids and alkalis will have a higher conductivity than weak acids and alkalis in aqueous solutions since they form more ions. The electrical characteristics of an electrolyte are much the same as for solid conductors. The conductivity of electrolytes varies greatly with concentration because as dilution increases the number of ions per unit volume decreases. Temperature also affects the conductivity of an electrolyte in the same as it does with solid conductors. As temperature rises more ions are formed and they travel faster in the solution. Conductivity therefore increases with temperature. In order to measure the effect of concentration the term molar conductivity (Λ) is used.

Λ = σ (S cm²/mol)

c where c is the concentration in mol/cm3. The measurement of electrical conductivities can also be a means of determining the concentration of a solution. Measurement of electrical conductivity is achieved by using conductivity cells. Conductivity cells allow a small, usually alternating current to be passed through a precise volume of liquid whose conductivity we wish to measure. The resistance of the solution between two electrodes of fixed shape and constant distance apart is measured. The conductivity of the solution can be calculated by the formula:

σ = α R

where α is the cell constant (cm-1), and R is the resistance between the electrodes. In order to be able to measure the full range of conductivities of aqueous solutions it is necessary to use cells of different cell constants within the range of 0.01 to 100.

73

Guide to cell constant for known conductivity range

Conductivity range (µS/cm) Cell constant (cm-1)

0.05 – 20 0.01

1 - 200 0.1

10 - 2000 1

100 - 20000 10

100 - 200000 50

In order to measure the conductivity of a solution accurately it is necessary to know the cell constant accurately. The cell constant can be determined using one of two methods:

(i) Measuring the conductance when the cell is filled with a solution whose conductivity is accurately known,

(ii) Comparing a measured conductance with that obtained from a cell of known cell constant when both cells contain the same solution at the same time.

The only solutions whose conductivities are known with sufficient accuracy to be used for reference purposes are aqueous solutions of Potassium Chloride (KCl).

Standard solutions for cell calibration

Solution σ (t=18ºC) σ (t=25ºC)

(g KCl/1000g of solution) (S/m) (S/m)

(A) 7.4191 1.1163 1.2852

(B) 0.7453 0.12201 0.14083

The electrode materials vary with the corrosive nature of the electrolyte. Materials used are platinum (black coated), graphite, stainless steel, nickel etc. They are generally mounted flush so that flow is not restricted.

74

6.3 Conductivity Cells Conductivity cells are of two main types: (i) Dip type (ii) Flow-through type Whichever type is employed, the body on which the electrodes are fixed is made of a plastic insulator material such as epoxy resin, PTFE etc.

Dip Type Cell

One common form of Dip type cell consists of a satinized stainless steel rod electrode surrounded by a cylindrical stainless steel electrode, having holes to permit the sample to flow freely through the cell. This is surrounded by an intermediate cylinder also provided with holes, and two o-rings which together with the tapered inner end form a pressure tight seal onto the outer body when the inner cell is withdrawn for cleaning, so that the measured solution can continue to flow and the cell will be replaced without interruption of the process. The outer body is screwed into the line through which the measured solution flows. The cell may be used at 110ºC up to 7bar pressure. Figure below shows the inserted cell as it is when in use and the withdrawn measuring element with the intermediate sleeve forming a seal on the outer body.

75

Flow-through Type

Many manufacturers offer a type of flow through conductivity cell with annular graphite electrodes, one form of which is shown below.

It consists of three annular rings of impervious carbon composition material equally spaced within the bore of an epoxy resin moulded body. Conduction through the solution within the cell takes place between the central electrode and the two outer rings, which are connected to the earthed terminal of the measuring instrument. Electrical conduction is confined entirely within the cell, where it is uninfluenced by the presence of adjoining metal parts in the pipe system. The use of impervious carbon composition material for the electrodes eliminates polarisation errors and provides conducting surfaces that do not require re-platinization, other than periodic cleaning with a bottle brush. Typical operating temperature and pressure limits for this type of cell are 100ºC and 7bar.

76

6.4 Instruments for conventional a.c. measurement The conductance of a conductivity cell may be measured by: (a) Wheatstone bridge methods or (b) by direct measurement of the current through the cell when a fixed voltage is applied. Wheatstone bridge method

A self-balanced Wheatstone bridge a.c. powered to prevent polarisation effects, with either manual (see figure a) or automatic temperature compensation (see figure b) may be used to measure a cell’s conductance. Any out of balance detected by the detector causes the contact of the slide wire at B to be moved to restore balance at the same time indicating and/or recording the conductance of the cell or initiating and act or process control.

Direct measurement method

This method measures the current in the cell. The current is directly proportional to the conductance so the output from the current amplifier is applied to the indicator and recorder. Temperature compensation is achieved by connecting a manual temperature compensator in the amplifier circuit as shown below.

77

6.5 Choice of measurement parameters

Since solutions can have a wide range of conductivities due to a wide range of concentrations it is important that certain parameters are adjusted so that measurement is as sensitive as possible. These parameters are: (i) the value of cell constant and (ii) the value of a.c. frequency. Choice of conductivity cell In dilute aqueous solutions of ions, there will be very few ions between the cell

electrodes and the resistance of the cell will be very high. In concentrated aqueous solutions of ions, there will be large numbers of ions between the cell electrodes and the resistance of the cell will be low. No measuring instrument is capable of measuring with the same sensitivity over such a wide range of resistance. This is overcome by selecting a cell with a constant which with a particular solution will have a resistance most appropriate for high instrument sensitivity. Thus, to measure conductivity in dilute solutions, a cell is chosen which has large electrode area and small distance apart of the electrodes i.e. low value of cell constant. To measure conductivity in concentrated solutions, a cell is chosen which has small electrode area and large distance apart of the electrodes i.e. high value of cell constant. Choice of a.c. frequency

Low frequencies must be used for dilute solutions whilst higher frequencies may be used for concentrated solutions. Polarisation errors

Electrolytic discharge of ions occurs at the electrodes causing errors due to polarisation of the electrodes by the deposited materials and due to local change of ionic concentration. This can be overcome by using an a.c. source of power so that no discharge takes place because the anode and cathode change identity at the rate of alteration. The higher the concentration of ions the higher the a.c. source frequency should be to ensure that no ionic discharge takes place. The polarisation of the electrodes results from a back emf which these deposits generate at the electrode surface, which decreases the applied emf by an amount roughly proportional to the ionic concentration and texture of the deposit. Platinum electrodes are usually coated with platinum black by electrolysis as this has two effects. It makes a better contact with the aqueous solution so that electron transfer at the electrode surface is improved and it increases the surface area (makes it rougher) so that quite small electrodes have a much larger surface area than their superficial measurements indicate.

78

Chapter 7: Electrochemical methods

7.1 Classification of electrochemical methods

An electrochemical method can be defined as one in which the electrical response of

a solution (sample) is measured.

The apparatus used can be divided as follows:

(a) The electrolyte which is a chemical solution capable of conducting current

(b) The measuring or external circuit which is used to apply and to measure

electrical signals (currents, voltages).

(c) The electrodes which are conductors that serve as contacts between the

measuring system and the electrolyte.

Electrodes are classed as anodes and cathodes.

At the anode, oxidation occurs i.e. electrons are removed from the electrolyte and

pass into the measuring circuit. At the cathode, reduction occurs i.e. electrons flow

from the cathode into the electrolyte (electrolyte gains electrons).

GALVANIC CELL

Working or indicator electrodes are those at which a reaction being studied is

taking place.

Reference electrodes are those which maintain a constant potential irrespective of

changes in current.

Counter electrodes are those which serve to allow current to flow through the

electrolyte and they do not influence the solution.

NOTE: The anode is positive in a device that consumes power (electrolytic cell), and the

anode is negative in a device that provides power (galvanic cell).

79

Electrochemical methods can be divided into two categories:

(i) those involving no net current flow (potentiometric) and (ii) all others.

In potentiometry we measure the equilibrium thermodynamic potential of the solution

without causing electrolysis or current drain because this would affect the existing

equilibrium.

In all other methods, a voltage or current is applied to an electrode and the resultant

current flow or voltage change of the solution (sample) is monitored.

Electrochemical cells

An electrochemical cell can be defined as two conductors or electrodes, usually

metallic, immersed in the same electrolyte solution, or in two different electrolyte

solutions which are in electrical contact.

Electrochemical cells are classed into two types:

(i) Galvanic (Voltaic) cell is one in which electrochemical reactions occur spontaneously when the two electrodes are connected by a conductor. These cells are often employed to convert chemical energy into electrical energy.

(ii) Electrolytic cell is one in which chemical reactions are caused to occur by the introduction of an external voltage greater than the reversible (galvanic) voltage of the cell.

80

7.2 The pH and its measurement Measurement of pH Values Colorimetric Method Some substances change colour when added to a solution of a particular pH value. For example litmus, a colouring material, turns red in acid and blue in alkali, but it should be understood that there is a dead zone in the region near the neutral point. The colour change is gradual and litmus papers are fully red at 4.5pH and fully blue at 8.3pH. There are a number of different dyestuffs which undergo a distinctive colour change and they are sufficient to cover the major portion of the pH scale. These substances are usually called indicators. A more complex chemical solution of a mixture of dyes, has a different colour for each value on the pH scale. This solution is called a universal indicator. To make a measurement, a small quantity of the indicator is added to a sample of the liquid being tested. The resulting colour is then compared with the colour standards which are normally supplied with the indicator. Colour matching techniques are often employed to increase the accuracy of measurement. Electrometric Method This method is used to monitor continuously the degree of acidity of alkalinity of process liquids. Basically, an electrode immersed in a liquid acquires an electrical potential. The value of the potential varies with the pH value of the liquid. It is impracticable to measure a single potential, therefore a reference electrode is employed in conjunction with the measuring electrode. The potential of a reference

electrode is a known constant value and will not change with solution composition. For pH measurement therefore, the e.m.f. (mV) generated by the measuring electrode is compared with that of the reference electrode and is converted to a pH scale.

pH is the measurement of the acidity of alkalinity of an aqueous solution and is the Hydrogen-ion concentration. The pH scale is a logarithmic scale and ranges from 0 to 14. Neutral solutions, and pure (distilled) water, which have neither acidic or alkaline content, have pH value of 7, acids have values below 7 and alkalis have values above 7.

81

The pH scale shown above was proposed by Sorensen in 1909. It uses the Hydrogen ion concentration as a measure of the degree of acidity of alkalinity of a solution.

pH = -log10 [H+] = log10 __1__ [H+]

A pH of 14 indicates a strong alkali whilst a pH of 0 indicates a strong acid. Ionisation

The process of ionisation, is the dissociation of a compound into charged particles called ions, and it occurs in aqueous solutions. Provided the solutions are not too concentrated, any strong acid when dissolved in water will produce positively charged Hydrogen ions and negatively charged Acid radical ions.

HA = [H+] + [A-]

where HA is the acid, [H+] is the Hydrogen ions, and [A-] the Acid radical ions. As all strong acids are fully ionised in aqueous solutions, a measure of the Hydrogen ion concentration will give strength of the acid solution. Some examples of strong acids are:

Hydrochloric acid HCl

Sulphuric acid H2SO4

Nitric acid HNO3

In a weak acid, such as Acetic, the number of dissociated molecules is small. Water in this context is also weak and ionises as follows:

H²

O = [H+] + [OH-]

where [H+] are positive Hydrogen ions and [OH-] are negative hydroxyl ions. It has been established that, for most practical purposes, the product of the Hydrogen ion and hydroxyl ion concentration at a temperature of 25ºC is given by the expression:

[H+] x [OH-] = 1x10-14

In pure water the number of Hydrogen ions and hydroxyl ions is the same, the concentration of each is equal to 1x10-7. The product of the Hydrogen ion and hydroxyl ion concentration remains constant even when other compounds are dissolved in the water. If a strong acid is added, and the Hydrogen ion concentration is increased to 1x10-4, the hydroxyl ion concentration will decrease to 1x10-10. Potassium Hydroxide: KOH Sodium Hydroxide: NaOH

82

Buffer solutions A buffer solution is defined as a solution that resists changes in pH as a result of either dilution or small additions of acids or alkalis (bases). The most effective buffer solutions contain large and approximately equal concentrations of a conjugate acid-base pair. Buffer solutions of certain pH ranges are used for calibrating pH measuring equipment. Weak acids commonly used in buffer solutions are phosphoric, boric, acetic and citric. These are partially neutralised with an alkali or a salt of the acid. It is important to achieve the correct proportions of salt and acid in order to obtain the correct pH. Standard buffer solutions have good characteristics and for pH 4, 7 and 9.2 are available commercially, as pre-weighted tables, sachets of powder or in solution form. Those unobtainable commercially are simple to prepare.

7.3 The Calomel reference electrode Calomel electrodes are the most widely used reference electrodes. The electrode consists of the inner electrode assembly which is placed within a guard tube as shown below.

The inner electrode assembly consists of a Platinum wire dipped into mercury (Hg), Mercurus Chloride (Hg2Cl2) Potassium Chloride paste (KCl). The inner electrode assembly is placed into a guard tube filled with saturated Potassium Chloride solution. At the bottom of the tube Potassium Chloride crystals may be seen. They are maintaining a saturated solution. Electrical connection between the inner electrode assembly and the liquid being tested is made via the potassium chloride solution. This made possible by the two porous plugs. The potential of this electrode is governed by the activity of the chloride ion.

Platinum wire

Screened lead

Liquid level

Potassium Chloride crystals

Potassium Chloride –Calomel paste

Porous

plugs

Guard tube

filled with

KCl solution

Mercury

83

A variety of Calomel electrodes are commercially available. The saturated Calomel electrode uses a saturated solution of KCl but is suffers from potential variations due to temperature changes. For accurate work 0.1M or 1M KCl Calomel electrodes should be used. Calomel electrodes using NaCl solutions are also available.

7.4 The Glass measuring electrode The glass measuring electrode consists of:

(i) a thin [H+] ion responsive glass membrane sealed to a stem of high resistance non-responsive glass, and

(ii) an internal reference electrode with a constant internal Hydrogen ion concentration.

The internal reference electrode may be either Ag/AgCl in HCl or Hg/Hg²Cl² in HCl solution. The solution of HCl is of known pH value. The entire cell requires an external reference electrode for operation.

84

When the electrode is placed in a solution, a potential difference is established across the membrane. The value of this potential difference depends upon the pH of the solution and also on the temperature of the membrane. Thus, for accurate measurements some form of temperature compensation is required. The glass measuring electrode does not suffer from the influences of oxidising and reducing agents. One disadvantage is that this type of cell has a very high impedance (1-500MΩ) and necessitates the use of a voltmeter with a very high input impedance or an electrometer. Glass electrodes must be calibrated fairly often with buffers of pH value within one unit of the pH to be measured. These electrodes cannot be used in highly acidic fluoride solutions. Also suffer from errors in highly acidic or highly alkali solutions due to the interference of anions and cations.

7.5 Electrical circuits for use with Glass electrodes For measurement of pH the e.m.f. (mV) generated by the glass electrode compared with that of the reference electrode has to be converted to a pH scale, that is, one showing an increase of one unit for a decrease in e.m.f. of approximately 60mV. The pH scale requires the use of two controls:

(i) the calibration control and (ii) the slope control.

The calibration control relates a measured e.m.f. to a fixed point on the pH scale. The slope control varies the number of mV to one pH unit. The slope control may be used as a temperature compensation control. A typical pH measuring system (glass electrode and reference electrode immersed in a solution) may have a high resistance as mentioned earlier. To obtain an accurate measurement of the e.m.f. developed at the measuring electrode, the electrical measuring circuit must have a high input impedance and the insulation resistance of the electrical leads from the electrodes to the measuring circuit must be extremely high (1x105 MΩ). The usual method of measurement is to convert the developed e.m.f. into a proportional current by means of a suitable amplifying system. With modern integrated circuit techniques it is possible to obtain an amplifier with a very high input impedance and very high gain, so that little or no current is drawn from the electrodes. Such a system is employed in the pH-to-current converter shown below.

85

It employs zener diode stabilised power supplies and feedback networks designed to give a high gain and a very high input impedance. The principle of another system which achieves a similar result is shown below.

It uses a matched pair of FET’s housed in a single can. Here the e.m.f. produced by the measuring electrode is fed to the gate of one of the pair. The output of the operational amplifier will be controlled by the difference in the potentials applied to the gates of the FET’s. The current flowing through the local and remote indicators

will be a measure of the change of potential of the measuring electrode.

86

If the e.m.f. given by the glass electrode is plotted against pH for various temperatures, it will be seen that there is a particular value of pH at which the e.m.f. is independent of temperature. This point is known as iso-potential point. The pH meter

A few commercial pH meters have variable iso-potential control so that they can be used with several different combinations of electrodes. It is more generally the case that pH meters have fixed iso-potential control settings and can only be used with certain combinations of measuring and reference electrodes. It is strongly recommended that, with fixed iso-potential control settings, both the glass and reference electrodes be obtained from the manufacturer of the pH meter. Temperature compensation circuits generally work only on the pH and direct activity ranges of pH meter and not on the mV range. Modern pH meters with analogue displays are scaled 0 to 14 pH units with the smallest division on the scale equivalent to 0.1 unit. The mV scale is generally 0-1400mV. Digital outputs are also available with the most sensitive ones reading to 0.001 pH units or 0.1mV. Instruments incorporating microprocessors are also available. These can calculate the concentration of substances from pH measurements and give readout in concentration units.

7.6 Alkaline and Asymmetrical errors

(i) Precision and accuracy

Measurements reproducible to 0.05pH units are possible in well buffered solutions in the pH range 3 to 10. In poorly buffered solutions reproducibility may be no better

than ±0.1pH unit and accuracy may be lost by the absorption of CO² or by the

presence of suspensions and gels.

(ii) Sodium Ion error Glass electrodes for pH measurement, although selective for Hydrogen ions, also respond to other ions such as Sodium especially at pH values greater than 11. This causes the pH value to be underestimated. To overcome this problem the electrode can either be standardised in an alkaline buffer solution containing the appropriate salt in a suitable concentration or use special lithium glass electrodes developed for use in solutions of high alkalinity. Electrode manufacturers produce tables of the errors involved with glass electrodes at various temperatures to allow corrections to be made.

(iii) Temperature errors The calibration slope and standard potential of ion-selective electrodes (including glass pH electrodes) are affected by temperature. If measurements are made at a temperature different from that at which the electrode was calibrated there will be an error. This will be small if the meter has an iso-potential settling. For the most accurate work the sample and buffer solutions should be at the same temperature, even if iso-potential correction is possible.

87

7.7 Industrial pH systems with Glass electrodes

Two types of electrode systems are in common use: the continuous-flow type of assembly and the immersion or dip-type of assembly.

(i) Continuous-Flow type of assembly The physical form of the assembly may vary a little from one manufacturer to another. Figure below shows a typical assembly designed with reliability and easy maintenance in mind.

It is constructed in rigid PVC throughout, it operates at pressure up to 2bar and temperatures up to 60ºC. For higher temperatures and pressures the assembly may be made from EN58J stainless steel. The standard measuring electrode is usually accommodated in toughened glass. A reservoir for potassium chloride (or other electrolyte) forms a permanent part of the electrode holder. A replaceable reference electrode fits into the top of the reservoir. A micro ceramic plug at the lower end of the reservoir ensures slow electrolyte leakage (up to six months continuous operation without attention is usually obtained).

88

The close grouping of electrodes makes possible a small flow cell, and hence a fast pH response at low flow rates. The flow through the cell creates some degree of turbulence and thus minimises electrode coating and sedimentation.

(ii) Immersion type assembly This assembly is similar to the flow-type except that the flow cell is replaced by a protecting guard which protects the electrode but allows a free flow of solution to the electrodes. Immersion depths up to 3m are available. Electrode assemblies should be designed so that the electrodes can be kept wet when not in use. When the electrode assembly is removed from the process, it can be immersed in a bucket filled with process liquid, water or buffer solution. The design of the electrode assembly is often modified to suit the use. When the assembly is immersed in a tank, care must be taken in the sitting to ensure the instrument is measuring the properties of a representative sample. The main cause to trouble in electrode assemblies is the fouling of the electrodes. In order to reduce this, two form of self cleaning are available and the choice of method is dependent on the application. Where the main cause of trouble is deposits on the glass electrode and mechanical cleaning is required, this may be achieved by a cleaning attachment incorporated with the electrode assembly. Alternatively an ultrasonic generator operating at 25kHz can be fitted to the electrode assembly, this greatly increasing the periods between necessary electrode cleaning.

7.8 Redox (Eh) potential and its measurement Oxidation and reduction, which involves a transfer of electrons, occur in some aqueous solutions. An electrode placed in one of these solutions will acquire a potential and its value will depend upon the degree of oxidation and reduction which has taken place. This potential generally referred to as Redox potential is given the symbol Eh. Oxidation

Originally the word oxidation was used to describe the process in which a substance combines with Oxygen, for example some materials were said to be oxidised when they were burned in the presence of oxygen. Later it was observed that materials such as magnesium appeared to burn in an atmosphere consisting of a gas other than oxygen. The change in the atomic structure of the magnesium was found to be the same whether the burning was in Oxygen or in another gas, this change being a loss of electrons.

The word oxidation is now used in a much wider sense. It is applied to chemical reactions in which the atoms of a material lose electrons. For example, copper is said to be oxidised when it forms copper sulphate.

89

Reduction

The process of reduction is the reverse of oxidation and the two occur simultaneously. Thus, as one substance is being oxidised, another is being reduced and the atoms of the substance undergoing reduction will gain electrons. As an example, during the production of magnesium chloride (MgCl2) the magnesium is oxidised and the chloride is reduced. What should be taken into account when measuring Redox potentials

Some chemical reactions involve the combination of two substances. The resulting product may contain either a deficiency or an excess of one of these in varying quantities. For product uniformity, the reaction must be maintained at the correct rate and measuring the Redox potential is one method of monitoring this. A fairly comprehensive knowledge of the process liquids is required prior to the measurement of a redox potential. It is necessary to know for example:

(i) The nature and behaviour of the chemical solution involved e.g. acids, bases, and salts.

(ii) That there are two substances in the solution, one of which is oxidised and the other reduced during normal plant operation.

(iii) Whether the speed of reaction is fast enough to permit satisfactory measurement and control.

(iv) It there are any substances present in the solution which may cause secondary reaction.

(v) The pH value of the solution and whether it is intended to control this. (vi) The variations in temperature of the solution.

Equipment required to measure Redox potential consists of:

(i) A measuring electrode which will acquire a potential having a value dependent of the state of the oxidation/reduction reaction in the solution.

(ii) A reference electrode in order that a potential difference may be measured. The potential acquired by this electrode must be independent of the redox reaction and the calomel electrode is normally employed.

(iii) An instrument for measuring the potential difference. This instrument must have a high input impedance as described previously.

90

Chapter 8: Chromatography

Chromatography is a general name given to the methods by which two or more components in a sample physically separate themselves by distributing themselves between two phases:

(i) a stationary phase which can be a solid or a liquid supported on a solid, and (ii) a mobile phase which can be either a gas or a liquid which flows

continuously around the stationary phase. The mobile phase flows continuously over the stationary phase and separation of individual components occurs by their different affinities for the stationary phase.

In liquid chromatography (LC) the mobile phase is a liquid. In gas chromatography (GC) the mobile phase is a gas. Chromatography can be considered as a chemical extraction process. Chemical extraction is a means of separating the components of a sample.

8.1 Classification of Chromatographic techniques Chromatographic techniques can be classified according to the type of stationary phase used i.e. solid or liquid. If a solid stationary phase is used the separation is achieved by adsorption. Separation by adsorption is caused by differences in the adsorption forces between the various components of the sample. Each component will be adsorbed at different rates and therefore adsorption occurs at different levels in the column. The adsorbent is generally an active porous solid with a large surface area such as silica gel, alumina or charcoal. If a liquid stationary phase is used then separation is achieved by partition. Separation by partition is similar to adsorption but depends upon the solubility of components of the sample in the liquid stationary phase. The stationary phase may be a liquid layer on a solid support i.e. water in a porous solid. The least soluble components travel more rapidly and therefore further down the column to emerge before the more soluble components.

91

8.2 Basic principles of Chromatography To illustrate the basic principles of chromatography we shall consider a hypothetical separation of a three component sample in a closed column. The stationary phase (packing) consists of solid porous particles contained inside a long narrow tube called the column. Figure below demonstrates the chromatographic process.

A small volume of sample solution is injected at the column inlet (see A). The mobile solvent phase (eluting agent) moves the sample through the column packing (see B). As the sample passes through the column it separates either by adsorption or partition as explained earlier. Each component X is distributed between the stationary phase (s) and the mobile phase (m) as it passes down the column. The corresponding distribution coefficient for component X is given by: Kx = (X)s (X)m A large value of Kx indicates that the component favours the stationary phase and moves slowly through the column, whilst a small value of Kx indicates that the component favours the mobile phase and moves quickly down the column. The different speeds of the components separate them along the column (see C). The eluting agent eventually carries the components out of the column. Different components will emerge at different times (see D). If one measures the concentration of each component as it exits from the column and plots it as a

92

function of the volume of mobile phase passed through the column or time, a chromatogram results.

Theory related to practice

The object of chromatography is separation in a reasonable time. Retention In order to achieve separation, one must have retention. The distribution co-efficient Kx is defined as a measure of the degree of retention for component X. The capacity factor K’x is a more practical quantity that can be determined directly from the chromatogram, and it is given by:

K’x = total moles of X in stationary phase = (

) Kx = tr - to

total moles of X in mobile phase to Vs is the volume of the stationary phase within the column, Vm is the volume of the mobile phase within the column, tr is the peak retention time, and to is the time required for solvent molecules to traverse the column. The fundamental equation for any chromatographic process relating to retention volume Vr is given by: Vr = Vm (1 + K’x ) = Vm + Vs Kx

93

Example Using the hypothetical chromatogram given previously calculate the capacity factors of peaks B and C. Assuming that 60% of the volume of a column 25cm in length by 0.40cm in internal diameter is occupied by solid packing particles, calculate the expected retention volume for peak C. (i) K’B = tr - to = 13-2 = 5.5 to 2 K’C = 21.5 - 2 = 19.5 = 9.8 2 2 (ii) Vm = 0.40 πr²L (this is the fraction column volume occupied by mobile phase, treat the column as a cylinder) Vm = 0.40 x 3.14 x (0.2)² x 25 = 1.256 cm³ Vr = Vm (1 + K’C) = 1.256 (1+9.8) = 13.565 cm³ Column efficiency

It describes the rate of band broadening as the solute travel through the column or across the plate or paper. The quantitative measure of efficiency is the number of theoretical plates N calculated from the chromatogram by using the following equation.

(

)

tw is the peak width measured in the same units as the retention time. It is obtained from the intersection of the base line with the tangents drawn on the sides of each peak. The theoretical-plate model, assumes a column to be made up of a series of plates. The higher the value of N, the more chance there is for separation to occur. Another useful parameter for column efficiency is the Height Equivalent to a Theoretical Plate (HETP). This is the H value and has units of length. H = L N L is the column length. A column of plate with a low H value is better than one with a high value. Values of H less than 1-3mm are commonplace in gas chromatography.

94

Example Using the hypothetical chromatogram given previously determine

(i) the number of theoretical plates for peak C (ii) assuming the column length is 25cm calculate the value of H in mm for

peak C.

(i) (

)

(

)

(ii) H = L = 250mm = 0.57mm N 440 Resolution The degree of separation is referred to as resolution R.

(

)

The hypothetical chromatogram given previously and the equation given above enable the examination of chromatographic peaks and the calculation of R. If two components A and B of the chromatogram given previously are examined, resolution is determined by the distance between the peak maxima and peak width. The separation between peaks is related to the selectivity factor , sometimes called relative retention.

= k’B

k’A Selectivity refers to the capability of a chromatographic system to distinguish between two components. As the selectivity factor approaches to unity separation becomes very difficult. To modify selectivity one must change the stationary phase, the mobile phase of both.

95

Figure below shows the influence of selectivity and efficiency on chromatographic resolution.

Figure A shows a two component separation displaying poor resolution. Figure B shows by decreasing peak width (i.e. more theoretical plates), resolution can be increased without affecting selectivity. Figure C indicates that better resolution results from increasing the distance between peal maxima (improving selectivity), even without increasing efficiency. Figure D shows poor resolution caused by low capacity factor K’x. Better resolution can be obtained by increasing Vs (using longer column or an absorbent with a higher surface area). In chromatography a value of R=1 is considered the minimum value for quantitative separation. Example

For the hypothetical chromatogram shown previously determine the resolution between peaks B and C and the value of the selectivity factor for the same peaks.

a) (

)

(

) = 2.7

b)

= 1.8

96

8.3 Gas Chromatography

On-line or process gas chromatographs are instruments which incorporate facilities to carry out the analytical procedure automatically. Samples are taken from process streams and presented in a controlled manner and under known conditions to the gas chromatograph. Successive analyses may be made on a regular timed basis. The main components of a typical process chromatograph system are shown below.

These components are:

(i) A supply of carrier gas to transport the sample through the column and detector,

(ii) A valve for introduction of known quantities of sample, (iii) A chromatographic column to separate the sample into its components, (iv) A detector and associated amplifier to sense and measure the components of

the sample in the carrier gas stream, (v) A programmer to actuate the operations required during the analytical

sequence and to control the apparatus, and (vi) A display of data processing device to record the result of the analysis.

Sampling system

It must present a representative sample of the gas or liquid to be analysed. The sample is treated (e.g. dried, filtered) and temperature and pressure adjusted. Discrete volumes of the treated sample are injected into the carrier gas stream by means of a gas sampling valve. It is important that the sample size is constant for each analysis and that it is introduced into the carrier gas stream rapidly. The sample should be allowed to flow continuously through the sampling system to minimise transportation lags. Chromatographic sampling or injection valves are specially

97

designed change-over valves which enable a fixed volume, defined by a length of tubing (the sample loop), to be introduced in either one or two gas streams with the minimum interruption of either stream. The design and operation of a typical sampling valve is shown below.

The inlet and outlet tubes terminate in metal (stainless steel) blocks with accurately machined and polished flat faces. A slider or soft plastic material, is held against the polished faces and moved between definite positions to fill the loop or inject the sample. The main difference between “gas” and “liquid” sampling valves is in the size of sample loop. In the “gas” sampling valve the loop is formed externally and typically has a volume in the range 0.1-10ml. For liquid sampling the volumes required are smaller and the loop is formed in the internal channels of the valve and may have a volume as small as 1µl.

98

Chromatographic columns The separating columns used in process chromatographs are typically 1-2m lengths of stainless steel tubing, 3-6mm outer diameter, wrapped as a helix for convenient housing, and packed with a solid absorbent. Separation of gases is normally carried out on columns packed with molecular sieve. The length of the column will influence the separation achieved. Too short column will give incomplete separation whilst too long column will mean that the retention time will be too long and the amplitude of the chromatographic peaks will be reduced. As column length increases selectivity increases but efficiency decreases. Carrier gas The carrier gas transports the components of the sample over the stationary phase in the chromatographic column. The carrier gas must not react with the sample and it is advantageous to use a gas of low viscosity particularly when using long columns. The primary factors determining the choice of carrier gas are its effect on component resolution and detector sensitivity. The carrier gas and type of detector are chosen so that the eluted components generate large signals. Helium is normally used (as carrier gas) with thermal conductivity detectors because of its high thermal conductivity. However, when measuring Hydrogen in trace quantities using a thermal conductivity detector, Helium should not be used as a carrier gas because Hydrogen and Helium have high and similar thermal conductivities. The flow rate of the carrier gas must be adjustable since it affects the retention time of a compound in the column and the shape of the chromatographic peaks. Mechanical or electronic flow controllers are employed to control the carrier gas flow rates. Controlled temperature enclosure

Temperature will influence the mass flow rate of the sample (at higher temperatures there will be fewer molecules in a given volume of gas). It is important that the temperature is kept constant. Certain components of the chromatograph, such as the sample injector, columns and detectors are usually enclosed in temperature controlled zones within the instrument (to control to ±0.1K or better). Two general methods are used to distribute heat in a temperature controlled enclosure. (i) an air bath and (ii) metal to metal contact The first is advantageous because of fast warming time and accessibility to all components in the enclosure. The second has a slower warming time and is more hazardous.

99

8.4 Component detection Thermal conductivity detector One of the most commonly used devices is the thermal conductivity detector which measures the change in thermal conductivity of a gas mixture. The thermal conductivity detector produces a large signal requiring no amplification. The detector cell has either two or four filaments arranged in a Wheatstone bridge circuit shown below.

In the four-filament model, two filaments in opposite arms of the bridge are surrounded by carrier gas flowing in a reference stream, the other pair by carrier gas flowing out of the column. When the bridge is balanced, no signal appears across points 1 and 2. Since the temperature of the filaments is proportional to the rate at which heat is transported to the cell body by the gas, and since resistance is proportional to temperature, a change in the heat conductivity of the gas will produce an output signal at points 1 and 2. Most organic vapours have low thermal conductivities compared to Hydrogen or Helium. For this reason, Helium is widely used as a carrier gas. If interested in analysing the noble gases (He, Ne, Ar, Kr), Nitrogen might be the choice. The TCD is reliable, simple, non-destructive and moderately sensitive. It responds to essentially all compounds.

Relative Thermal Conductivities of some gases

Air 1.00 Carbon monoxide 0.96

Oxygen 1.01 Carbon dioxide 0.59

Nitrogen 1.00 Sulphur dioxide 0.32

Hydrogen 4.66 Water vapour 1.30

Chlorine 0.32 Helium 4.34

100

Note: Thermal conductivity depends upon molecular size, mass and temperature. It is independent of pressure over a wide range. At very low pressure the thermal conductivity of a gas is proportional to its pressure. Flame ionisation detector (FID)

It has a wide linear range and high sensitivity, and is quite reliable. It consists of a Hydrogen-air flame polarised in an electrostatic field. The flame ignites and ionises the combustible sample components as the carrier gas passes into it. The ions are collected at the electrodes producing a current.

The FID does not respond fully to Oxygenated carbons. Also, it does not respond at all to water or to the permanent gases (N2, O2, CO2, etc.) making it ideally suited for trace analysis in aqueous solutions and atmospheric samples. Electron capture detector (ECD)

The electron capture detector takes advantage of the affinity of certain functional groups for free electrons. The sample and the carrier gas pass through a cell containing a beta source, which ionises them both. The source can be Pt foil saturated with H2. A Ni foil is used more frequently because of its higher temperature stability. The beta particles ionise the carrier gas molecules and produce electrons (-ve charges) which migrate to the anode (Collector electrode) under an applied potential of 1-100V. The electrons produced by the ionised carrier gas are attracted to the Collector electrode (anode +ve) and act as a load. This load is needed to produce a standing current.

101

Positive charges (electron capturing species) from the ionised sample entering the column will react with the electrons to form an ion or neutral molecule. The net result is a reduction in the number of electrons i.e. a drop in the standing current. A peak in ECD detection is therefore actually a detector current drop since the maximum current is found in the absence of capturing species. Response is non-linear, but a linear range of 0.5 – 1x10³ can be achieved by pulsing the polarising voltage. The pulse duration is long enough for electron collection, but not for ion collection. The major advantage of the ECD is its selectivity. It is not common for FID and ECD to be combined, displaying the response of both detectors on the same chart using a two-pen recorder. The many applications of ECD include analysing pesticides and organometallics (e.g. lead) and tracing SF6 in flue and stack gases.

Sample + Carrier gas

entering the column

Potential 1-100V is applied between the

Collector electrode and the ECD casing

Standing current I

102

8.5 Operation and Data handling of a Chromatograph Programmers

The operation of gas chromatograph involves the execution of a series of operations. Process gas chromatographs incorporate programmers which enable the analytical sequence to be carried out automatically. The most basic type of programmer will consist of mechanical or electromechanical timers which operate relays or switches at the appropriate time to control the sequence of analysis. Some chromatographs have built in microprocessors capable of controlling and monitoring many more parameters as well as data logging. When the process chromatograph is operated in automatic mode all timed sequences are under programmer control (sample stream selection, sample injection, column or detector switching, automatic zero and attenuation adjustment, and back flushing). Other functions which the programmer may carry out are: fault detection and identification, alarm generation and automatic shutdown when a fault is detected. Data Processing The output from a gas chromatograph will normally be an electrical signal. The simplest method of presenting the data is as a chromatogram of the sample which is obtained by recording the detector output as a function of time. This is not satisfactory, however, since further data processing is necessary for interpretation of the results. The processing system must therefore be able to identify peaks in the chromatogram corresponding to the components of interest and then measure a suitable parameter which relates to the concentration of the component. Identification of the peaks in the chromatogram is made on the basis of retention time. If all other parameters remain constant (e.g. temperature, carrier gas flow rate) the retention time is characteristic of a given compound on a particular column. Small changes in operating conditions will change the retention time, therefore data processing systems must be capable of allowing for these changes. The concentration of a component can be determined from the output signal by measuring the height of the peaks or the area under it. In either case a calibration curve must be produced before analysis. These are produced from the analysis of standard mixtures. The calibration data can be stored in the system and the necessary calculations carried out after each analysis and the results of the analysis are simply printed out. Automatic updating of the calibration is also possible. Simple data processing systems relate peak height to concentration. It is usually better to use peak areas which compensates for changes in peak shape. To measure peak area the system must include an integrator. Gas Chromatograph Integrator There are a number of different types of integrator in use to measure the areas under the peaks of a chromatograph. The area is obtained by adding a number of individual measurements of the detector output during the peak. The peak area is expressed in mV/sec. The differences between the different types of integrator are the method employed to process the detector output signal and the facilities available for subsequent processing of the peak area values. All instruments convert the analogue detector output signal to digital form.

103

REFERENCES Existing process sampling methods & new sampling devices for the process

industry by Kevin Cook www.tycovalves.com

DOPAK sampling systems www.dopak.com

Instrumentation Reference Book by Walt Boyes, Fourth Edition, Butterworth-

Heinemann publications

Instrumental Analysis, Henry Bauer, Gary Christian, James O’Reilly, International

Student edition, Allyn and Bacon, Inc