ICP Operations Guide

8/11/2019 ICP Operations Guide

1/44

ICP Operations Guide

A Guide for using ICP-OES and ICP-MS

by Paul R. Gaines, PhD


2/44

Inorganic Ventures has over twenty-five years experience specializing in themanufacturing of inorganic certied reference materials (CRMs) and nearly a decadeaccredited to ISO 17025 & ISO Guide 34 by A2LA . Tis singular focus has enhancedthe quality of our manufacturing, the depth of our technical support and the caliber ofour customer service.

Te pursuit of excellence in these areas has lead to the creation of the ICP OperationsGuide. Te purpose of this guide is to assist ICP / ICP-MS operators with thenumerous tasks they encounter on a daily basis. Te topics are fundamental in natureand are intended as an aid for the analyst who is completely new or somewhat new to thetechnique of ICP.

copyright2011 by Inorganic Ventures, Inc.


3/44

ICP Operations GuideA Guide for Using ICP-OES and ICP-MS

Tis guide is intended or anyone operating and preparing samples and standards or measurementusing ICP (ICP hereafer reers to either ICP-MS or ICP-OES). Our last guide, race Analysis:A Guide or Attaining Reliable Measurements, ocused on the task o achieving reliable tracemeasurements by ICP. Tis series will not ocus on any single topic, but rather upon a multitudeo day-to-day tasks required by all ICP operators. Te topics will be undamental in nature and are

intended as an aid or the analyst who is completely new or somewhat new to the technique o ICP.

able of contentsT

Multi-Element Standard Blends...................................4

1. Elemental and Matrix Compatibility

2. Quality Issues

3. Handling, Preparation and Storage o Standards

Sample Introduction....................................................11

4. Sample Introduction Systems

5. Nebulizers, Spray Chambers and orches

6. Compatibility and Precision Issues

Performance Characteristics.......................................16

7. Linearity and Detection Limits

8. Spectral Intererence: ypes, Avoidance and Correction

9. Key Instrument Parameters

Calibration Techniques...............................................24

10. Calibration Curve and Standard Additions echniques

11. Internal Standardization and Isotope Dilution

Problem Elements.......................................................28

12. Common Problems with Hg, Au, Si, Os and Na

13. Common Problems with Ag, As, S, Ba, Pb and Cr

Basic Calculations.......................................................3214. Accuracy, Precision, Mean and Standard Deviation

15. Signicant Figures and Uncertainty

16. raceability

by Paul R. Gaines, PhD

1


4/44

* Visitinorganicventures.com/tech/icp-operations/ foradditionalinformationfrom thislink

ulti-Element Standard BlendsElemental and Matrix CompatibilityM1

Nitric Acid Matrices

Most analysts preer nitric acid (HNO3) matrices due to the solubility o the nitrates as well as its oxidizing ability and

the relative reedom rom chemical and spectral intererences as compared to acids containing Cl, S, F, or P. In addition,HNO

3is very popular in acid digestion sample preparations.

Te elements that are stable/soluble and commonly diluted in aqueous/HNO3are shaded in red below:

1. Os should never be mixed with HNO3due to

the ormation o the very volatile OsO4.

2. Cl is oxidized to molecular Cl2which is volatile

and adsorbs on plastic.3. Br and I are oxidized to molecular Br

2and I

2

which adsorb onto plastic.4. Dilutions o Hg and Au in HNO

3below 100

ppm should be stored in borosilicate glass due toHg+2adsorption on plastic.5. Not soluble above concentrations o 1000 gmL.

6. race levels o HCl or Cl- will orm AgCl,which will photoreduce to Ag0.

Fdenotes that the element can be diluted in HNO3i complexed with F-.

Cldenotes that the element can be diluted in HNO3i complexed with Cl-.

HFdenotes that the element should have excess HF present when diluted with HNO3.

Tdenotes that the tartaric acid complex can be diluted in HNO3.

Hydrochloric Acid Matrices

Te use o hydrochloric acid (HCl) is the next most popular acid matrix. HCl is volatile and it is corrosive to theinstrument and it's electronics thereore, exposure should be kept to a minimum.

Te elements that can be diluted in HCl are shaded in blue below:

1. Concentrated (35%) HCl will keep up to 100g/mL o Ag+in solution as the Ag(Cl)X-(X-1)complex. For more dilute solutions, the HCl canbe lowered such that 10% HCl will keep up to 10g/mL Ag in solution.NOE: Te Ag(Cl)X-(X-1)complex is photosensitiveand will reduce to Ag0when exposed to light.HNO

3solutions of Ag+are not photosensitive.

2. Parts-per-billion (ppb) dilutions o Hg+2in HClare more stable to adsorption on the containerwalls than are dilutions in HNO

3.

Fdenotes that the element is more stable tohydrolysis i complexed with F-. In the case o Si and Ge the uoride complex is generally considered a necessity.


5/44


Water at pH of 7

Dilutions in water at pH 7 are not as common or

most elements but may be required to preventchemical reactions o some o the compoundscontaining the element. Please note that solutionsat pH 7 may support biological growth andthereore the long-term stability should bequestioned.

Tose elements that may have an advantage tobeing diluted in water at pH 7 are shaded inyellow to the right:

Hydrouoric Acid Matrices

Hydrouoric acid (HF) requires the use o HF-resistant introduction systems. Tese systems aremore expensive than glass, have longer washouttimes, and give a larger measurement precision.However, there are times when the use o HFoffers a major advantage over other reagents.

Tose elements where an HF matrix may beoptimal are shaded in green below:

1. HF is used or Si3N

4preparations and other

nitrides.

Sulfuric Acid Matrices

Suluric acid (H2SO

4) is commonly used in preparations and thereore added to standards in combination with other acids.

Elements that either benet or comortably tolerate the presence o H2SO

4are shaded in orange below:

1. Dilutions o Hg and Au in H2SO

4below 100

ppm should be stored in borosilicate glass due toadsorption on plastic.2. race levels o HCl or Cl-will orm AgCl,which will photoreduce to Ag0.

Fdenotes that the element can be diluted inH

2SO

4i complexed with F-.

Cldenotes that the element can be diluted inH

2SO

4i complexed with Cl-.

HFdenotes that the element should have excessHF present when diluted with H

2SO

4.

Tdenotes that the tartaric acid complex can bediluted in H

2SO

4.

Phosphoric Acid Matrices

Phosphoric acid (H3

PO4

) is not commonly used in preparations since it attacks glass, quartz, porcelain, and Pt containers atelevated temperatures (greater than 100 C). However, the presence o

3PO

4will not adversely effect any o the elements at low

g/mL levels and below.


6/44


Quality Issues

Tere are several quality issues that are important with respect to multi-element chemical standards:

Many o the topics above have been discussed in other publications on our site. Please use the links provided throughout this

article to gain a better understanding o the issues discussed below.

AccuracyTe accuracy o a certied reerence material (CRM) standard is dependent upon:

*

or multi-element blend).

* (Note that

uncertainty calculations will be discussed in part 3 of this series).

*

Elements at ppb Concentration Levels*or detailed inormation on physical stability.

*

in place helps to prevent laboratory blunders. See ISO Guide 34, 17025, and 9001 Explained*to learn more about out

which International Organization o Standardization (ISO) standards are most important or trace analyses.

PurityPurity becomes an issue when using starting materials o single element blends to prepare multi-element blends. Te degree o

importance increases as the relative order o magnitude o the components increases. Known purity and hopeully very clean

materials are critical in the execution o ICP-OES spectral intererence studies. Tese studies typically involve the aspiration o

a 1000 g/mL solution o a single element while collecting the spectral regions o analytes that may be interered with.

Inorganic Ventures laboratory has purchased many materials claiming a purity o 5 to 6-9s. However, its never a bad idea to

conrm a manuacturers claims. For more inormation regarding purity considerations, please consult the ollowing online

articles:

*

*

*

Chemical CompatibilityIts important or the multi-element blends to be compatible with the containers in which they are prepared and stored. Its

equally important that they are compatible with the introduction system o the instrument(s) used to analyze the blend and

with the other analytes within the blend. Some points to consider:

or chromate, ppt o the alkaline and rare earths with F- in HF matrices, ppt o uorinated elements like Sn(F)x-y in the

presence o elements that would complex with the uoride and thereore pull it away rom the metal stabilized as the uoride

complex, etc.

2


7/44


chloride to the very volatile and toxic OsO4when nitric acid is added. Volatile compounds may not be lost rom the standard

solution but will give alse high readings due to a disproportionate amount o the element making it to the plasma where the

nebulization efficiency is greater due to the added mode o transport to the plasma as the vapor state.

Stability

the two should be made to conrm stability. I there are chemical concerns rom the beginning then a resh blend should

be prepared the next analytical day or comparison. Reer to Stability o Elements at ppb Concentration Levels*or more

inormation.

AvailabilityConsider the ollowing:

Some o these questions may appear as i they belong in other sections but they all impact the availability o the standard in

important ways. For example, blends that must be kept rerigerated or rozen cannot be used until allowed to come to room

temperature. Tis is ofen the case with blends manuactured within the biological pH range o 4-10.

DocumentationAlthough documentation may seem less important than the above topics, it is paramount or less obvious reasons. Tink

about the ollowing questions:

ISO has issued a document reerred to as ISO Guide 31. Tis document details what the international scientic community

considers to be critical to the analyst when using chemical standard solutions or CRMs. Our guide to Certicate o Analysis

Components*offers explanations o each section o an ISO Guide 31-compliant Certicate o Analysis.

Traceability be more critical than you realize. raceability has been dened as the property o the result o a measurement or the value o

a standard whereby it can be related to stated reerences, usually national or international standards, through an unbroken

chain o comparisons all having stated uncertainties. Tis denition has achieved global acceptance in the metrology

community. Reer to our article NIS raceability*or additional inormation.

Calculations,Handling, Preparation and Storage of Standards3Handling

Observing the ollowing recommendations will save considerable time, money, and rustration:

1. Never put solution transer devices into the standard solution. Tis precaution avoids possible contamination rom thepipette or transer device.

2. Always pour an aliquot rom the standard solution to a suitable container or the purpose o volumetric pipette solutiontranser and do not add the aliquot removed back to the original standard solution container. Tis precaution is intended to


8/44


avoid contamination o the stock standard solution.

3. Perorm volumetric pipette solution transer at room temperature. Aqueous standard solutions stored at lower temperature or the concentrations units are in wt./wt. rather than wt./volume.

4. Never use glass pipettes or transer devices with standard solutions containing HF. Free HF attacks glass but it is sometimesconsidered sae to use glass when the HF is listed as trace and /or as a complex. However, many uorinated compounds willattack glass just as readily as ree HF.

calculated provided the density o the standard solution is known. Tere are too many possible pipetting errors to risk a

volumetric transer without checking the accuracy by weighing the aliquot.

6. Uncap your stock standard solutions or the minimum time possible. Tis is to avoid transpiration concentration o theanalytes as well as possible environmental contamination.

7. Replace your stock standard solutions on a regular basis. Regulatory agencies recommend or require at least annual

* possibility o an operator error through general usage (more ino)*. A mistake may occur the rst time you use the stockstandard solution or it may never occur with the probability increasing with use and time. In addition, the transpirationconcentration effect occurs whether the standard solution is opened /used or not and increases with use and increased vaporspace (transpiration rate is proportional to the ratio o the circumerence o the bottle opening to vapor space).

Calculations

Te concentration units or chemical standard solutions used or ICP applications are typically expressed in g/mL(micrograms per milliliter) or ng/mL (nanograms per milliliter). For example, a 1000 g/mL solution o Ca+2contains 1000micrograms o Ca+2per each mL o solution and a 1 g/mL solution o Ca+2contains 1000 ng o Ca+2per milliliter o solution.o convert between metric concentration units the ollowing conversions apply:

Te difference between ppm and g/mL is ofen conused. A common mistake is to reer to the concentration units in ppm asa short cut (parts per million) when we really mean g/mL. One ppm is in reality equal to 1 g/g. In similar ashion ppb (partsper billion) is ofen equated with ng/mL. One ppb is in reality equal to 1 ng/g. o convert between ppm or ppb to g/mL orng/mL the density o the solution must be known. Te equation or conversion between wt./wt. and wt./vol. units is:

(g/g) (density in g/mL) = g/mL and/or (ng/g) (density in g/mL) = ng/mL

Suffix

kilo- (k)

milli- (m)

micro- ()

nano- (n)

pico- (p)

= 103

= 10-3

= 10-6

= 10-9

= 10-12

= 1000 g

= 0.001 g

= 0.000001 g

= 0.000000001 g

= 0.000000000001 g

kilogram (kg)

milligram (mg)

microgram (g)

nanogram (ng)

picogram (pg)

Scientic Notation Decimal Equivalents Example Units

Table 3.1: Mass portion of concentration unit where g = gram

Suffix

milli- (m)

micro- ()

nano- (n)

pico- (p)

= 10-3

= 10-6

= 10-9

= 10-12

= 0.001 L

= 0.000001 L

= 0.000000001 L

= 0.000000000001 L

milliliter (mL)

microliter (L)

nanoliter (nL)

picoliter (pL)

Scientic Notation Decimal Equivalents Example Units

Table 3.2: Volume portion of concentration unit where L = liter


9/44


Tereore, i we have a solution that is 1000 g/mL Ca+2and know or measure the density to be 1.033 g/mL then the ppm Ca+2= (1000 g/mL) /(1.033 g/mL) = 968 g/g = 968 ppm.

(mL

A)(C

A) = (mL

B)(C

B)

For example, to determine how much o a 1000 g/mL solution o Ca+2required to prepare 250 mL o a 0.3 g/mL solution oCa+2we would use the above equation as ollows:

(mLA)(1000 g/mL) = (250 mL)(0.3 g/mL), (mL

A) = [(250 mL)(0.3 g/mL)]/(1000 g/mL), (mL

A) = 0.075 mL = 75 L

Preparation

Weight Volume

Standard chemical solutions can be prepared to weight or volume. Te elimination o glass volumetric asks may be necessary

to eliminate certain contamination issues with the use o borosilicate glass or to avoid chemical attack o the glass. It is ofenassumed that 100 grams o an aqueous solution is close enough to 100 mL to not make a signicant difference since thedensity o water at room temperature is very close to 1.00 (0.998203 at 20.0 C). Diluting /preparing standard solutions byweight is much easier. Still, the above assumption should not be made. Te problem is that trace metals standards are mostcommonly prepared in water + acid mixtures where the density o the common mineral acids is signicantly greaten than1.00. For example, a 5% v/v aqueous solution o nitric acid will have a density o ~1.017 g/mL which translates into a xederror o ~1.7%. Higher nitric acid levels will result in larger xed errors. Tis same type o problem is true or solutions oother acids to a degree that is a unction o the density and concentration o the acid in the standard solution as described bythe ollowing equation (to be used or estimation only):

dS= [(100-%) + (d

A)(%)] /100

dS= density o nal solution% = Te v/v % o a given aqueous acid solutiond

A= density o the concentrated acid used

For example, lets estimate the density o a 10% v/v aqueous solution o nitric acid made using 70% concentrated nitric acidwith a density o 1.42 g/mL.

DS= [(100-%) + (d

A)(%)]/100 = [(100-10) + (1.42)(10)]/100 = (90 + 14.2)/100 = 1.042 g/mL

Acid Content

match the standard and sample solutions to avoid a xed error in the solution uptake rate and/or nebulization efficiencysometimes reerred to as a matrix intererence. I a solution is labeled as 5% HNO

3

70% concentrated nitric acid and dilute to a volume o 100 mL then this is 5% HNO3(v/v) where the use o 70% concentrated

acid is assumed. However, nitric acid can be purchased as 40%, 65%, 70%, and > 90%. Tereore, note the concentration o theconcentrated acid used i different rom the norm as well as the method o preparation i.e. v/v or wt/wt or wt/v or v/wt. Tewt. % concentrations o the common mineral acids, densities, and other inormation are shown in the ollowing table:

Acid

Hydrochloric

Hydrouoric

Nitric

Perchloric

Phosphpric

Suluric

36.46

20.0

63.01

100.47

97.10

98.08

1.19

1.18

1.42

1.67

1.70

1.84

37.2

49.0

70.4

70.5

85.5

96.0

12.1

28.9

15.9

11.7

14.8

18.0

Mol. Wt. Density (g/mL) Wt. % Molarity

Table 3.3: Wt. % Concentrations


10/44


Acid Content in Molarity

It is important to know what the concentration units o the concentrated acid being used mean. aking 70% concentrated

nitric acid as an example means that 100 grams o this acid contains 70 grams o HNO 3. Te concentration is expressed at70% wt./wt. or 70 wt. % HNO

3. Some analysts preer to work in matrix acid concentrations units o Molarity (moles/liter). o

calculate the Molarity o 70 wt. % nitric acid we calculate how many moles o HNO3are present in 1 liter o acid. Lets say that

we tare a 1 liter volumetric ask and then dilute to the mark with 70.4 wt. % HNO3

solution to be 1420 grams. Knowing that the solution is 70.4 wt % would then allow us to calculate the number o grams oHNO

3which would be (0.704)(1420g) = 999.7 grams HNO

3per liter. Dividing the grams HNO

3by the molecular weight o

HNO3(63.01 g/mole) gives the moles HNO

3/L or Molarity which is 15.9 M. Te above logic explains the ollowing equation

used or calculating the Molarity o acids where the concentration o the acid is given in wt %:

% = wt. % o the acid

d = density o acid (specic gravity can be used i density not available)

Using the above equation to calculate the Molarity o the 70 wt % nitric acid we have:

[(70.4 x 1.42) /63.01] x 10 = 15.9 M

Dilutions o the concentrated acid to prepare specic volumes o specied Molarity can be make using the (mLA)(C

A) = (mL

B)

(CB) equation.

Avoiding Precipitates

In the preparation o mixtures o the elements, it is good to avoid the ormation o precipitates. It is common to ormprecipitates when concentrates o elements that are considered compatible (see part 1 o this series) are mixed. Manyprecipitates are not reversible (i.e., will not go into solution upon dilution). It is thereore best to add all o the acid and mosto the water to the volumetric ask or standard solution container (dilutions to weight) beore adding the individual element solution is above room temperature. Tereore allow the solution to cool to room temperature and adjust to the mark with DIwater. It is best to prepare the dilution the day beore needed to allow or proper volume adjustment.

Storage of Standards

Te ollowing are some considerations you may want to make beore the storage o chemical standard solutions:

1. Know the chemical stability o your standard. Chemical stability can be altered by changes in starting materials and

preparation conditions. It is thereore advisable to perorm stability studies on all standard solutions to avoid time consumingand costly delays or mistakes and to strictly adhere to preparation methodology, including order o addition or multi-component standard solutions.

2. Note the temperature during storage and attempt to maintain a storage temperature at or around 20 C. Some standards arenot stable or long periods at room temperature and require rerigeration or even reezing.

3. Perorm the stability study in the container material selected or storage. It is not advisable to use volumetric asks asstorage containers due to expense, contamination, and transpiration issues.

4. Determine i the standard is photosensitive and/or store in the dark i there is a concern. Tis is an issue with some o theprecious metals and is a unction o matrix. Photosensitivity will increase in the presence o higher energy light (sunlightas opposed to articial light) and trace or minor amounts o organics especially i there is an extractable proton alpha to anelectron withdrawing unctional group such as a carbonyl group. Te presence o chloride may increase instability to photo

reduction. A classic example is Ag+

in HCl solutions.5. Store the standard in containers that will not contribute to contamination o the standard. LDPE is an excellent containeror most inorganic standards.

will decrease with time.


11/44


Sample Introduction Systems4ample IntroductionS

Te most common orm o ICP sample introduction is liquid. Te purpose o this section is to introduce the beginner to themost popular components o liquid sample introduction systems used or the introduction o samples to ICP-OES and ICP-MS instrumentation (hereafer reerred to as ICP) and to alert the reader to some common problems.

SystemComponentsBeore continuing any urther, I strongly encourage you to read the ollowing: A Beginner's Guide to ICP-MS Part II: heSample-Introduction System*

In the above article, author Robert Tomas gives an excellent overview o the most popular commercially available nebulizersand spray chambers. He also provides guidance and basic theory behind the available designs, as well as an overallunderstanding o ICP introduction systems.

Te key elements o a sample introduction system start with the sipper tube and end with the torch. Tey are listed as ollows:

1. Sipper (typically plastic)

2. eon tubing going rom the sipper to the peristaltic pump tubing

3. Peristaltic pump tubing

4. eon tubing going rom the peristaltic pump tubing to the nebulizer

5. Spray chamber

6. orch

Troubleshooting

Connection Checks

Te main difficulty I have experienced with introduction system ailure is that o connections between components. Teconnections are listed as ollows:

1. Sipper to eon tubing

2. eon tubing to peristaltic tubing (both into and out o)

3. eon tubing rom peristaltic pump to nebulizer

4. Nebulizer to spray chamber

5. Spray chamber to waste drain tube6. Spray chamber to torch

I any one o these connections is not airtight, the operator will experience anything rom poor precision to an inability tolight the plasma. One o the many reasons I preer concentric glass nebulizers is that they are ree ow (i.e., the liquid willow rom the sample container to the nebulizer without assistance rom the peristaltic pump). A simple check is to determinei you obtain a ne steady mist (using water as the sample) without the peristaltic pump (pressure lever released) so that reeow can occur. Tis can be done with the nebulizer disconnected rom the spray chamber (plasma has not yet been lit) so thatthe mist can be easily visualized. You can also check or the appearance o any small air bubbles in the eon tubing, whichshould never be present and indicate a poor connection somewhere between and /or including the sipper and the nebulizer.

Another connection that is ofen taken or granted is the spray chamber drain/waste tube connection. Tis connectionis absolutely critical. One way to test this connection is to put some water in the spray chamber using a wash bottle and

determine i it drains smoothly and without leaks. Poor precision or the inability to light the plasma is a common symptomo a poor drain tube connection. During this test you should also observe the absence o water droplets in the spray chamber(assuming glass construction). A dirty spray chamber will leave water droplets and cause poor precision and carryoverproblems. Make sure the plasma is not lit whenever you perorm this test.

Spray Chambers

Spray chambers can be made o all glass, all plastic, and glass with plastic end caps. I you do not use HF (all plastic systemsmust be used with HF) and thereore have the luxury o using glass components, attempt to use a spray chamber without


12/44


the plastic end cap (i.e., all glass). Tey are typically used with glass concentric nebulizers and use only two O rings toconnect the nebulizer to the spray chamber. I have ound that the plastic end cap may cause longer washout times, carry overproblems, and is a very large connection surace where connection problems can occur. Using a glass concentric nebulizer and

all glass spray chamber a precision o between 0.2 and 0.5% RSD should be observed. I an all glass system gives a precision o1% RSD or greater, then there is most likely a connection problem or the nebulizer gas ow rate is too high (look or spittingwhen checking the nebulizer ree ow and do not be araid to lower the gas pressure {argon sample ow} to the nebulizer).

Peristalic Pump Tubing

and the pressure can be set to give a steady mist when the pump in running. Te problem is that the pump tubing stretchesand either the pressure is not enough to drive the solution through the tubing or you over tighten and get a pulsating mistspray. Tis is a problem that each analyst has to be aware o and solve through experimentation. Tis problem is particularlytroublesome or ICP-MS users because the argon ow changes as the tubing stretches. Tis causes a relative increase in thesensitivity o the higher atomic number elements.

MaintenanceI preer glass components because o their ease o operation and cleaning. It is always best to start the day with a cleannebulizer, spray chamber, and torch. Cleaning the torch daily will also extend its lie. Tere are many cleaning solutions thatcan be used. Some o our analysts preer 1:1 nitric acid/water and others preer suluric acid and hydrogen peroxide.Another common cleaning solution is 1:1 HCl/nitric. All o these solutions will work depending upon the nature o thecontaminants. Te suluric/peroxide is generally a severe approach and needed only i organics such as grease are suspected.

Be advised that ultrasonic baths are great or cleaning. However, NEVERuse them to clean a glass concentric nebulizer.Glass concentric nebulizers are cleaned by leaching and occasionally by applying a backpressure with water to remove lodgedparticles. Te use o a cleaning wire or ultrasonic bath is a sure way to destroy the nebulizer.

In summary, when it comes to ICP introduction systems there is no substitute or experience. Relatively speaking,

introduction systems are simple but they are not easy to maintain and they are challenging to operate to their maximumpotential.

Nebulizers, Spray Chambers and Torches5Tere has been a tremendous activity in the area o sample introduction over the past 30 years since ICP has been

commercially available. Te objective o this section is to acquaint the reader with the basic options available to the ICP

operator or the introduction o liquid samples.

Some o the considerations in selecting an introduction system include dissolved solids content, suspended solids presence,

presence o HF or caustic, detection limit requirements, precision requirements, sample load requirements, sample sizelimitations, and operating budget. In the last section, the concentric nebulizer and all glass introduction systems were given

top billing but they may not work at all or your application. Te analyst is lef with the task o choosing the best introduction

components afer taking into account the appropriate considerations.

Nebulizers

Pneumatic Nebulizers

Te term pneumatic is dened as o or relating to or using air or a similar gas. Te word nebulizer is derived rom the

Latin nebula meaning mist and is dened as an instrument or converting a liquid into a ne spray. Tereore, a pneumatic

nebulizer is literally an instrument or converting a liquid into a ne spray that uses a gas as the driving orce.

Some o the most popular ICP pneumatic nebulizers are:


13/44


Te concentric and xed cross-ow are still the most common designs. he construction o both types is described in the ollowingarticle by ICP expert Robert homas (see Figures 4 & 5): A Beginner's Guide to ICP-MS Part II: he Sample-Introduction System *ICP manuacturers will give you an option as to the type o nebulizer to use depending upon your analytical requirements and

the instrumental design.

Ultrasonic Nebulizer with Axial ICP-OES*

Sound can be used instead o a gas as the energy source or converting a liquid to a mist. Tese nebulizers use an ultrasonic

generator at a requency o between 200 kHz and 10 MHz to drive a piezoelectric crystal. A pressure is produced that breaks

the surace o the liquid - air interace. Ultrasonic nebulizers are more expensive and difficult to use but they will improve(lower) detection limits by about a actor o 10.

Spray Chambers

Te basic designs that have remained over the years are the Scott

double-pass and the Cyclonic. o review the designs o these two

components, see Figures 8 & 9 in Robert homas' article: A Be-

ginner's Guide to ICP-MS Part II: he Sample-Introduction System*

Te Cyclonic design is relatively new but is very popular. Te

purpose o the spray chamber is to remove droplets produced by

the nebulizer that are > 8m in diameter. Considerations includethe wash-in-time, washout time, stability, and sensitivity. Te

drainage characteristics are important in part due to pressure

changes that may occur during drainage. It is important that

the drainage process be smooth and continuous. Te analyst

may observe aster washout times with the Cyclonic design. Te

chamber material o construction as well as the sample matrix

and the chemistry o the element will inuence the washout time.

In addition, the analyst may observe aster washout times with glass construction than with polymers. Tis is due in part to

better wet ability o the glass (lack o beading). Both designs are excellent and the analysts may wish to experiment with eachto determine which yields the best perormance or their specic analyses.

TorchesTe two basic torch designs are the Greeneld and Fassel torches. Te Greeneld torch requires higher gas ows and RF

powers. Te Greeneld torch is more rugged (less likely to extinguish due to misalignment and introduction o air) whereas

the Fassel torch requires less Ar and power. Both designs produce similar detection limits.

Some nebulizer designs work better with one torch design over another. Beore experimenting with torches, it is best to

contact your instrument manuacturer to determine the torch design recommended or your instrument as well as any design

specications, operating conditions, and dimensions that must be observed.

Considerations

Te ollowing are some questions you may want to consider, whether you are looking to purchase a new ICP or already have

one or more existing units:

your annual Ar expense).

Assorted spray chambers

For more inormation on ultrasonic nebulizers, visity the ollowing

link: CEAC U-5000+


14/44


Solutions Containing HF

Te presence o HF causes the vast majority o compatibility problems between the sample matrix and the introductionsystem components. I you are preparing samples containing one or more o the ollowing elements, then you are likely usingHF in your sample preparation:

When HF Attacks

Te introduction o solutions containing HF should be o concern to the instrument operator, especially i he/she is concentration o the HF and the type o glass or quartz. It is the HF molecule that does the attacking; not the uoride anion(F-1). Tere is absolutely no attack by neutral solutions o F-1upon any orm o glass or quartz (note that there is watersolubility o amorphic and crystalline orms o silica that is a unction o the surace area, impurities, and structure).

Te HF attack is enhanced by the presence o a strong acid, such as HNO3or HCl, by:

1. Increasing the relative amount o HF through a shif in the equilibrium o equation 6.1 below and;2. By adsorbing as the hydronium ion on the solid silica surace where it behaves as a catalyst (i.e., the reaction o HF with asolid silicate can be described by two equations that work in parallel).

In addition, the crystalline orm o the silicate inuences the rate o attack. Te net result being that quartz is not attacked asreadily as glass. (Tis is a generalization - please note that there are four production types of quartz in addition to natural quartzwhere different solubility and contamination characteristics can be expected from each. It may be more appropriate to think ofglass as amorphous silica and quarts as structured or better yet crystalline silica).

Equation 6.1:

H+1+ F HF (K = 8.9 x 10 )-1a

-4

It ollows that solutions containing HF that are neutralized with a base to eliminate HF will not attack silicates provided thatthe HO-1concentration is not too high (i.e., the pH is not above 8). Tis is why organic amines such as triethanol amine are sogood at eliminating HF attack simply through neutralization o the HF as opposed to NaOH, which will attack silicates i highenough in concentration.

Compatibility and Precision Issues6

=


15/44


introduction components. It is common practice to react HF with boric acid (typically, 1 gram o boric acid is added or every1 mL o 49 % HF) to orm the mono-uoroboric acid. Unortunately, uoroboric acid will attack glass (including concentricnebulizers) and the attack o silicates, in general, is not greatly altered. Te ormation o the uoroboric acid will diminish thetendency to orm insoluble uorides such as CaF

2which is why it was originally added.

Glass Introduction Systems

Glass introduction systems are generally preerred by analysts because they are less expensive, have shorter washout times,and give better precision than plastic. Tis is why many analysts opt to use all-glass introductions provided the HF contentis < 100 ppm. Quartz is less reactive than glass and is sometimes used i the analyst is concerned with making low level Bmeasurements in a trace HF matrix.

Our laboratory uses a ype C glass concentric nebulizer at an Ar ow o ~ 0.75 L /min, a pressure o 30-35 PSI, and a sampleintroduction rate o 0.7 mL/min. Te spray chamber is an all glass cyclonic and the torch is made o quartz. A typicalmeasurement precision is between 0.2 and 0.5 % RSD and the washout times are excellent or all elements, including B and Hg( Hg takes ~ 75 seconds o rinse with 10 % (v/v) HNO

3). race levels o HF are easily tolerated even when elements such as Si

and B are measured.

Recommendations

HF concentrations 0.1 % will attack both glass and quartz and cause considerable problems or the analyst attempting todetermine Si, B, or Na. It is necessary to either switch to an HF-resistant introduction system or neutralize the HF with a base.Our laboratory introduces 1000 to 20000 g/mL solutions o all the HF elements using the neutralization (triethanol amine)option with the addition o H

4EDA when required or chemical stabilization, while other laboratories get excellent results

using the HF-resistant (plastic) introduction systems. Te PFA concentric nebulizer is popular with a PFA or PEEK spraychamber and Al

2O

3(inner tube) torch. I would suggest checking with your instrument manuacturer or power supply and gas

ow compatibility beore investing in an HF resistant system.

High Dissolved SolidsFor conventional xed cross-ow and concentric nebulizers, high dissolved solids may be a problem. Te problem lies in thesalting out o the matrix component(s) in the nebulizer. Tis occurs in the nebulizer at the point where the solution goesrom a liquid to a mist, resulting in a temperature drop and reduced solubility. I the solution component is well below its

Te answer is relative to the solubility o the matrix. I you are aspirating a 0.7 % solution o B as boric acid then salting outwill occur. A 4 % solution o Cu as the nitrate or chloride will not salt out. Salting out is indicated by poor precision and agradual loss o signal. Te analyst has several options:

1. Dilute the sample.2. Humidiy the sample Ar stream.3. Use one o the high solids or high pressure concentric nebulizers mentioned in part 5 o this series.4. Increase the solubility o the culprit.

Our laboratory uses option 1 or 4 in order to retain the excellent characteristics o the type C concentric glass nebulizer. Teaddition o EA is made to high boric acid solutions. Tis greatly increases the boric acid solubility and eliminates salting out.Other matrices are best dealt with through dilution, where the lowest concentration o the matrix metal that can be toleratedby a type C concentric - in our experience - is 10000 ppm.

Suspended SolidsSamples containing suspended solids may cause a problem with the conventional xed cross-ow or concentric nebulizersdepending upon particle size. Solids that will pass through a 0.3 m lter will not plug these nebulizers and will behave asi they are in solution with respect to the entire sample introduction process. Particles > 10 m will not aspirate normallyand are not likely to cause plugging. Many sample types have particulate that is easily visible to the naked eye and will causedifficulty with the cross-ow and concentric nebulizers. Te Babington V-Groove, GMK Babington, Hildebrand dual grid,Ebdon slurry, Cone Spray, and Noordermer V-groove nebulizers are all popular choices. Other options include ltration toremove the solids and chemical treatments such as usion, ashing, or acid digestion to dissolve the solids.

Closing RemarksHF, high dissolved solids, and suspended solids are the most common compatibility issues acing the ICP analyst. Te waysaround these problems are ofen expensive, time consuming, and result in lowered detection limits, longer wash out times,and poorer precision. In extreme cases, alternate analytical measurement techniques are required. It is always best to consultwith your instruments manuacturer beore switching introduction components outside the realm o those recommended/supplied by the manuacturer.

Tere is a general misunderstanding that the addition o boric acid will eliminate HF attack, allowing the analyst to use glass


16/44


erformance CharacteristicsLinearity and Detection LimitsP7

Dening ICP Performance Characteristics

Te ollowing steps are intended as a practical guide or the determination o an ICPs perormance characteristics:

1. Read the operating manual and amiliarize yoursel with the sofware, key instrumental parameters and preerred settingsbeore the instrument is installed.

Most instruments are supplied with optimization and wavelength or mass calibration standards that will be used during set-upby the service technician and are intended for use on a regular basis by the operator. Discuss the optimization process with themanufacturer as well as the preferred settings for the key instrumental parameters.

Te remaining steps assume that the operator fully understands and is able to perform the optimization process that has beendened by the manufacturer as well as the spectral limitations of the instrument.

2. Select the lines to be studied or each element (lines is used in this document to mean either wavelength or mass).

Line selection is based upon spectral interference issues, detection limit requirements and working range requirements. Select asmany lines as possible within practicality for each element. Te greater the number of lines, the greater the exibility.

3. Prepare single element standards over the anticipated working range or each element. Te range o standards dependsupon the analytical requirements. Te ollowing ranges are suggestions only:

Tis step is important because these data can be used to determine instrument detection limits (IDL), linear working ranges, modern (i not all) instruments, the spectra obtained or each element at each concentration can be saved or review later. Inaddition, the sofware will calculate the IDL and BEC plus the linear regression o each line will establish the linear workingrange. All o this is typically done or the operator by the sofware that comes with the instrument. I at all possible, attempt to:

standards manuacturers provide this inormation with their single element standards. Tese data are important in identiyingdirect spectral overlap intererences and in not identiying an impurity as an intererence o this type.

you are interested in possibly using up to 6 lines or roughly 72 elements, then each solution spectrum totaling 72 x 6 =~ 432 lines per solution and ~ 432 x 5 = 2160 spectra or each element need to be stored or uture reerence. Most ICP-MSapplications would require ar ewer data to be collected due to the reduced number o lines available and /or easible.

beginning o each element concentration series. Look or the presence o the prior element analyzed to conrm that it hasbeen completely washed out o the introduction system.

4. Having the data available on a desktop computer is convenient and allows the analyst to construct potential spectra bycalling up the element and the anticipated concentration or each element in the analytical sample. Having several linesavailable makes the job o line selection easy as well as the estimation o the lines sensitivity and linearity. Constructingthese composite spectra rom pure single element solutions eliminates conusion as to the identity o the line. Te ollowingexample is intended to illustrate the process:

Examples of Spectra

FYI: All spectra were obtained using a concentric glass nebulizer with no problems around salting out or plugging.


17/44


Te ollowing example is or an application where a submitter has been obtaining minor levels (0.1 to 1.0 %) o Cr in analloy containing roughly equal amounts o Fe and Ni. Te laboratory where this alloy is analyzed uses a procedure where 0.2grams o the sample is dissolved in 5 mL o a 1:1 HNO3 /HCl mixture and diluted to 1000 mL with DI water. Te analyst is

inormed that a limit o detection (LOD = 3SD0) o 1 ppm Cr based upon the original sample and the ability to quantiy theCr to within 10 % relative at the 10 ppm level is an absolute minimum requirement.

calculation and determines that using the most sensitive Cr line and the current procedure, the lowest possible detection limitis 4 ppm and a more realistic estimation would be ~ 4 times the IDL or ~ 16 ppm. Te analyst then pulls up the ollowingspectra, instrument detection limits, and linear regression data which were obtained on their radial view instrument aboutour years ago when installed using pure single element solutions as described above.

Te 205.552 nm Cr line was ound to be the most sensitive o the 16 Cr lines originally characterized with an IDL o 4.0 ppm= [ (0.0008 g/mL Cr IDL) x 1000 ] /0.2 based upon original sample size and dilution as described above. However, thespectrum o a 0.1 ppm Cr standard shows signicant intererence rom both Ni and Fe at a concentration o 100 ppm makingthe line useless at low ppm Cr levels (see Figures 7.1 and 7.2).

Te analyst then begins the relatively simple process o identiying a Cr line with the most sensitivity that is spectrally clean.Figures 7.3 and 7.4 show the line identied using the same scan data shown or the 205 Cr line. Te 267.716 nm Cr line looksclean at the current dilution actors and has an IDL o 0.0016 g /mL Cr which increases the detection limit to somewherebetween 8 to 32 ppm.

Te good news is that the 267.716 line looks spectrally clean and the possibility o increasing the sample size while loweringthe nal volume by a actor o 100 is possible (i.e., 2 grams sample up to 100 mL using 20 mL o 1:1 HCl/HNO

3). Te

concentrations o the Fe and Ni in the nal solution would be ~ 10,000 g /mL each. Tis capability was conrmed when40,000 g/mL solutions o both Fe and Ni were scanned as shown in Figure 7.5. Tese spectral data indicate a realisticdetection o


18/44


Te spectra in Figure 7.5 were used to articially produce Figure 7.6 which approximates signals that would be measured ora Fe/Ni alloy where 2 grams to 100 mL dilution were made on a sample containing 1.25 ppm Cr. Te entire investigationwas perormed using spectra that had been stored on computer (i.e., the analyst can literally provide an answer as to projecteasibility while speaking on the phone with the client).

Te above process is not intended to take the place o method validation, but rather to arm the analyst with sufficient data tomake intelligent choices during the initial stages o method development.

Conrm Basic Performance Criteria

Te ollowing excerpt was taken rom Part 17: Method Validation*o our race Analysis series*. Tis section discussesperormance criteria conrmation during the method validation process. Please note that the validation process is more

detailed and specic.

Te method must t the purpose as agreed upon between the client and the analyst. In the case o trace analysis, theollowing criteria are typically evaluated as part o the method development process:

Specicityinvolves the process o line selection and conrmation that intererences or the ICP-OES or ICP-MSmeasurement process are not signicant. A comparison o results obtained using a straight calibration curve (without internalstandardization to that o internal standardization and/or to the technique o standard additions) will give inormationconcerning matrix effects, drif, stability, and the actors that inuence the stability. Te various types o spectral intererencesencountered using ICP-MS and ICP-OES should be explored.

Accuracy or Biascan be best established through the analysis o a certied reerence material (CRM, or SRM i obtainedrom NIS). I a CRM is not available, then a comparison to data obtained by an independent validated method is the next

best approach. I an alternate method is not available, then an inter-laboratory comparison, whereby the laboratories involvedare accredited (ISO/IEC 17025 with the analysis on the scope o accreditation) is a third choice. Te last resort is an attempt toestablish accuracy through spike recovery experiments and/or the use o standard additions.

Repeatability(single laboratory precision) can be initially based upon one homogeneous sample and is measured by thelaboratory developing the method. Te repeatability is expressed as standard deviation.

Limit of Detection (LOD) is a criterion that can be difficult to establish. Te detection limit o the method is denedas 3*SD

0, where SD

0is the value o the standard deviation as the concentration o the analyte approaches 0. Te value o

SD0can be obtained by extrapolation rom a plot o standard deviation (y axis) versus concentration (x axis) where three

concentrations are analyzed ~ 11 times each that are at the low, mid, and high regions o interest. Tis determination shouldbe made using a matrix that matches the sample matrix.

Sensitivityor delta C = 2 (2)1/2SDc, where SD

cis the standard deviation at the mid point o the region o interest. Tis

represents the minimum difference in two samples o concentration C that can be distinguished at the 95% condence level.

Limit of Quantitation (LOQ)is dened as 10 SD0and will have an uncertainty o ~ 30% at the 95% condence level.

Linearity or Rangeis a property that is between the limit o quantitation and the point where a plot o concentration versusresponse goes non-linear.

Figure 7.5:Spectra o pure 40,000 ppm Fe and Ni solutions, 0.1 ppm Cr

and a water blank at the 267.716 nm Cr wavelength

Figure 7.6:Simulated spectrum o a solution produced rom 2 grams 100 mL solution o a 50 /50 wt.

% Ni/Fe alloy containing 1.25 ppm Cr at the 267.716 nm Cr wavelength


19/44


Spectral Interference: Types, Avoidance and Correction

8Types of Spectral Interference: ICP-OESTe types o spectral intererences most commonly encountered or ICP-OES are discussed in the Spectral Intererencessection o Part 15: ICP-OES Measurement o our Reliable Measurements series. You may wish to review this inormationbeore continuing.

As noted in part 7 o this guide, the collection o spectra at different concentrations on all elements and lines available willsave a lot o time in the line selection process.

Avoidance: ICP-OES

Several modern ICP instruments have the capability o avoiding the spectral intererence by going to another line. Manyinstruments can make measurements simultaneously on several lines or 70+ elements in the same time it used to take tomake a measurement on a single line/element combination. I you have the opportunity, I would strongly encourage theavoidance approach over attempting to make correction on a direct spectral overlap or wing overlap intererence. Backgroundcorrections are another manner and can be routinely dealt with.

Correction: ICP-OES

Background Interference

Background radiation is a potential source o error that requirescorrection. Te source o the background radiation is rom a

combination o sources that cannot be easily controlled by theoperator. Figure 8.1 shows the spectra or a highly concentratedCa sample as compared to a nitric acid blank.

Te background radiation intensity or the nitric acid blank is~ 110,000 counts at 300 nm whereas the background radiationor the Ca containing solution is ~ 170,000 counts at the samewavelength. Although background radiation can be loweredsomewhat by adjusting instrumental parameters, it cannot beeliminated and corrections are typically necessary. It can be seenthat the highly concentrated Ca matrix contributes some to thebackground radiation but there are greater contributions rom other sources independent o the sample matrix.

It can be argued that matrix-matched standards and samples will eliminate the need or background correction where theanalyst only has to measure the peak intensity. It would ollow that the precision o the measurement would be better (lower)and or some instruments the measurement time will be shorter. However, the problems with matrix matching are obviousand may offset any advantage gained when you dont make them.

Te correction or background radiation is typically made by rstselecting background points or regions and then a correctionmode or algorithm. Te algorithm or correction mode dependsupon the curvature o the background, as is illustrated below.

Figure 8.2 shows a at background where correction was madeon both sides o the line. In this case the instrument allows

or the selection o background regions thereby improvingthe accuracy o the estimated background radiation. I theinstrument only allows or selection o background points thenintensities are taken at set wavelengths, averaged and subtractedrom the peak intensity. For at backgrounds the distance o eachpoint rom the peak intensity is not important provided thereis no intererence rom other lines in that vicinity. Figure 8.2

Figure 8.1:Spectrum o 6% Ca solution vs. nitric acid blank

Figure 8.2:Flat background correction


20/44


21/44


It was assumed that the precision o measuring the intensity o the As or Cd contributions at 228.802 nm is 1%. In addition, itwas assumed that the best-case precision or making a correction is calculated using the ollowing equation:

SDcorrection= [(SDCd I)2+ (SDAs I)2]1/2

where:SD

correction= standard deviation o the corrected Cd intensity;

SDCd I

= standard deviation o the Cd intensity at 228.802 nm;SD

As I= standard deviation o the As intensity at 228.802 nm

a best-case detection limit or Cd at 228.802 nm in the presence o 100 ppm As would be 2 x SD

correction, then the calculated

detection limit is 0.1 ppm. In reality, the detection limit wouldbe closer to .5 ppm. Te detection limit or the Cd 228.802 nmline is 0.004 ppm (spectrally clean) showing roughly a 100-oldloss. Furthermore, the lower limit o quantitation has been

increased orm 0.04 ppm (10 x the DL) to somewhere between1 and more realistically 5 ppm Cd. Figure 8.6 illustrates thesituation with the spectra o 1 and 10ppm Cd solutions withand without 100 ppm As present.

Correcting or the intererence o As upon Cd would requirethat (1) the As concentration in the solution be measuredand that (2) the analyst already have measured the counts /ppm As at the 228.802 nm line (sometimes called correctioncoefficient). Tis inormation allows or a correction bysubtracting the calculated intensity contribution o As uponthe 228.802 nm Cd line, thereby making the correction.Tis approach urther assumes that slight changes in the

instrumental operating parameters and conditions will inuence both the analyte (Cd) and the interering element (As)equally (i.e., an assumption many analysts are not willing to make).

Te problems associated with direct spectral overlap make it difficult or the analyst to perorm quantitative measurements.Each case should be reviewed. I a spectral correction is ound to be necessary, the reader is advised to consult their operatingmanual where a dened procedure will be outlined using the instruments sofware.

Types of Spectral Interference: ICP-MS

he types o spectral intererences most commonly encountered or ICP-MS are discussed in the Intererences section o Part16: ICP-MS Measurement*o our Reliable Measurements series. You may wish to review this inormation beore continuing.

Avoidance: ICP-MS

Te ollowing are possible avoidance pathways:

Conc. Cd

ppm

B

1000

100

10

1

A

0.1

1

10

100

C

13193

124410

1242401

11196655

D

132

1244

12424

111967

E

672850

672850

672850

672850

F

6729

6729

6729

6729

G

6730

6843

14129

112169

H

5100

541

54

6

I

51.0

5.5

1.1

1.0

Rel conc.

As/Cd

Cd 228.802

net intensity

Estimated

SD on clean

Cd line

100 ppm

As Net

Intensity at228.802

Estimated

SD on 100

ppm As at228.802

Estimated

SD of 100

ppm As +

corr. Cd conc

at 228.802

Uncorrected

Relative

Error (%)

Best- Case

Corrected

RelativeError (%)

Table 8.1:Estimated Errors o As on Cd 228.802 nm line

Figure 8.6:1 and 10 ppm Cd with and without 100 ppm As


22/44


Te above approaches are just examples o some o the approaches that have been taken to avoid intererences. For a givenapplication, it is suggested that a literature search be perormed in an attempt to benet orm the vast amount o research

that has been conducted in this area. In addition, instrument manuacturers are constantly revising and updating theirinstrumentation and sofware in an attempt to take advantage o new technologies. Tus, consulting with the manuacturermay help when intererences are encountered.

Te act is that the mass spectra o elements are much less detailed than in optical emission spectroscopy. Most elements have go to another isotope even i it is less abundant. Te difficulty in obtaining low detection limits in ICP-MS with intererencecorrection is a unction o the relative signal intensities and measurement precision as illustrated above or ICP-OES. I acorrection cannot be avoided, many analysts seek alternate techniques rather than run the risk o reporting unreliable data.

Key Instrument Parameters9Te perormance characteristics o an ICP is a unction o a variety o instrumental parameters. Current instrumentationhas many parameters that are xed by the manuacturer and all instrumentation will come with recommended settings orthose parameters that are not. Te purpose o this section is to point out the key parameters that will require adjustment ona regular basis. Tis discussion will be limited to the introduction o the analyte as a nebulized solution and Ar as the plasmagas.

Gas Flow Rates

Tere are three gas ow rates or the common torch designs. Te outer gas ow is sometimes reerred to as the coolant orplasma gas ow; the middle or intermediate gas ow is sometimes reerred to as the auxiliary gas ow; and the central gas

ows do not have a great impact upon the perormance characteristics and the values suggested by the manuacturer shouldbe used or common applications. However, the sample gas ow rate will vary between nebulizers o the same design andrequire adjustment on a regular basis.

Sample Ar Gas Flow for ICP-OES

Assuming sample solution is not signicantly limited, the main consideration when adjusting the sample Ar gas ow is thato precision. Increasing the sample Ar gas ow does not necessarily increase the emission intensity. Te objective in settingthis ow rate is to obtain the best detection limit. Noisy signals will typically result rom higher ow rates that will serve todegrade the stability o the plasma, increase the short-term measurement precision and consequently give poorer detectionlimits.

Te ollowing considerations should prove helpul:

determining the optimum ow setting or a given nebulizer.

matrix deposition, or an ailing mass ow controller are possible causes or a change in the optimum setting or an inability toreproduce the same precision as when the nebulizer was new.

Applied Power for ICP-OES

Te second key parameter that the operator may wish to vary is the applied power. Higher applied power will increase the netsignal intensity but not necessarily improve the detection limit.


23/44


view). Over the years manuacturers have determined the optimum power and observation height settings. Tereore, rst tryusing the settings recommended by the manuacturer.

IMPORTANT: Sample Ar Gas Flow cannot be separated rom Applied Power and Sampling Depth or ICP-MS.

Te sample Ar gas ow or ICP-MS systems is a parameter that is more complex than with ICP-OES instrumentation.Assuming the goal is to obtain the maximum signal intensity, the Ar gas ow is closely related to the applied power andsampling depth. Tere is not a single set o optimum power, sampling depth, and sample Ar ow settings. For example, a

higher applied power will increase the signal intensity but change the optimum sampling depth and sample Ar ow. However,the higher sample Ar ow rates required at high power bring about some degradation in other perormance characteristics. Ithe applied power is constant or every method, then the optimum sampling depth will change as the sample Ar is changed.Te consideration o MO (metal oxide) ormation and different sensitivities at different mass ranges must also be made withincreased sample Ar ow.

Here are some nal observations that may prove useul:

relative to the heavy masses.

your particular model instrument using a suite o elements covering the mass range. A mixture o Mg, Rh, Ce, and U shouldsuffice where the CeO and Ce+2masses are measured as well.

Te above observations may seem conusing, but in reality they give the operator a degree o exibility that the ICP-OESoperator does not have in that you can optimize the instrument or selected mass ranges. For example, we know that a higher

the nest step is to adjust the sampling depth to give the optimum signal while aspirating a solution containing a combinationo light, mid-range, and heavy elements such as Mg, Rh, Ce, and U. I the double ion or MO signals are higher than desirable,a reduction in the peristaltic pump tubing diameter or pumping speed should lower these signals. Tese initial adjustmentswill take a lot o time and patience but they are well worth the effort. As the operator makes adjustments in these keyparameters, a pattern will begin to unold allowing the operator to optimize the instrument or selected mass ranges.

It is suggested that new ICP-MS operators take the time to determine the trends when changes in applied power, sample Arow, sampling depth and peristaltic pump speed are made.

Te ollowing inormation may prove useul:


24/44


Calibration Curve and Standard Additions Techniques10 alibration TechniquesCBoth the accuracy and precision o ICP measurements is dependent, in part, upon the calibration technique used. Tis sectionis ocused upon errors (both xed and random) that can be introduced through the use o different calibration techniquesusing accurate calibration standards, samples that have been prepared accurately to within dened error limits, and aninstrument that has been set-up correctly using a procedure programmed where there are no spectral/mass intererences thatinclude background correction. You may believe that i the above errors have been conned to within acceptable and knownlimits that there is nothing else to worry about. Unortunately, this is not the case.

Te most common calibration technique options or ICP measurements are calibration curve and standard additions. Inaddition, the option o using internal standardization is available or the calibration curve technique and the ability o matrixmatching may also be available. ICP-MS has the added option o using an internal standard that is an enhanced isotope othe element being measured (i.e., isotope dilution ICP-MS). Tis discussion will be limited to the above approaches, to theintroduction o the analyte as a nebulized solution, and to the use o Ar as the plasma gas.

Basic Considerations

Beore reading ahead, it may be helpul to restate the assumptions made above and make some additional considerations:

measurement is a comparison process. instrumental response that is described by the equation or a straight line. uncertainty o the prepared standard solution is known and has been calculated). dened limits o time, matrix, concentration, temperature/humidity, and container material(s). o the analyst. Tis assumption is made to allow us to ocus completely upon the potential errors involved with the calibrationprocess. errors. It is thereore assumed that the uncertainty in preparation can be described by the random and known sampling,

weighing and volume dilution errors. Again, this is an assumption that is ofen not the case but is made to allow us to ocuscompletely upon the potential errors involved with the calibration process.

Calibrations Standards

ICP is a matrix-dependant technique. Based upon the above assumptions and the act that ICP is a comparative method, theprime concern is the availability and use o appropriate calibration standards. Te problem analysts ace is that ICP (ICP-OESand ICP-MS) is extremely matrix-dependent. Tereore, the ideal situation is that the matrices o the standards and samplesbe identical.

Recommendations

Tis section lists several recommendations. Discussions relating to these recommendations are provided in the next sectionor the reader who would like more detail.

Recommendation (a)- Match the acid content o your calibration standards and samples in both the type o acid used and theconcentration o the acid.

Recommendation (b)- Match the elemental matrix components o your calibration standards and samples to the greatestextent possible. In this situation, the analyst who knows the composition o the sample has this capability.


25/44


Recommendation (c) technique o standard additions. However, this approach is slow as compared to the calibration curve technique with the useo internal standardization.

Recommendation (d)- Te use o internal standardization is very effective in many cases but may introduce--or not corrector--all errors. Tis statement does not apply to isotope dilution ICP-MS that is considered to be a primary analyticaltechnique.

Recommendation (e)- Chemical calibration is an approximation at best. Te analytical chemist must be constantlyaware o the possibility o bias introduced by the nature o the standards used, which may be the major source o bias inthe analytical data. Appropriate reerence materials should be used to evaluate this and other aspects o the measurementprocess.1

Discussions

Discussion (a and b)- Te matrix will inuence the nebulization efficiency, which is proportional to the signal intensity.Nebulization efficiency is the percent o solution that reaches the plasma. Tereore, i the nebulization efficiency is 1 %, then

99 % o the solution is going to waste and 1 % is making it to the plasma. ypically, nebulized solution mist particles that aregreater in diameter than 8 microns will go to waste. I a matrix component changes the efficiency rom 1.0 % to 0.8%, thena relative drop o ~ 20 % would be expected rom this effect alone. Te droplet size distribution o a pneumatic nebulizer isgoverned by the physical properties o the solution as well as the volume ow rates o liquid (inuenced by peristaltic pumpspeed and tubing diameter) and gas (sample Ar ow rate). Te physical properties claimed to inuence the droplet sizedistribution are the surace tension, viscosity, and density.See Inductively Coupled Plasmas in Analytical Atomic Spectrometry;Montaser, A., Golighty, D. W., Eds.; VCH Publishers: New York, 1992 - page 703 for more detail and additional references on this topic.

For the ICP analyst, the most common matrix component that will alter the physical properties o a solution is the acidcontent. Tis is not to say that other differences such as the presence o trace organics (added intentionally or not) shouldnot be considered. However, the identity and concentration(s) o one or more acids is an issue that virtually all ICP analystshave to decide upon. Te ICP analyst is most commonly involved in the preparation o samples where one or more inorganic

mineral acids are required to bring about dissolution o the sample and/or to maintain solution stability o the analyte(s) ointerest. Te acids most commonly used are HNO3,HCl, HF, HClO

4, H

2SO

4, and H

3PO

4and are listed in the order o best to

worst.

Te effect o acid matrix upon nebulization efficiency is such that a change in acid content rom 5 to 10 % v/v will cause adecrease in efficiency o 10 to 35 % depending upon the acid used, nebulizer design and liquid and gas ow rates. Matchingthe matrix to within 1 % relative is necessary or the most accurate (we use the term assay) work (i.e., a 5 % HNO

3acid

solution would be made to 5.00 0.05 %.

Te matrix will inuence the plasma temperature, which is related to the signal intensity or ICP-OES. Te other effect matrixcomponents have on the ICP cannot be explained by a change in nebulization efficiency. Te effect is one where the matrixcomponents give the appearance o taking power away rom the plasma (lowering the temperature o the plasma). It has beenreported that this effect is related to the excitation potential o the line and that the effect increases as the excitation potential

increases. A similar effect would be seen by decreasing the applied RF power or by increasing the sample (nebulizer) Ar owrate since both result in a reduction o the plasma temperature. Tereore different lines o the same element would be affecteddifferently according to their excitation potentials. In addition, when choosing an internal standard element it ollows that theexcitation potentials o the internal standard and analyte lines should be as close as possible, unless the calibration standardsand samples are matrix matched. For more inormation and additional reerences, see:

Inductively Coupled Plasmas in Analytical Atomic Spectrometry; Montaser, A., Golighty, D. W., Eds.; VCH Publishers: New York,1992 - pages 279-281.

ICP-MS suffers rom nonspectral matrix effects. Te effect most commonly encountered is reerred to as quenching andis thought to be due to deocusing o the ion optics by space charge effects. Generally, as the concentration o the matrixelement(s) increases, the analyte signal will be suppressed. Quenching increases in effect as the matrix element absoluteconcentration increases, the matrix element mass increases and the analyte mass decreases. Tis effect is absolute in nature

and not a unction o the relative concentrations o the matrix elements and analyte elements. Tereore, when sensitivityallows, it can be diluted out. It is also greater in effect as the RF power is lowered. Te effect is such that an element matrix

Inductively Coupled Plasma Mass Spectrometry; Mantaser, A., Ed.; Wiley-VCH: New York, 1998 - page 543.


26/44


Internal Standardization and Isotope Dilution

11Matrix effects are arguably the subtlest danger to the ICP-OES analyst. Slight differences in the matrix can cause aconsiderable systematic error. Te most common calibration technique options or ICP measurements are calibration curveand standard additions.

Standard Additions

Te technique o standard additions is used when the matrix is quite variable and/or when an internal standard that correctsor plasma related effects couldnt be ound. Tis technique is also useul in conrming the ability o an internal standardcalibration curve technique to correct or both nebulizer and plasma related effects (see part 10 o this series or more onnebulizer and plasma related matrix effects). Te ollowing considerations may prove useul in perorming the technique ostandard additions:

remove exactly 50.00 grams o solution to a separate clean container or spiking. range or each analyte. spike the unknown solution with a concentrate o the analyte(s) o interest to levels o between 2x

and 3x

where x

represents

the unknown concentration(s) o the analyte(s) o interest.

, 3x

, 4x

, and 5x

). As with all techniques, a primary

concern is in making an accurate spiked addition. For ICP, an additional concern is drif. Te objective is to make anaccurate measurement. Rather than making multiple spiked additions where drif is given more ground to introduce error,it is suggested that the analyst measure the sample along with a single spiked sample several times to account or drif. A

reasonable measurement sequence would be:

blank -> sample -> blank -> spiked sample -> blank -> sample -> blank -> spiked sample -> blank -> sample -> blank

where an average o all measurements is taken or the nal calculation. Te above analysis sequence assumes linear drif thatshould be conrmed beore acceptance o the data. relative error. I larger spiking aliquots are required then an equal volume o 18 MO water should be added to the unspikedsample portion to cancel out volume dilution errors. with x,y coordinates o 0,0 as ollows:

(1) YI= mx

I= intensity o the sample, m = slope, and x

= concentration o the unknown analyte.

(2) Yk= m(x

+ x

s) = mx

+ mx

s

s= the concentration contribution rom the spike addition to the analyte concentration and Y

k= intensity or the

spiked sample.

concentration o zero. It is thereore necessary that the signal intensities be background corrected.

Subtract the intensity o the spiked rom the unspiked sample solution and divide this by the concentration o the analytespike to calculate the slope (m)

Yk- Y

I= mx

+ mx

s- mx

= mx

s

(Yk- Y

I) /x

s= m

Substitute the value or m into equation (1) along with the intensity (YI) to calculate the unknown analyte concentration (x

)


27/44


Te technique o standard additions offers the best possible solution to matrix intererence through plasma related effects.Te technique it requires an accurate background correction o the analytical signal intensities and does not account or matrices, it is possible to have severe spectral and background correction problems. It is cautioned here that at least twospectral lines should be used and the spectral region careully scanned and studied.

Internal Standardization

Te calibration curve technique is the most popular calibration technique. I the sample matrices are known and consistentthen matrix matching the calibration standards to the samples is an excellent option. Even when matrix matching is an option,many analysts still use an internal standard. It is suggested that the analyst consider the ollowing questions beore using aninternal standard:

8. I your plasma temperature were to go up or down, is the IS likely to ollow the same pattern o intensity change as the difficult [at best] to nd or each analyte while avoiding other issues listed above).

As discussed in the last part o this series, the matrix can inuence the plasma as well as the nebulizer. Internalstandardization is very effective in correcting or nebulizer related effects and may be effective or correcting plasma relatedeffects. It is obviously important that the matrix effect inuence both the internal standard to the same extent as the analyte.

Tis should be the case or nebulizer related effects but it may not be so or plasma related effects where the matrix inuence isrelated to the excitation potential o the emission line (as discussed in Part 10). It may be difficult to nd an internal standardthat has a similar excitation potential as the analyte in measurements where several analytes are involved. Te analyst isadvised to conrm that the matrix inuences the internal standard and analyte signal intensities proportionately.

Isotope Dilution Mass Spectrometry

As discussed in part 10, ICP-MS suffers orm matrix related effects upon the nebulizer and the signal intensity (quenching).In addition, even slight deposition on the sampler cone will cause drifing. Due in part to drifing, analysts have chosen to usethe calibration curve technique with internal standardization over the technique o standard additions. Although the standardadditions technique should work well in theory, the drifing associated with ICP-MS is too pronounced. Te use o a ratiotechnique such as internal standardization is a reasonable compromise with the understanding that the internal standardis not inuenced to exactly the same degree as the analyte signal. Tis is due to mass dependence. Te internal standards

commonly used are only used over relatively narrow mass ranges making the use o multiple internal standard elementsrequired or broad mass range applications. Te most common internal standard elements listed rom low to high mass are 6Li(isotope 6 enriched), Sc, Y, In, b and Bi.

ICP-MS has the unique capability o using an enriched isotope o the element o interest as the internal standard. Tistechnique, which is known as isotope dilution mass spectrometry (IDMS), has been known or nearly 50 years1. IDMS ismade possible through the availability o enriched stable isotopes o most o the elements rom the electromagnetic separatorsin Oak Ridge, ennessee (U.S.A). IDMS is thereore not applicable to monoisotopic elements.

Te IDMS technique involves the addition o a known amount o an enriched isotope o the element o interest to the sample.Tis addition is made prior to sample preparation during which the spiked addition o the enhanced isotope is equilibratedwith the sample. By measuring the isotope ratio o the sample and sample + spike isotope addition and knowing the isotopicratio o the enhanced addition, the sample concentration can be calculated. Te entire measurement is based upon ratiomeasurements o one isotope o the element to another. Drif, quenching and other related matrix effects do not present

an intererence with IDMS. Tis technique is considered a denitive2method and is well suited and established or thecertication o certied reerence materials.

IDMS is ree rom matrix effects (physical intererence) but it is not intererence-ree in that mass intererence must still bedealt with (isobaric, MO+, M++, etc.) in addition to correction o the signal intensity or detector dead time and mass biasintererence.

o view and example o an IDMS method, reerence EPA Method 6800*


28/44

*Visit inorganicventures.com/tech/icp-operations/or additional inormation rom this link

26

1. Hintenberger, H, Electromagnetically Enriched Isotopes and Mass Spectrometry, Proceedings Conference, Harwell, (1955): pg177; Butterworths Scientic Publications, London.

2. Denitive is dened as, A method of exceptional scientic status, which is suffi ciently accurate to stand alone in thedetermination of a given property for the Certication of a Reference Material. Such a method must have a rm theoretical

foundation so that systematic error is negligible relative to the intended use. Analyte masses (amounts) or concentrations must bemeasured directly in terms of the base units of measurements, or indirectly related through sound theoretical equations. Denitivemethods, together with Certied Reference Materials, are primary means for transferring accuracy -- i.e., establishing traceability.

raceability is dened as, Te property of a result or measurement whereby it can be related to appropriate standards, generallyinternational or national standards, through an unbroken chain of comparisons.

Common Problems with Hg, Au, Si, Os and Na12roblem ElementsP

Tis part o our ICP Operations guide provides some suggestions when attempting to work with mercury, gold, silicon,osmium, or sodium.

Mercury (Hg)

In March o 2003, the EPA published a bulletin describing the use o Au to stabilize Hg solutions: Mercury Preservationechniques. When working at the ppb level we have ound that using HCl rather than nitric acid will maintain the stabilityo Hg+2 solutions in plastic (LDPE) containers.

Te stability o mercury-containing solutions has been a topic o concern or all trace analysts perorming Hg determinations.Our in-house stability studies have yielded the ollowing conclusions.

Mercury Stability

1. Hg is stable in glass (only borosilicate glass studied) in 5% nitric acid at room temperature at all concentrationsstudied (0.05 to 1000 g/mL) or 1 year.

2. Hg is stable in glass (only borosilicate glass studied) in 5% nitric acid at 4C at 0.05 g/mL or 14 months. 3. Hg is stable in glass (only borosilicate glass studied) in 5% nitric acid at room temperature at 5 g/mL or 2 years

and 8 months.4. Hg appears to be stable in 10% v/v HCI in LDPE. A detailed stability study is in progress or 10% HCI Hg

containing solutions. 5. Hg is stable in LDPE in a water /5% absolute nitric acid matrix or at least 5 months.

Mercury Instability

1. Hg is not stable in MEBs containing Sb at the artrate. 2. Hg standards at 0.1, 1.0, 5, 10 and 100 g/mL were studied in LDPE and it was ound that Hg is lost. Te loss at the

100 g/mL is relatively small.3. Hg looses up to 1 g/mL Hg in LDPE over time. Tereore, Hg standards < 100 g/mL should be packaged in

borosilicate glass with a 5% nitric acid matrix. 4. Te most dramatic result o Hg loss: In a comparison o 5 g/mL Hg standards in 5% nitric acid stored at room temperature in glass and LDPE over

a period o ~2.5 years, it was ound that the glass was stable. Te LDPE container lost greater than 99% o the Hgindicating amounts o Hg > 1 g/mL can be lost in LDPE with time.

Another problem with Hg is loss during sample preparation. When perorming acid digestions, the use o closed vessel digestionor the use o condensers should be considered. Ashing should be avoided. Only use validated sample preparation procedures.

ICP_Booklet indd 26 2/9/11 4:58 PMCP_Booklet indd 26 2/9/11 4:58 PM


29/44

*Visit inorganicventures.com/tech/icp-operations/or additional inormation rom this link

27

Here are some additional suggestions when working with mercury:

Te presence o reducing agents in the solution may reduce Hg to the metal causing alse high results due to the volatility o

the element where the introduction system delivers more Hg to the plasma as a result. Te use o plastic introduction systems will cause unusually long washout times. Glass is preerred and the use o HCl ratherthan nitric acid will reduce the washout time. Te use o nitric acid matrices or ppb Hg determinations by ICP-MS should only be attempted using Au as a stabilizingagent (see above link).

Hg elemental data*

Gold (Au)

Te chemical stability o Au is very similar to that o Hg. Te ollowing suggestions may be helpul:

Nitric acid solutions o Au at the low ppm and ppb levels are not stable. Use HCl matrices.

Do no use Pt crucibles when ashing samples containing Au. Au will alloy with the Pt. When measuring Au in the presence o signicantly greater amounts o Pt using ICP-MS, be aware o the resolvingcapability o your instrument.

Au elemental data*

Silicon (Si)

Te ollowing suggestions are advised when working with silicon:

Si is a common contaminant. In addition to the obvious use o laboratory glassware, common sources o contaminationinclude silicon oil/grease, plastics containing catalyst residue, and air particulates. Si0is easily dissolved using an equal mixture o HF:HNO3:H2O. SiO2 is readily soluble in either HF or NaOH. Regardless othe mode o dissolution, solutions should be stored in plastics known to contain no catalyst residues or that have been leachedwith dilute HF or 48 hours. Exercise caution when heating solutions containing Si and HF. Si may be lost as the volatile H

2SiF

6when heated. When water

is present H2SiF

6will not orm. I you wish to remove Si rom the sample then add suluric acid and heat in a Pt crucible.

Silicon dioxide is soluble in caustic media. When acidied it is stable at low ppm levels but will slowly polymerize andprecipitate out o solution. Common preparations involve sodium carbonate usions in Pt crucibles and dissolution o theuseate with HCl - make sure the ppm level o Si upon dilution is low ppm and the solution is not allowed to sit or extendedperiods. HF (even low ppm levels o HF) containing samples should not be put through glass or quartz in

Date post:	03-Jun-2018
Category:	Documents
Upload:	jamesbarnesiv
View:	227 times
Download:	0 times

ICP Operations Guide

Documents