Astm stp 1468 2005

The ASTM International D02 Committee on Petroleum Products and Lubricants through its Subcommittee 3 on Elemental Analysis has played a large and crucial role in the last several decades

Elemental Analysis of Fuels and Lubricants: Recent Advances

and Future Prospects

R A Kishore Nadkarni, editor

ASTM Stock Number: STP 1468

INTERNATIONAL

ASTM International

100 Barr Harbor Drive

PO Box C700 West Conshohocken, PA 19428-2959 Printed in the U.S.A

Kishore Nadkarni, editor

p c m ~ S T P ; 1468)

Includes bibliographical references and index

ISBN 0-8031-3494-0 (alk paper)

1 Fuel Analysis 2 Lubrication and lubricants Analysis I Nadkami, R.A II

Series: ASTM special technical publication; 1468

Peer Review Policy

Each paper published in this volume was evaluated by two peer reviewers and at least one edi- tor The authors addressed all of the reviewers' comments to the satisfaction of both the technical editor(s) and the ASTM Committee on Publications

To make technical information available as quickly as possible, the peer-reviewed papers in this publication were prepared camera-ready as submitted by the authors

The quality of the papers in this publication reflects not only the obvious efforts of the authors and the technical editor(s), but also the work of the peer reviewers In keeping with long-standing publication practices, ASTM maintains the anonymity of the peer reviewers The ASTM Committee

on Publications acknowledges with appreciation their dedication and contribution of time and effort

on behalf of ASTM

Printed in Baltimore, MD September 2005

This publication, Elemental Analysis of Fuels and Lubricants: Recent Advances and Future Prospects, contains selected papers presented at the symposium of the same name held in Tampa, Florida, on 6-8 December 2004 The symposium was sponsored by Committee D02

on Petroleum Products and Lubricants The symposium chairman and editor was R A Kishore Nadkarni

Analysis of Gasoline and Diesel Fuel Samples by Inductively Coupled Plasma

Atomic Emission Spectrometry (ICP-AES), Using Pneumatic Nebulizer

and Standard Spray Chamber -c c ONYESO

Elemental Analysis of Lubricating Grease by Inductively Coupled Plasma

Atomic Emission Spectrometry (ICP-AES) B s FOX

The Use of Microwave Digestion and I C P to Determine Elements in Petroleum Samples J D HWANG, M HORTON, AND D LEONG

Advances in ICP-MS Technologies for Characterization and Ultra-Trace

Speciation as a Tool for the Petroleum Industry J PASZEK, K J MASON,

A, S MENNITO, AND F C MCELROY

Direct Trace and Ultra-Trace Metals Determination in Crude Oil and

Fractions by Inductively Coupled Plasma Mass Spectrometry

S DREYFUS C PECHEYRAN, C MAONIER, A PRINZHOFER C P LIENEMANN AND

O F X DONARD

Fuel Analysis by Filter Furnace Electrothermal Atomic Absorption

Spectrometry P TII~rARELLL M PRIOLA S RICCHIUTO, D A KATSKOV, AND

P NGOBENI

Rotrode Filter Spectroscopy: A Recently Improved Method to Detect and

Analyze Large Wear and Contaminant Particles in Fluids M LUKAS

R J YURKO AND D P ANDERSON

Analysis of Fuels, Lubricants, and Greases Using X-Ray Fluorescence

Spectrometry J WOLSKA, B VREBOS, AND P BROUWER

Determination of Sulfur Content in Crude Oil Using On-Line X-Ray

Transmission Technology s FESS

Low-Level Sulfur in Fuel Determination Using Monochromatic W D X R F -

ASTM D 7039-04 z w CHEN, F WEI, I RADLEY, AND B BEUMER

Latest Improvements on Using Polarized X-Ray Excitation EDXRF for the

Analysis of Low Sulfur Content in Automotive Fuel D WISSMANN

Rapid Determination of Sulfur in Liquid Hydrocarbons for At-Line Process

Applications Using Combustion/Oxidation and UV-Fluorescence

Detection s TARKANIC AND J CRNKO

Pyro-Electrochemicai On-Line Ultra Low Sulfur Analyzer J R RHODES

DP-SCD and LTMGC for Determination of Low Sulfur Levels in

Hydrocarbons R L GRAS, J C LUONG R V MUSTACICH, AND R L SHEARER

Sampling and Analysis of Mercury in C r u d e O i l - - s M WILHELM

D A KIRCHGESSNER L LIANG, AND P H KARiHER

Determination of Total Mercury in Crude Oil by Combustion Cold Vapor

Atomic Absorption Spectrometry (CVAAS) B s Fox K J MASON AND

F C MCELROY

Mercury Measurements in Fossil Fuels, Particularly P e t r o c h e m i c a l s - -

P B STOCKWELL, W T CORNS, AND D W BRYCE

181

196

207

O T H E R H E T E R O A T O M S

Recent Advances in Gas C h r o m a t o g r a p h i c / A t o m i c Emission Hetero-Atom

Selective Detection for Characterization of Petroleum Streams and

e r o d u c t s - - F , p DISANZO AND J W DIEHL

Improvements in the Determination of Fluorine in Fuel and Lubricants by

Oxidative Combustion and Ion-Selective Electrode D e t e c t i o n - - L J NASH Phosphorus Additive Chemistry and its Effects on the Phosphorus Volatility of Engine Oils T w SELBY, R J BOSCH AND D C FEE

Analysis of the Volatiles Generated During the Selby-Noack Test by 31p NMR

Spectroscopy R J BOSCH, D C FEE, AND T W SELBY

In spite of being a mature science, elemental analysis continues to play a vital role in product manufacturing and quality characterization in many sectors of all industries Re- search divisions in both industry and academia continue devising new ways of lowering the elemental detection limits so that even the minutest amounts of elements in products could

be determined in as accurate and precise a fashion as possible

The ASTM International D02 Committee on Petroleum Products and Lubricants through its Subcommittee 3 on Elemental Analysis has played a large and crucial role in the last several decades in standardizing numerous elemental analysis methods used in the oil in- dustry Currently there are about 75 standard test methods under the jurisdiction of SC 3, and additionally at least 6 more are under active development and moving towards standard designations I have no doubt that this activity will continue in the future These standards comprise virtually all known modem techniques for elemental analysis of petroleum products and lubricants

The first ASTM D02 symposium on this subject was held in New Orleans in December

1989 at which 20 papers were presented Of these, 13 were published as a book, Modern Instrumental Methods of Elemental Analysis of Petroleum Products and Lubricants, ASTM STP 1109 The current and second "quindecennial" (i.e., every 15 years) was held in Tampa, Florida in December 2004 This was attended by over 120 people Thirty papers were pre- sented on diverse subjects from 64 authors from nine different countries: Brazil, France, Germany, Italy, the Netherlands, South Africa, Switzerland, U.K., and U.S Of these, 12 papers were from the oil industry, 15 from the instrument manufacturers, l0 from national research organizations, and 4 from the universities

The objective of this symposium and this book is to acquaint the readers with the latest advances in the field of elemental analysis and to focus on what avenues of future research

to explore in this area The subjects included are various elemental analysis techniques such

as atomic absorption spectrometry, inductively coupled plasma emission and mass spectrom- etry, isotope dilution mass spectrometry, X-ray fluorescence, ion chromatography, gas chro- matography-atomic emission detection, other hyphenated techniques, hetero-atom micro- analysis, sample preparation, reference materials, and other subjects related to matrices such

as petroleum products, lubricating oils and additives, crude oils, used oils, catalysts, etc

Of the 30 papers presented at the symposium, 23 papers were published in the Journal of ASTM International (JAI), and are included in this ASTM publication As far as possible,

the papers have been arranged by analytical techniques used, although in some cases there

is some overlap: ICP-AES, XRF, sulfur, mercury, other hetero-atoms

The first article is from the plenary lecture given at the symposium by the symposium chairman Kishore Nadkarni It covers total quality management practices advocated for ob- taining a "perfect" analysis Proper staff training, sampling, calibration and quality control practices, adherence to test method details, participation in proficiency testing, accreditation from national bodies, benchmarking, etc., are some of the critically important approaches that need to be taken to achieve the ideal state of analytical Zen perfection

Atomic Spectroscopy

Among the seven atomic spectroscopy papers in this book, five concern various aspects

of ICP-AES, a technique widely used for the determination of metals in petroleum products

vii

and lubricants Onyeso (Ethyl Corporation) presents an ICPAES method for the determina-

tion of additive elements and wear metals, principally manganese, in gasoline and diesel fuels, with simple dissolution in kerosene and using yttrium internal standard Accessories such as direct injection nebulizer, ultrasonic nebulizer, chilled spray chamber, etc., were not necessary for this analysis

Fox (ExxonMobil Research and Engineering) presents an ICPAES method for the deter-

mination of additive elements and wear metals in lubricating greases Since such samples cannot be directly nebulized in the ICP plasma, alternate sample dissolution techniques were employed: dry sulfated ashing, microwave assisted dry ashing, microwave assisted acid di- gestion with both open and closed vessels This method is being developed into an ASTM standard test method and is expected to be published by YE05 Hwang and Leong (ChevronTexaco) also discuss the use of microwave acid digestion for sample preparation

before ICPAES measurements

Elemental speciation using mass spectrometry in conjunction with ICPAES is a latest advance in atomic spectroscopy, which is becoming popular in analytical research labs

Mason et al (ExxonMobil Research and Engineering) show how linking ICP-MS to various

liquid chromatographic techniques has enabled determination of ppm levels of metals in hydrocarbons to ppb level measurements in refinery effluent streams Hyphenated ICP-MS techniques were used to provide speciation information on nickel and vanadium in crude oils and assist in development of bioremediation options for selenium removal in wastewater treatment plants Similar ICP-MS technique without sample demineralization was used by

Lienemann, et al (lnstitut Francais du Petrole) to determine the trace and ultra-trace amounts

of metals in crude oils and fractions

Lukas et al (Spectro Inc.) describe an improvement made in rotating disc electrode atomic

emission technology by incorporating a filter device in the rotrode, which enables to detect particles greater than 10 i~m size

Tittarelli et al (SSC, Milan) employed a transverse heated filter atomizer with atomic

absorption spectrometry to determine a number of trace elements in automotive and jet fuels Sub-ppm detection limits were obtained The use of filter furnace reduces the risk of ele- mental loss during drying and pyrolysis steps, and decreases the interferences due to molec- ular absorption and light scattering

X-Ray Spectroscopy

Similar to atomic emission spectroscopy, equally widely used technique for elemental analysis in the oil industry is X-ray fluorescence (XRF) There are four papers in this book using this technique, three of which deal with the determination of sulfur in gasoline and diesel

Wolska et al (Panalytical BV) compared performance of three XRF technologies: high

power and low power WDXRFs and a bench top EDXRE There are large differences in the sensitivities and hence varying lower limits of detection or qualification and sample through- put, for these technologies

trace amounts of sulfur in fuels of the future as evident from seven papers on this subject published in this book

Nadkarni (Symposium Chairman) reviewed the alternate methods available for sulfur de- termination in fuels Out of about 20 ASTM standard test methods available, only about five (D 2622 WDXRF, D 3120 microcoulometry, D 5453 UV-fluorescence, D 6920 pyro- electrochemical, D 7039 MWDXRF) are appropriate for ultratrace amounts of sulfur in gasoline or diesel However, in their actual industrial use only D 2622 and D 5453 predominate

Chen et al (XOS Inc.) describe a newly developed technology instrument based on mon- ochromatic WDXRF for low sulfur analysis of fuels The instrument has a significant ad- vantage over existing WDXRF instruments in terms of increased sensitivity and improved signal to noise ratio This technique has been recently given the ASTM designation D 7039 Another new instrument recently developed for sulfur by XRF determination is described

by Wissmann (Spectro, Inc.) This method uses polarized EDXRF, considerably reducing background scatter, and achieving detection limit comparable to that of WDXRF Recent developments in detector technology and in closed coupled static geometry have resulted in further improvement of sensitivity for this application This method is also in the develop- mental stage for ASTM method designation

Shearer et al (Ionic Instruments and Dow Chemicals) describe a novel technique devel- oped tbr low levels of sulfur in hydrocarbon matrices using a low thermal mass temperature programmable and dual plasma chemiluminiscence detector The method with appropriate modification can measure individual sulfur species similar to ASTM method D 5623

On-line Sulfur Analysis

Increasingly refineries, plants, and pipeline operators are focusing on obtaining quick turn- around for sulfur analysis rather than wait ['or time-delayed laboratory analysis A large number of such installations are being operated in the industry around the world Three papers in this book discuss applications of such on-line technology for sulfur determination

in fuels In an on-line application of X-ray transmission technology, Fess (Spectro, Inc.)

describes the basis of this technology and its application to classification and blending of crude oils that contain between 0.1 and 3.3 m % sulfur Commercial instruments based on this technology are being used in the field

In a second on-line application paper, Tarkanic and Crnko (Antek/PAC) describe an on- line instrument based on ASTM Test Method D 5453, UV-Fluorescence Detection The latter

is a widely used method in the oil industry for low and ultra-low levels of sulfur The on- line instrument appears to be very stable and fast (< 1 min per analysis) over extensive periods of field operations

In a third on-line application paper tbr sulfur analysis, Rhodes (Rhodes Consulting), ASTM Test Method D 6920 is applied for on-line application This method uses pyro-combustion followed by electrochemical detection

Mercury Determination

Although adverse effects of mercury emissions on environment and humans has been known lbr decades, in recent years there has been concern regarding the mercury content of crude oils, and its emission through petroleum refining process There are three articles in this book discussing this issue

Wilhelm et al (Mercury Technology Services/EPA et al.) provide a review of the presence

of mercury in various parts of the world, its speciation, and alternate methods of determining

low ppm and sub-ppm levels Fox et al (ExxonMobil Research and Engineering) describe

a method for the determination of ppb levels of mercury in crude oils and distillation cuts

using combustion cold vapor atomic absorption spectrometry technique Stockwell et al (PS Analytical Ltd.) describe the technique of atomic fluorescence spectrometry for the deter-

mination of mercury both before and after mercury removal from petrochemicals The tech- nique has been used for on-line measurements in installations operating around the clock for

at least 2 years

Other Heteroatoms

DiSanzo and Diehl (ExxonMobil Research and Engineering) used GC-AED for the de-

termination of elements such as carbon, nitrogen, sulfur, oxygen, and phosphorus in fuels and petroleum fractions A simplified version of comprehensive GC x GC is coupled with atomic emission detector to reduce the hydrocarbon matrix interference using simple and rugged modulation along with rugged wide bore capillary columns The technique together with other spectroscopic techniques such as GC-MS can provide information on many se- lected elements and compounds that may be present in fuels as additives or contaminants

In a pair of papers, Selby et al (Savant, hzc and Astaris LLC) describe using phosphorus

as an indicator of volatility of engine oils Phosphorus is volatilized during Noack volatility test (ASTM D 5800) The volatile material is trapped and analyzed for total phosphorus using ICP-AES, and for phosphorus species using 3~p NMR spectroscopy

An oxidative combustion followed by ion selective electrode detection method is proposed

by Nash (Antek/PAC) for the determination of fluorine in fuels and lubricants An ASTM

method based on this technique is in development stage

Unpublished Symposium Papers

Some papers were presented at the Symposium; however, they were not submitted for publication by the authors Nevertheless, they represent interesting approaches to some spe- cific elemental analysis issues in the petrochemical industry It would be useful if the authors eventually publish these articles for the benefit of others in the industry These presentations include the following:

I Kelly et al (NIST) describe an isotope dilution thermal ionization mass spectrometry

method for the determination of sulfur in fossil fuels The method is being used in NIST for certification of a number of liquid fuels at low sulfur concentration levels

2 Kelly et al (NIST) also describe a "designer" calibration standard method for sulfur

determination in fossil fuels for users to prepare NIST traceable working standards with known concentrations and uncertainties

3 Manahan and Chassaniol ( Cosa Instruments and Dionex) describe an oxidative com-

bustion followed by ion chromatographic conductometric method for the determination

of a number of nonmetallic elements such as sulfur and halogens in liquid and gaseous hydrocarbons A standard based on this technique is under development in ASTM for designation as a standard method

4 Long et al (NIST) describe another method for mercury determination in crude oils

using isotope dilution-cold vapor-inductively coupled plasma-mass spectrometry tech- nique The method has very high sensitivity, very low blank and high accuracy The technique is being used to determine mercury in a large number of crude oil samples from Department of Energy strategic petroleum reserve in the mercury concentration range of 0.02-10 ng/g

5 Finally, Mason et al (ExxonMobil Research and Engineering) describe the approaches

used for assay of fresh and spent reformer catalysts to determine the precious metals (platinum and rhenium) in them Methods such as WDXRF, ICPAES, and classical wet chemistry methods are used for such analysis Precise and accurate methods are critical for these analyses, since small errors in analysis can have a large impact in commercial transactions of these catalysts between the catalyst vendors and the oil companies Hopefully, the papers included here will provide the readers with the current state-of-the- art and future research trends in the field of elemental analysis in the oil industry Most modern techniques used in the field are represented here

Acknowledgment

I want to thank various ASTM staff members (particularly David Bradley, Dorothy Fitz- patrick, Crystal Kemp, Hannah Sparks, and Roberta Storer) for their prompt response and cooperation that made the symposium and subsequent efficient publication in JAI and of this volume possible My thanks are also due to the reviewers who did a very good job of providing technical reviews of all original paper submissions Their invaluable assistance in reviewing the papers made the final publication a much better quality product

R A Kishore Nadkarni

Chairman, D02.SC 3 and Symposium Chairman

R A K i s h o r e Nadkarni 1

Journal of ASTM International, March 2005, Vol 2, No 3

Paper ID JAI12964 Available online at www.astm.org

Zen and the Art (or is it Science) of a Perfect Analysis

ABSTRACT: An analytical laboratory in any industry plays a crucial role in product quality management and ultimate customer satisfaction Some factors need to be considered for an aspiring laboratory to become a perfect performer These range from sampling, calibration, contamination control, and use of valid test methods to statistical quality assurance Some approaches may be utilized to achieve

a perfect analysis including: staff training, participation in proficiency testing, use of standard reference materials in the analytical sequence, internal and external audits, agency accreditation, continuous improvement program, benchmarking, etc Laboratories managed in this way show demonstrated superiority in data precision and accuracy over the labs which do not practice such quality management Well-managed industrial laboratories can have insignificant laboratory sigma compared with manufacturing variability in the plant production For a flawless perfect analysis, determination to excel, mental discipline to stay the course, willingness to overcome inertia and resistance, and focus on producing a perfect analysis at all levels of laboratory staff are essential

KEYWORDS: analysis, quality management, perfect analysis

As Robert Pirsig wrote in his landmark iconic autobiographical novel "Zen and the Art of Motorcycle Maintenance," the art o f motorcycle maintenance is primarily a mental phenomenon [1] One m a y have the tools, but unless there is mental preparation to achieve high goals, the tools alone will not help A similar mindset is needed to achieve excellence in a laboratory to make it into a perfect laboratory that produces flawless performance Tools may be available, but

i f there is no organizational passion and will to excel, the laboratory will not become a perfect laboratory

The culture of excellence must be pervasive throughout the laboratory organization from the laboratory manager to the laboratory technician Higher management especially needs to show through visible actions that only the best will do Perfection cannot be achieved through short- term stop-gap measures A long-term improvement plan must be in place and followed upon to

be effective

What is a perfect laboratory? It is a laboratory which delivers the product (i.e., accurate and precise data) on time; if necessary, continuously improves on itself; makes the analysis "Right the First Time," thus eliminating repeat analysis and giving erroneous information to the customers; communicates with its customers and sometimes educates them when necessary This laboratory cares about the success o f its customers' business

Customer Services

A laboratory is a microcosm of its parent organization The product delivered from a laboratory is quality data Hence, the primary objective of a laboratory should be to be the best in quality A laboratory needs to deliver a consistent product on time which meets or exceeds customers' expectations, and which increases customers' confidence in the laboratory's

Manuscript received 7 September 2004; accepted for publication 19 October 2004; published March 2005

i Millennium Analytics, Inc., East Brunswick, NJ 08816

reliability and dedication to quality A customer needs to know that the laboratory cares "Total Care" is the sum of impressions formed during contact with the customers A perception by the customer that the laboratory is a caring organization can convert a customer from being forced to

be a customer to becoming a customer by choice Substantial or continuing violations of a customer's justified expectations will cause the customer to feel that the laboratory, organization simply does not care

Pillars to Build a Perfect Laboratory

There are at least twelve components which help to produce flawless laboratory performance

In the approximate order in which an analysis is performed, these include but may not be conEmed to the following (See Fig 1 on page 3):

1 Training

2 Representative Sampling and Contamination Control

3 Calibrations

4 Technical Details of Test Methods

5 Statistical Quality Assurance

6 Use o f Certified Reference Materials

to identify ways to improve the methodology, to obtain better precision and accuracy, and to improve the turnaround time

Some of the areas in which a laboratory staff member must be fully trained include safety, data security, laboratory instrumentation, test methods used, calibration protocols, statistical quality assurance, use o f certified reference materials, long term analytical needs and goals, and ethical behavior In today's culture, it is still up to the supervisors and management to filter the Zen attitude down to the working level people The very fundamental first step toward obtaining

a perfect analysis is through staff training Dr Derek Bok, former president of Harvard University once commented, "If you think education is expensive, try ignorance."

Sampling

Obviously the first critical step in any analytical sequence is the integrity and validity of a sample being analyzed More often than not this sampling step does not involve actual laboratory

staff; usually the chain of custody for a sample starts with the receipt of the sample in the laboratory Once the laboratory acquires the sample, however, it is the laboratory's responsibility

to have a system for unique identification o f each sample, sample handling, storage and retention procedures, as well as safe disposal procedures Identification of the population from which the sample is to be obtained, selection and withdrawal o f valid gross samples of this population, and reduction of each gross sample to a laboratory sample suitable for the analytical technique to be used are some o f the key steps to be considered in obtaining a representative sample for analysis [2] Equally important is documented chain o f custody procedures to authenticate and maintain the sample integrity Several ASTM standards deal with sampling aspects for the analysis of petroleum products and lubricants:

* D 4057: Manual Sampling of Petroleum and Petroleum Products

9 D 4177: Automatic Sampling o f Petroleum and Petroleum Products

9 D 4840: Sampling Chain o f Custody Procedures

o D 5842: Sampling and Handling o f Fuels for Volatility Measurements

9 D 5854: Mixing and Handling of Liquid Samples of Petroleum and Petroleum Products Additionally, some ASTM standards give instructions for specific sampling requirements for specific analytical tests Attention must be paid to these caveats to obtain reliable test results

Contamination Control

Gross contamination of the sample in any analysis and in particular for trace analysis is a serious problem which, if unchecked, will completely negate the validity of the analytical results The problem can become particularly insidious as one is working in the range o f ppm and sub- ppm levels o f analytes Contamination from particulates in the air, impurities in reagents, trace elements from the sample containers as well as glass- or plastic-ware used during analyses are all potential sources of contamination [3] An accompanying "blank" sample used throughout the analysis sequence may or may not accurately measure the extent of contamination, since such contamination from air or glassware, etc may not be uniformly present when in contact with the blank and a real sample The point is that both a blank determination and a rigorous protocol for contamination control in the laboratory are essential for obtaining perfect results, particularly in the area of trace analysis An excellent source book for discussion on contamination control is given in [4]

Calibration or Verification

Virtually all analytical test methods require some form of calibration or verification before actual samples are analyzed Different test methods require different calibration intervals Thus,

a decision about appropriate calibration frequency must be made on a ease by case basis There

is a tendency among many laboratories to do the bare minimum calibrations similar to their approach toward quality control requirements This is not the way to achieve superior performance Moreover, if an instrument is out-of-calibration, under no circumstances can data from that instrument be reported to the customers

Appropriate calibration standards must be utilized during analysis A wide variety of such standards are available from commercial sources, NIST, etc Many laboratories have capabilities

o f preparing reliable in-house standards Calibration standards identical to the samples being analyzed would be ideal, but failing that, at least some type of standards must be used to validate the analytical sequence In physical measurements this is usually achievable, but it is often difficult or sometimes almost impossible in chemical measurements Even the effects o f small deviations from matrix match and analyte concentration level may need to be considered and evaluated on the basis o f theoretical or experimental evidence Sometimes the use of standard additions technique to calibrate the measurement system is a possibility But because an artificially added analyte may not necessarily respond in the same manner as a naturally occurring analyte, this approach may not be always valid, particularly in speciation work9

An important aspect of calibration is the decision on calibration intervals, i.e., the maximum

period between successive recalibrations Two basic and opposing considerations are involved: the risk of being out o f tolerance at any time o f use and the cost in time and effort The former should be the major concern because o f the dilemma o f what to do with the data obtained during the interval between the last known in and the first known out o f calibration However, an overly conservative approach could be prohibitively expensive A realistic schedule should reduce the risk o f the former without undue cost and disruption to work schedules The factors that need to

be considered in a realistic schedule include:

9 Accuracy requirement for the measured data

9 Level of risk involved

9 Experience of the laboratory in use o f the equipment or methodology

9 Experience of the measurement community

9 Manufacturer's recommendations

9 External requirements for acceptability o f data

9 Cost o f calibration and quality control

Quality control measurements can help a great deal in deciding what calibration frequency intervals should be used

Test Method Details

A laboratory must have fully documented test methods that are used for analyses, and the staff members must be competent in the details in each test method that they will be using for analyses Experience has shown that a major source o f analytical error is deviation from the prescribed standard test method, whether intentional or inadvertent Most o f the time the details given in a standard test method are there for a purpose A laboratory wishing to deviate from the standard test method must document the deviation and show that the modified version produces statistically equivalent or (preferably) better results in terms o f precision and accuracy

In most laboratories, staff members periodically and sometimes frequently change; hence, it

is important to have a system in place for periodic checking that the laboratory practice is indeed

in conformance with the test method requirements9

Statistical Quality Control and Assurance

One cannot control what one cannot measure, and one cannot improve what one cannot control Every measurement system is beset with variation and noise, and the only way to control

and reduce variation is by identifying its cause, establishing its extent, and interpreting what it is indicating Variability arises because no two things are absolutely alike either in nature or in laboratory measurements All one can try to do is to minimize it as much as possible The primary step in measuring variance is the use o f statistical quality control (SQC) or assurance (SQA) charts Quality assurance must be viewed as an integral part of a complete analytical sequence and not as an added burden with additional costs

Calibration(" ) Sample Analysis< -) SQA The only way to prove data integrity and reliability is through SQA The frequency of quality control (QC) analysis will depend on the type o f analysis, the instrument involved, stability o f the measurement process, importance o f decisions based on the test results, and magnitude o f changes between batches The same considerations described in the calibration section above need to be involved in making the decision on SQA frequency But at least one QC standard must be analyzed with each set of samples to be analyzed Preferably one QC standard each should be analyzed bracketing a series of sample analyses Ideally, a few QC standards should be interspersed with samples to continuously monitor the data quality Indeed there are some ASTM test methods where such protocol is mandatory, e.g., D 4951 and D 5185 ICP-AES methods for metals analysis A general rule o f thumb used in many laboratories is one QC sample per 5-10 actual samples Although there are statistical protocols for reducing the QC frequency, this is overridden if a customer (and many do) requires QC accompanying their sample analyses If a

QC analysis indicates poor performance, immediate remedial action must be taken before continuing the analysis or reporting the data to the customers

data as well A chart based on statistical control limits must be plotted and interpreted on time A control chart by itself can only indicate that there is an upset and not the cause of the upset Laboratory staff must investigate, identify, and eliminate the cause(s) of the upset and bring the process back to a state o f statistical control I f a control chart is merely plotted and filed away without taking action when required it is simply a waste of time and effort

There are a number of statistical run rules which are used to indicate out-of-control situations Some laboratories use only one rule of a data point beyond 3 sigmas This in itself may not be adequate, since it will ignore several short term and more importantly long term indications o f system instability

Based on the QC data, monthly or quarterly laboratory capability should be calculated to compare the laboratory precision with that given in the standard test method where available One way of estimating lab capability is the the equation below:

Test Performance Index = Laboratory Standard Deviation / Test Method Repeatability Further discussion on the numerical criteria to be used for evaluating TPIs can be found in ASTM standard D 6792 Some laboratories prefer to compare their standard deviation against the test method reproducibility or to use an inverse calculation from that given above In either case

an individual best laboratory will consistently produce a superior precision compared to the industry average precision Such periodic performance feedback should be a key feature of any continuous improvement program

Such a continuous improvement (CI) program can be initiated at any level and may include obtaining a better instrument or changing to a better alternate test method Since the lab worker

staff is usually busy in day-to-day analyses, the initiative for such CI program usually needs to come from supervisory or managerial staff Such a program can continue indefinitely looking at various ways o f improving diverse areas of laboratory operations A project ends, but a process continues!

Use of Certified Reference Materials

Where available, reference materials need to be used for calibration or quality control The National Institute of Standards and Technology (NIST) and many commercial sources supply such materials Many high performing technology laboratories are capable of preparing their own reference materials using the same approach used by NIST in certifying their Standard Reference Materials (SRM) The calibration reference materials (RM) should be traceable to national standards, and the traceability must be preserved in the laboratory's documentation system

It is not a good idea to use the same material for both a calibration standard and quality control The latter should be similar in matrix to the type of samples being analyzed, although this is not always possible due to the lack of availability of suitable certified materials At least for QC it is easy enough to obtain a reasonably large quantity of plant product, analyze it by the tests of interest multiple times, and calculate the average value and the sigma limits to initiate the control chart

Documentation

Laboratory analytical reports must contain all information necessary for a customer to understand or interpret the results This may include the unique laboratory sample ID, equivalent customer sample identification, test methods used and any modifications done to them during analysis, actual analytical results, supporting QC results where available, signature of responsible authorized laboratory staff, etc Records of calibration or QC records must be maintained for an adequate period of time

Errors in Data Reporting Reporting blunders occur more frequently than one might think

We have often seen the laboratories taking part in the ASTM interlaboratory erosscheck programs reporting the data without paying adequate attention to the report formatting A perfect analysis may have been done, but if the results of this perfect analysis are reported incorrectly, nothing is gained We have repeatedly found that the numbers are often transposed, or wrong decimal places are used When we check back with these laboratories, almost always it turns out that there was indeed an error in reporting Some of the glaringly obvious examples that we have come across include:

9 The total metal content is <100 ppm, yet the ash (D 482) is reported as 3 m%

9 The ash (D 482) and sulfated ash (D 874) results are widely different by an order of magnitude

9 The kinematic viscosity (D 445) of a fuel sample is reported as 6300 cst @ 100~ that would make this a practically solid sample

9 One laboratory reported 1000~ as the flash point (D 93) of a fuel sample!

9 The pour point (D 97) o f a fuel oil sample was reported as 54~ that would make it a solid oil sample at room temperature

9 The sulfur content (D 2622) of a reformulated gasoline sample is reported as 94 %! This should have been 94 ppm

9 The water content (D 6304) of a lubricating oil sample is reported as 9 % This should have been 9 ppm

All of these errors could have been prevented easily by reviewing the data before reporting One would not send a defective product from a plant to a customer Why then send the defective data out? It takes only a few moments to go over the final report form to check for such errors; a laboratory's credibility would be much better for having done so

Internal and External Audits

Every laboratory striving for perfection needs to do periodic audits to check that its quality systems are still working properly Audits o f test methods should be conducted (perhaps annually) to confirm adherence to the documented test methods The performance of the entire test should be observed and checked against the official specified test method Many errors in results are derived from taking liberties with the test method's detailed requirements

It needs to be recognized that the ISO 9000 audits alone by outside registrars do not adequately cover the technical aspects of a laboratory performance It is imperative that internal audits be done by the laboratory staff since they are far more familiar with the required technical details of each test that the laboratory performs Through such thorough reviews, a laboratory can eliminate defects, improve on its strengths, and become a perfect laboratory satisfying customer expectations Findings and recommendations of internal or external audits must be promptly reviewed and acted upon, otherwise there is little point in spending efforts in doing audits

Proficiency Testing

A laboratory needs to participate in relevant crosschecks organized within the company circuit, with its customers if requested, or with the industry organizations A completely unbiased assessment o f a laboratory's capability for precision and accuracy can only be judged by a blind crosscheck If the laboratory results are equivalent to those of other laboratories in the industry, it assures the laboratory itself and its customers that the laboratory is producing reliable results On the other hand, if the results are less than satisfactory, it also helps the laboratory to correct its operation, which may be deficient in some aspect Participating in a crosscheck without following up on deficiencies if detected is a waste o f resources

One such program widely used in the oil industry is described by Nadkami and Bover [5] Participation in proficiency testing should not be considered as a substitute for in-house quality control and vice versa These are two independent and necessary activities for a well-managed laboratory

Laboratory Accreditation

Laboratory accreditation in the U.S is handled by more than 150 accrediting private and quasi-government bodies Thus, there is a bewildering array to choose from as to which is the best accreditation for a laboratory This will depend to some extent on the industry and the customers with which a laboratory does business The most well known program is ISO 9000

registration Often when a chemical plant or a refinery gets ISO 9000 registered, the testing laboratory associated with that site is also registered It must be recognized however that having

an ISO 9000 registration does not necessarily guarantee the best laboratory data unless supplemented by additional internal quality protocols [6,7]

Two other meaningful accreditation programs for the testing laboratories are by the American Association for Laboratory Accreditation (A2LA) [8] and ISO Guide 17025 [9] There

is a similar National Measurement Accreditation Service (NAMAS) in the United Kingdom and parts of Europe, which is administered by the U.K National Physical Laboratory [10] An advantage o f these three accreditation schemes is that they involve actual test method performance, laboratory quality system, and other issues vital to a credible laboratory's performance Given the widespread utilization of such accreditation schemes, many customers will forgo their own quality audits of contractor laboratories if the latter have such external certifications

Benchmarking

Benchmarking is the search for those best practices that will lead to a company's superior performance, and it incorporates these practices in one's own organization To become the best, one must learn from the best This learning should not be limited only to the organizations in one's industries; companies outside a specific industry may also have superior standards which can be utilized beneficially Ideas should be shamelessly stolen from the best, provided they are not company proprietary materials

Japanese people call benchmarking "Dontotsu," literally meaning "striving for the best of the best." Xerox Corporation was a pioneer in industrial benchmarking An excellent text on this subject has been published by Xerox [1 t] Some Baldrige National Quality Award winning companies incorporate benchmarking as an essential tool in their quality management process Perfect laboratories also need to look at their counterparts to fred out how to improve themselves

to the level o f the best laboratories

Many laboratories calculate the so-called Laboratory Productivity Index (LPI), defined as the total tests/total laboratory workers per month or per quarter, etc However, experience shows that these indices may vary widely from laboratory to laboratory It is not certain that LPI is a reliable measure for interlaboratory comparison, since:

a Different tests need different amounts of time to complete

b The same analysis may be done automatically by analyzers in some laboratories and manually in others

c The test requirements can be different depending on the types of products analyzed

d The clientele and the test slate can be different for research and production laboratories LPI may be useful to track laboratory productivity in one site but needs caution when comparing one laboratory against another

Ethics

One would think that it is a simple fact that the results obtained in the laboratory should be reported honestly; however, several incidents reported in last few years make one wonder

whether ethics should be a mandatory training course in colleges and laboratories for analytical chemists!

In April 1996, the U.S Department o f Justice t'med a major additives company $4.75 million for providing additive packages to the government falsely certifying that they met the military product specifications and passed specified testing

A Linden, NJ commercial laboratory was fined $1 million in 2001 for doctoring the laboratory reports that showed that they met EPA's Clean Air Act standards for cleaner burning fuels The company was also put on probation for three years, and its former president was convicted of fraud

Some of the causes of the erroneous and false data can be ascribed to sloppiness in analysis, lack of training, use of improper techniques, inappropriate data manipulation, taking shortcuts in analysis, failure to follow prescribed steps in the test methods, improper calibration of instruments, etc In a practice called "dry labbing," a worker merely copies the data from a previous sample and submits it as a new sample analysis with a new identification number Commercial, business, and regulatory decisions based on such inaccurate data can often lead to wrong decisions for all affected parties Nancy Wentworth, Director o f quality staff at EPA's Office of Environmental Information, sums up advice to laboratories as: Get the right data; get the data right; and keep the data right [12]

Laboratory Quality Improvement through Quality Management

Using all or many steps outlined above, is it possible to achieve ultimate laboratory perfection? Is such a thing as a "perfect laboratory" or "perfect analysis" possible at all? The answer is in the affirmative if efforts are spent on the underlying foundation of good laboratory practices These are:

9 The tests must be conducted exactly as written without any deviations

9 The calibrations, if part of a test, must be carried out properly; if necessary, each time the test is carried out

9 Quality control must be an integral part o f the analytical sequence The QC data must be plotted, the trends analyzed, and corrective actions taken when so indicated

9 At least annually, all tests performed in a laboratory must be audited by personnel other than those performing these tests or managing the laboratory to check that all details of the test method are properly followed Any deviations found must be corrected promptly Performance of a group of laboratories which are well-controlled in terms of the above requirements has been shown to be superior to that o f other industry laboratories based on the in- house repeatability and their performance in ASTM interlaboratory crosscheck programs Their repeatability in the laboratory is even superior to that advocated in the ASTM test methods themselves Such precise data have implications beyond simple good analysis A properly calculated product specification is usually set based on both the manufacturing and analytical testing variance

Total sigma = (Manufacturing sigm a2 + Analytical sigma2) ~

If a testing laboratory is being managed to its top performance levels, often the analytical sigma becomes virtually insignificant compared to the manufacturing variability resulting in the width of a product specification being dependent almost solely on the manufacturing variance Thus, a superior analytical precision helps in both manufacturing a reliable product, eliminating quality complaints particularly in regards to the data, and keeping the customers happy and loyal

Lessons from Baldrige Award Winners

Since the inception of the Malcolm Baldrige National Quality Award in 1988, a number of U.S organizations has been recognized by the U.S Department of Commerce as the best in total quality management in the country There are a number of lessons to be learned from these winners which are equally valid in the laboratory environment to convert it into a perfect laboratory producing perfect analysis The basic strategies among the Baldrige winners have been summarized as [13]:

I Leadership and management commitment

2 Total customer delight

3 Long term efforts

4 Teamwork with employees, suppliers, and customers

5 Employee involvement and satisfaction

6 Continuing training for all

7 Statistical measurement o f progress

8 Continuous quality improvement

9 Benchmarking against the best in class

10 Total and open communications with employees, suppliers, and customers

Concluding Remarks

A paradigm shift in the mindset is necessary at all levels of laboratory staff to produce a superior product Think of a ham and egg breakfast; the chicken is involved, but the pig is committed You have to be the latter Nor are the results an instant pudding Again, think of a bamboo farmer; the sapling is watered for 4 years, and suddenly at the end the tree shoots up to

60 feet in 90 days! A constancy o f purpose, clear vision, a prepared mind, commitment to long term pursuit of goals, and dedication to all technical aspects o f pillars of good analysis are what will differentiate an ideal laboratory from an also-ran laboratory The choice is yours to make

As W Edwards Deming wrote, "Charles Darwin's law of survival of the fittest, and that the unfit

do not survive holds in free enterprise as well as in natural selection It is a creel and unrelenting law Actually, the problem will solve itself The only survivors will be the companies with constancy of purpose for quality, productivity, and service" [14]

Similar to Robert Pirsig's hypothesis of Zen and the art o f motorcycle maintenance [1], a mental exercise is at the heart of managing a "perfect laboratory" The mind needs to be prepared

to perform a flawless analysis and whatever is required to make it so; otherwise a perfect analysis is simply not possible Determination to excel, mental discipline to stay the course, willingness to overcome inertia and resistance of others, and focus on producing perfect analysis

at all levels of laboratory staff are essential Improvement in laboratory quality performance is a process, not a project A project ends, but a process continues

Inherently we all know the difference between a good laboratory and a not-so-good laboratory The difference is like the difference between a star in a play and the understudy The not-so-good laboratory (and the understudy) do their jobs and produce results, but a good laboratory (and the star) are needed to produce exceptionally perfect results

The technical quality management o f an analytical laboratory can be a hard pill to swallow, but it need not be It can help prevent excessive waste in time and money, improve productivity

by eliminating duplication and waste, and most importantly lead to increased customer satisfaction and increased profits Unless strong and sustained measures are taken, improvement

in data quality cannot be achieved or maintained No pain, no gain!

Only a prepared mind can meet the challenge The bottom line is that one needs to have a dream, a passion, and a vision to want their laboratory to be the best there is - the flawless perfect laboratory Unless one nurtures and pursues that vision, as Yogi Berra said, "You can't

be there if you don't know where you are going"

Additional information on this subject may be found in Refs [15-18]

References

[1] Pirsig, R M., "Zen and the Art of Motorcycle Maintenance," Williams Morrow and Co., Inc., New York, 1974

[2] Kratochvil, B and Taylor, J K., "Sampling for Chemical Analysis," Analytical Chemistry,

Vol 53, No 8, 1981, 924A-938A

[3] Mitchell, J W., "Ultrapurity in Trace Analysis," Analytical Chemistry, Vol 45, No 6, 1973,492A-500A

[4] Zief, M and Mitchell, J W., "Contamination Control in Trace Element Analysis," John Wiley & Sons, New York, 1976

[5] Nadkarni, R A and Bover W J., "Bias Management and Continuous Improvements through Committee D02's Proficiency Testing," ASTM Standardization News, Vol 32, No

[8] Locke, J W., "Quality System for Testing Laboratories," A2LA Reprint, 1992

[9] de Leemput, Van, "Testing Their Calibre - A Standard to Ensure Laboratories are up to Standard," ISO Bulletin, 16-18, June 2000

[10] Broderick, B E., "Laboratory Accreditation: The Operation of an Established Scheme,"

[11] Camp, R C., "Benchmarking," ASQC Quality Press, Milwaukee, WI, 1989

[12] Hogue, C., "Ferreting out Erroneous Data," Chemical and Engineering News, pp 5 and 49-50, April 2002

[13] Nadkami R A., "A Not-So-Secret Recipe for Successful Total Quality Management,"

[14] Deming, W Edwards, "Out o f Crisis," MIT Press, Cambridge, MA, 1982

[ 15] "Quality Assurance for the Chemical and Process Industries: A Manual o f Good Practices," American Society for Quality, Milwaukee, WI, 1987

[16] Nadkarni, R A., "The Quest for Quality in the Laboratory," Analytical Chemistry, Vol 63,

Chris C Onyeso, Ph.D.t

Available online at www.astm.org

Analysis of Gasoline and Diesel Fuel Samples by Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES), Using Pneumatic Nebulizer and Standard Spray Chamber

ABSTRACT: A method has been developed for accurate quantitative determination of additive elements and wear metals in gasoline and diesel fuel in the concentration range of 0 to 50 ppm using inductively coupled plasma atomic emission spectrometry (ICP-AES) This method requires an ICP capable of detecting metals at the 0.02 mg/kg level, and capable of linear calibration over the range of 0.05 to 20.0 mg/kg with a correlation coefficient of 0.9999 or better No additional ICP-AES accessory or lengthy sample preparation is required The PerkinElmer Optima 4300 has been shown to meet this requirement

KEYWORDS: gasoline, diesel fuel, additive elements, ICP, wear metals, and analysis

I n t r o d u c t i o n

Trace metals in fuels, except in the case o f additives, are usually undesirable and normally occur in very low concentrations, requiring sensitive techniques for their determination Also, there is a need to determine the concentration o f additive elements such as manganese (Mn) in gasoline The industry method currently available for the determination of manganese (Mn) in gasoline is ASTM D 3831 [1], which requires bromine reduction o f the gasoline followed by measurement with Atomic Absorption Spectrometer (AA) This method has inherent problems with high olefin content fuels and cannot accurately measure low concentrations o f Mn in gasoline

Use of lCP-AES

ASTM test methods D4951 [2] and D5185 [3] are commonly used for the determination o f additive elements and wear metals in fresh and used lubricating oils respectively The problem with the use o f ICP for gasoline analysis is that gasoline is highly volatile and extinguishes the ICP plasma [4,5] A few examples o f ICP analysis o f gasoline are found in the literature, but

these methods involve the use of expensive ICP accessories such as direct injection nebulizer [6], ultrasonic nebulizer with micro-porous membrane desolvator [7], chilled spray chamber [8,9], a thermostated condenser between the spray chamber and the plasma torch [10], or lengthy sample preparation with possibility o f contamination, such as emulsification with surfaetants [ 11-14]

In this paper, an accurate method for determining the concentration of manganese and other metals in gasoline and diesel fuel by ICP without the use o f chilled spray chamber, direct injection nebulizer, ultrasonic nebulizer with micro-porous membrane desolvator, thermostated condenser, or emulsification is discussed

Manuscript received 15 November 2004; accepted for publication 13 May 2005; published November 2005

Presented at ASTM Symposium on Elemental Analysis of Fuels and Lubricants: Recent Advances and Future Prospects on 6-8 December 2004 in Tampa, FL

I Ethyl Corporation, Richmond, VA

Copyright 9 2005 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959

17

Analysis

Calibration standards are prepared from certified Conostan| standards, using kerosene containing Yttrium internal standard as the solvent The kerosene containing internal standard is also used as the solvent for the gasoline and diesel fuel samples The calibration standards and the gasoline samples are aspirated into the ICP instrument Each element present in the sample emits light at discrete wavelengths The intensity o f light at each element's wavelength is compared against a calibration curve generated from standards containing metals at k n o w n concentrations to provide quantitative determination This method also outlines the preparation

of the instrument, samples, and standards used in the analysis The method was originally developed for the determination o f manganese (Mn) in gasoline and diesel fuel, and later extended to include additive elements and wear metals

Instrument

9 PerkinElmer Optima 4300

9 The appropriate wavelength (nm) for each element (analyte) was chosen

9 Internal Standard is Yttrium (Y)

9 Quartz Torch (no slot) for organic solvents

9 Quartz Torch (single slot) for aqueous solution

9 Spray chamber - cyclonic spray chamber for organic solvent

9 Scott spray chamber for aqueous solution

9 Nebulizer - GemCone nebulizer for organic solvent

9 Crossflow nebulizer for aqueous solution

9 Injector Tube - Quartz injector tube (1.2 m m i.d.) for organic solvent - Sapphire injector tube (2.0 m m i.d.)

Spectrometer

Resolution was set at High with three replicates and integration time o f five seconds

Plasma Parameter

9 Plasma View: radial

9 Peristatic Pump set at a flow rate o f 2.50 ml/min

Calibration

9 The manganese calibration standards concentrations are in mg/kg (ppm)

9 The instrument was calibrated using the following standards: blank (0 ppm), 5 (5.2333 ppm), 10 (10.7836 ppm), 15 (15.6748 ppm), and 20 (20.5006 ppm)

9 A good calibration curve with a correlation coefficient o f 0.99996 was obtained (see Fig

1 below)

Sample Analysis

After calibration, gasoline samples o f known Mn content (as added) were analyzed Results are shown in Table 1

FIG 1 Calibration curve

TABLE 1 Determination of Mn in gasoline samples by ICP

Gasoline Sample Given Cone Measured Cone Difference

7) Gasoline from unknown source 8.3 9.2 0.9

Based on accurate results obtained on the gasoline samples with known Mn content (the gasoline samples were spiked with a known amount of Mn), many gasoline and diesel fuel samples were analyzed for Mn content and the ICP results (Alton) compared with those obtained

by AA at a contract Lab (Tables 2 and 3) The target values are the spiked values

TABLE 2 Determination of Mn in diesel fuel

Sample Type Contract Lab D3831 Alton R & D Target Value

TABLE 3 Determination of Mn in gasoline samples

Sample Contract Lab Alton R & D Target Value

Precision

The manganese method has a repeatability of 0.24 mg/kg At 12.0 mg/kg Mn level, the precision of the method is +/- 1.16 mg/kg Diesel fuel was diluted 100-fold to keep the method consistent; detection limits and precision could be better for diesel at lower dilutions since diesel

is not as volatile as gasoline

Expansion of the Method to Include Additive Elements and Wear Metals

Having successfully used this method to determine Mn in gasoline and diesel fuel, it was desirable to extend the method to include determination of additive and wear metals The method was calibrated with a Conostan| standard containing the desired elements, and a good calibration curve was obtained as shown below (Table 4)

TABLE 4 Multi-element calibration and limits of detection

TABLE 5 Elements detected in samples

Aqueous Analysis

Another way that gasoline can be analyzed by conventional ICP without the use o f additional accessories or lengthy sample preparation is to analyze it in aqueous solution For gasoline and

diesel fuel samples containing solid impurities or particulates that are insoluble in kerosene, aqueous analysis is a very viable option Determination of metals after total destruction o f organic matter is a very reliable procedure for petroleum and petrochemical samples

The gasoline samples were digested with nitric acid in a microwave digestion oven The resulting acidic solution was diluted with de-ionized/distilled water down to about 5-10 % HNO3 and analyzed by aqueous ICP method (Table 6) Sample dilution in the aqueous analysis is about

1 to 50, unlike organic analysis, which is 1 to 100

TABLE 6 Aqueous ICP analysis of gasoline

Gasoline Samples (results in mg/kg)*

Elements Gasoline 1 Gasoline 2 Gasoline 3 Gasoline 4 Gasoline 5 Gasoline 6

*Sn, Cr, P, Ni, Ba, and Mn were below the detection limit in these six gasoline samples

The differences between the results shown for Fe, Pb, Ca, and Si in Tables 5 and 6 are due to the fact that the gasoline samples are different, the dilution factors are different, and the matrices are different (kerosene versus water)

Conclusions

An ICP method for the accurate determination of additive elements and wear metals in gasoline and diesel fuel has been developed The gasoline and diesel sampleswere dissolved in kerosene with yttrium internal standard and analyzed by ICP without any additional accessory Samples were bracketed with matrix matched standards, and an internal standard was also used Based on good calibration curves (correlation coefficients o f 0.9999 and higher), lack of spectral interference, and accurate results obtained on gasoline samples with known amount of Mn, the results for the samples with an unknown amount of metals must be correct For the organic ICP method, the standard deviations for four measurements o f each of the twelve (12) gasoline samples ranged from 0.00 to 0.21 This study has demonstrated the ease of gasoline and diesel fuel analyses both in organic and aqueous matrices by ICP without the use of additional expensive accessories, such as direct injection nebulizer, ultrasonic nebulizer with micro-porous membrane desolvator, chilled spray chamber, a thermostated condenser between the spray chamber, and the plasma torch or lengthy sample preparation

Acknowledgments

Thanks go to Ken Garelick and Penny Hurt for their assistance in the preparation and analysis

of the samples and to Afton Chemical Company

PA

[4] Botto, R I., Spectrochim Acta, Part B, Vol 42, 1987, p 181

[5] Kreuning, G and Maessen, F J M J., Spectrochim Acta, Part B, Vol 44, 1989, p 367 [6] Botto, R I., Canadian Journal of Analytical Sciences and Spectroscopy, Vol 47, No 1,

2002, pp 1-13

[7] Botto, R I., J AnaL At Spectrom., Vol 8, No 1, 1993, pp 51-57

[8] Ryan, A., Chemistry in New Zealand, Vol 61, No 6, 1997, pp 9-11

[9] Hausler, D W and Taylor, L T., AnaL Chem., Vol 53, 1981, p 1223

[10] Maessen, F J M J., Seeverens, P J H., and Kreuning, G., Spectrochim Acta, Part B, Vol

39, 1984, p 1171

[11] Saint Pierre, T D., Dias, L F., Pozebon, D., Aucelio, R Q., Curtius, A J., Welz, B.,

[12] Brenner, L B., Zander, A., Kim, S., Shkolnik, J J., AnaL At Spectrom., Vol 11, No 2,

1996, pp 91-97

[13] A1-Swaidan, H M., Atomic Spectroscopy, Vol 14, No 6, 1993, pp 170-173

[14] Reimer, R A and Miyazaki, A., Journal of Analytical Sciences, Vol 9, No 1, 1993, pp 157-159

B r i a n S F o x 1

Available online at www.astm.org

Elemental Analysis of Lubricating Grease by Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES)

ABSTRACT: Lubricating grease is a mixture with complex properties that poses far more difficult analytical challenges than do its individual components At many petroleum testing laboratories, Inductively Coupled Plasma Atomic Emission Spectroscopy, or ICP-AES, is used for the analysis of additive elements and wear metals in oil However, it may also be used to measure the concentration of both additive metals and contaminants in lubricating grease However, unlike some of the other matrices that are routinely analyzed, grease cannot be simply diluted for direct sample introduction into the instrument without some form of acid digestion This paper will discuss how ICP-AES is used as a grease analysis tool Several sample preparation schemes will be covered, including classical sulfated ash

on a hot plate, microwave assisted dry ash, and microwave assisted acid digestion Comparison data from different digestion techniques will illustrate potential problems that may be encountered by the analyst in each of these methods The advantages of closed-vessel microwave digestion for the analysis of elements that are often volatilized trader normal digestion conditions will be discussed

KEYWORDS: lubricating grease, additives, ICP-AES

Introduction

In 1885, Grant McCargo formulated one o f the earliest petroleum grease mixtures b y blending oil and soap together Today, lubricating grease finds use in almost all bearings around the world Lubricating grease is composed o f approximately 90 % additized oil and soap or other thickening agent [1] In the absence o f an oiling system, grease provides a film o f additized oil to protect the metal surfaces that are in contact with one another The thickener in the grease helps to keep it located in the bearing assembly, where it can do its job, Some of the other additives in lubricating grease mirror the functions that they perform in typical lubricating oil In lubricating oil, the additive and wear metal concentrations are determined by several atomic spectroscopy methods, among them being ASTM D 4951 or ASTM D 5185 However, these tests have no counterparts for testing grease samples

The focus o f most standardized grease testing has been upon performance and appearance, rather than determining the concentration of chemical components In part, this is likely due to the relative volume o f grease demand when compared to other lubricants But the primary reason m a y be that grease is a rather difficult matrix with which to work from the perspective o f the analytical laboratory Lubricating oil m a y be diluted in a solvent and introduced directly into

an Inductively Coupled Plasma Atomic Emission Spectrometer (ICP-AES) for rapid elemental measurements Grease, on the other hand, does not dissolve in any common laboratory solvent that is compatible with ICP-AES

Manuscript received 18 November 2004; accepted for publication 7 March 2005; published September 2005 Presented at ASTM Symposium on Elemental Analysis of Fuels and Lubricants: Recent Advances and Future Prospects on 6-8 December 2004 in Tampa, FL

i Advanced Research Associate, ExxonMobil Research and Engineering, Paulsboro, NJ 08066

Copyright 9 2005 by ASTM International, 100 Ban" Harbor Drive, PO Box C700, West Conshohockon, PA 19428-2959

24

In addition to ICP-AES, an X-ray fluorescence spectroscopic technique is sometimes used to quantify elemental concentrations in grease samples However, it suffers from two major drawbacks The first is the inability to detect two important lubricating grease components, lithium and boron Lithium compounds are used as a thickener in the vast majority o f lubricating grease types on the market Boron is sometimes used with lithium as a complexing agent The second drawback emerges in the case o f the analysis o f used grease obtained from a bearing A grease sample that is displaced b y fresh grease and obtained from outside a bearing housing m a y contain dirt from the environment along with wear metals The grease that is o f primary interest

is that which is sampled from the load zone o f a disassembled bearing [2] Obtained from this source, there is often insufficient material available to fill the X-ray sample cup for a proper analysis

Despite the difficulty in working with grease samples, there are well over a dozen elements

o f interest to the lubricating grease supplier in the form o f additives, contaminants, and wear metals Table 1 shows some o f these elements in lubricating grease and the role that each m a y play Determining their concentrations can be an important aspect o f grease manufacture In addition, a reliable analysis technique can also assist in the process o f troubleshooting problems with new and used grease in the field For example, some o f the thickeners used in the various types o f lubricating grease m a y be incompatible with one another [3] Quantifying the signature elements o f these thickeners could determine the level o f accidental mixing o f two incompatible grease types

T A B L E 1 - - P o t e n t i a l elements found in lubricatin~ ~rrease

Aluminum, A1 Clay or other thickener <20 to 3000 #g/g

Antimony, Sb Extreme pressure additive <20 to 1000 #g/g

Boron, B Complexing additive for thickener <20 to 1000 #g/g

Bismuth, Bi Extreme pressure additive <10 to 1000 #g/g

Calcium, Ca Thickener and extreme pressure additive <20 to 3000 #g/g

Chromium, Cr Wear metal, contaminant Trace to 1000 #g/g

Iron, Fe Wear metal, clay thickener <10 to 5000 #g/g

Lead, Pb Extreme pressure additive <20 to 1000/zg/g

Magnesium, Mg Clay thickener minor element Trace to 100 #g/g

Manganese, Mn Clay thickener minor element Trace to 100 #g/g

Molybdenum, Mo Molybdenum disulfide / extreme pressure additive <20 to 3000 #g/g

Phosphorous, P Additives such as zinc dithiophosphate (ZDTP) <20 to 1000 #g/g

Sulfur, S Additives such as ZDTP <20 to 1000 #g/g

Silicon, Si Contaminant, clay thickener trace element <10 to 1000 #g/g

This paper examines various techniques that are available to prepare lubricating grease samples for analysis b y ICP-AES Each o f the techniques has specific advantages and m a y be applied to different grease sample types, depending upon composition and expected levels o f analyte concentration Special attention will be given to the use o f high-pressure, microwave assisted acid digestion This technique provides the optimal solution for complete digestion o f

the organic matrix, while minimizing loss of additive elements that may be volatilized during other preparation schemes

Instrumentation

The measurements were performed using a Thermo Elemental ]RIS Inductively Coupled Plasma Atomic Emission Spectrometer (ICP-AES) A 2 kW crystal-controlled radio frequency (RE) generator operating at 27.12 MHz powers the plasma source An Echelle optical system with a 381-ram focal length diffracts the light from the plasma source before it is focused onto the Charge Injected Device (CID) camera detector [4]

A fully de-mountable ABC TJA IRIS Radial torch was equipped with a Twister cyclonic spray chamber for the non-hydrofluoric (HF) acid analyses Both are available from Glass Expansion Inc, Pocasset, Massachusetts, USA When analyzing samples prepared with a dilute

HF acid solution, an alumina center tube and HF resistant spray chamber (Glass Expansion ) were substituted The nebulizer used was a Teflon Mira Mist that is available from Burgener Research Inc., Mississauga, Ontario, Canada

The microwave digestion oven used for sample preparation is an MDS 2000 It is available from CEM Corporation, North Carolina, USA The microwave digestion vessels are high pressure, Teflon lined, 45-mL Model 4782 bombs that are available from Parr Instrument Company, Moline, Illinois, USA For microwave assisted dry-ashing of lubricating grease samples, a CEM Microwave Ashing System MAS 7000 was used The open vessel microwave digestion apparatus that was used for this work is a Prolabo Microdigest 401

Reagents and Materials

All water used is 18 Mfl de-ionized water polished by a Simplicity ion exchange system available from Millipore Corporation, USA Trace Metal Grade sulfuric, nitric, and perchloric acids that were used in the procedures are available from Fisher Scientific, Pittsburgh, Pennsylvania, USA Commercially prepared, multi-element, aqueous ICP-AES standards and the blank acid matrix were obtained from VHG Labs, Inc., Manchester, New Hampshire, USA The multi-element, oil standards used for validation are available from Conostan Division, Conoco Specialty Products, Ponca City, Oklahoma, USA

For sulfated ash digestions, either 250-ml Vycor beakers or dishes may be used (Fisher Scientific) The microwave digestion vessels used were high pressure, Teflon lined, 45-mL Model 4782 bombs that are available from Parr Instrument Company, Moline, Illinois, USA The polypropylene volumetric flasks used to dilute digested samples to volume are available from Fisher Scientific

Procedure

The following descriptions of the digestion methods are simplified versions Depending upon sample difficulty, other steps may beadded to obtain a clear solution that is relatively free

of the organic matrix and un-dissolved material

Sulfated Ash Digestion

The sample size depends upon expected analyte concentration Approximately 1-2 g of

material are weighed into a Vycor container of suitable size The sample is then charred on a hot plate until it is reduced to about 0.5 g A heat lamp is used to assist the process Then 1-2 mL o f sulfuric acid are added to the residue and heated until the fumes cease to evolve The charred sample is placed in a muffle furnace at 530~ until the black color is gone This typically takes about 2 h, but it was originally felt that a sample may be left overnight to obtain complete decomposition Approximately 5 mL of hydrochloric acid are added, and the sample is gently heated to dissolve remaining solids The solution is brought to volume in a 50-mL volumetric flask with water

Microwave Assisted Dry Ashing

Approximately 1 g of sample is weighed into a crucible Vycor or platinum crucibles have both been used for this work The MAS 7000 furnace insert allows the use of metal vessels in the microwave unit The crucible is placed into the microwave furnace The temperature is ramped up to 525~ over a two-hour period It then holds the temperature for one hour After the sample cools, approximately 5 mL of a 1:1 nitric acid-water solution are added to the crucible and heated gently to dissolve the ash It is critical that nitric acid only be added after the organic matrix has been ashed and cooled The sample is quantitatively transferred to a 25-mL volumetric and brought to volume with water

Open- Vessel Microwave Digestion

Approximately 1 g of sample is weighed into the digestion vessel, and 10 mL of nitric acid are added The microwave unit is set to 30 W to cause a gentle reflux of the nitric acid After about one hour or when the sample is nearly decomposed, the drop-wise addition of perchloric acid will typically complete the digestion Perchloric acid is an extremely strong oxidizer that may react violently with organic material Experienced personnel only should use it with utmost caution The sample is heated until about 5 mL of solution remain in the vessel The contents are washed into a 50-mL volumetric flask and brought up to volume with water

Closed- Vessel Microwave Digestion

From a safety perspective, it is critical that no more than 0.1 g of sample be used for this technique The sample is weighed into the Teflon digestion vessel Approximately 4 mL of nitric acid are added The vessel is capped and placed into the microwave oven Four vessels are simultaneously processed The microwave is set at 125 W for 15 min The oven then ramps

up to 190 W for another 15 rain Care must be taken not to keep internal temperature and pressure within the capability of the vessels Excessive heat and pressure will cause the digestion bombs to deform and potentially leak After the cycle is finished, the vessels are placed into an ice bath for at least one hour to cool The dissolved sample is washed into a 25-ml volumetric flask and brought to volume with water

ICP-AES Instrument Conditions

The instrument operating conditions are shown in Table 2 Optimization of conditions is important to attain a robust plasma that is minimally affected by differences in acid concentration between standards and samples Calibration was done using a series of external standards with

yttrium (Y) as the internal standard The internal standard was continuously added to standards and samples via a "tee" fitting and mixing loop

TABLE 2 ICP-AES operating conditions

Coolant Gas, Argon 16 1 min "l Auxiliary Gas, Argon 0 1 min "1 Nebulizer Gas, Argon 0.65 1 min -l Sample uptake rate 1.1 ml min 1

Discussion

Lubricating grease does not possess the required qualities to make it a suitable Standard Reference Material, or SRM The oil in the grease tends to separate from the thickener after a few months on the shelf Most of the following work uses samples with expected concentrations, based upon known quantities of well-characterized additive Commercial certified standards were also used to validate the digestion procedure and to determine detection limits

Table 3 shows the elements in lubricating grease, which was prepared with an organic thickener as opposed to a metallic soap All of the results are expressed in weight percent Four repeat analyses were made using each digestion scheme The expanded uncertainty of each measurement is calculated by multiplying the value for of one experimentally derived standard deviation of the data times the critical value The critical value of 3.18 is determined from a student's double-sided t-distribution [5] with 3 degrees o f freedom and a confidence level of

95 % The average of the sulfated ash digestion indicates that nearly 12 % o f the phosphorous is lost from the sample The loss is likely due to volatilization of phosphorous as opposed to sample spatter, based upon the fact that the zinc result is closer to the expected value The dry ash technique shows less average phosphorous loss However, it is still slightly more than that of Parr Bomb or closed vessel, high-pressure microwave digestion scheme

TABLE 3 Additive concentration results in lubricating grease with organic thickener

The grease in Table 4 represents a more typical lithium soap formulation Phosphorous loss

is almost negligible in this grease sample However, boron suffers a nearly 25 % loss from the

expected value when the sample is prepared with the sulfated ash technique A nearly complete loss of boron has also been observed when digesting a commercial multi-element oil standard by the sulfated ash technique Open vessel microwave digestion results were consistent with expected values However, this technique has limited utility due to safety concerns about using perchloric acid to complete the digestion Another problem with the perchloric acid reagent was the presence of boron Despite the use of trace metal grade acid, an unacceptable level of boron was detected This forced the use of a blank correction that could lead to expanded uncertainty Subsequent to the generation of these data, a new source of perchloric acid was found that appears to have suitably low levels of boron contamination

The four digestion schemes were applied to a second lubricating grease sample with lithium soap thickener These results are shown in Table 5 The sulfated ash digestion resulted in loss o f antimony, boron, and phosphorous, while lithium and zinc remained consistent with expected values

TABLE 4 Additive concentration results in lithium thickener lubricating grease

*Result is reported to two significant figures due to loss of precision from acid blank correction

TABLE 5 Additive concentration results in lithium thickener lubricating; grease

*Result is reported to two significant figures due to loss of precision from acid blank correction

The dry ash did not appear to result in nearly as much analyte loss due to volatilization However, this procedure has shown a higher propensity to spatter or ignite, resulting in ejection

of analyte from the container There is some evidence for this in the fact that the dry ash technique sometimes shows the lowest results for non-volatile grease components It also can be more difficult to dissolve the ash from a dry combustion, especially in used grease samples This may be the result of the formation of recalcitrant oxides that require severe heating of the acid to complete the dissolution of the ash

In the absence of a certified reference material that matches the sample, a generally accepted validation technique involves subjecting an analogous certified material to the digestion procedure and extrapolating the results to samples This validation exercise was done on a commercial, multi-element oil standard The results are shown in Table 6 As expected, an average of nearly 65 % of the boron was lost due to volatilization during the sulfated ash

TABLE 6 -Additive elements from commercially prepared metal in oil standard

Element Certified Concentration 2-h Sulfated Ash Parr Bomb Microwave

Element

TABLE 7 Comparison of ZDDP additive elements by ICP-AES

Concentration ASTM D 5185 Sulfated Ash Microwave

a greater portion of volatile analyte is lost When the samples are allowed to heat overnight at 530~ it was observed that nearly all of the antimony and boron are lost More than half of the phosphorous was also lost

The advantages of retaining volatile analytes, while ensuring complete digestion o f a grease sample when using the high-pressure Parr Bomb, start to become apparent One of the major disadvantages of the closed vessel microwave digestion scheme is the sample size limit o f approximately 0.1 g Since 4 mL o f acid are used to digest the sample, the analyst is either faced with an extremely large dilution factor or must deal with the effects o f a strong acid concentration in the final solution

Because these digestion schemes may result in widely varying acid concentrations in the final solution, the ICP-AES conditions require careful optimization for this work Researchers have seen that increasing acid concentration often causes a depression in signal intensity for some lines when using pneumatic sample introduction systems [6] The effect may be especially prominent under non-robust plasma conditions The ICP-AES conditions were optimized using

the Mg II 280.270-nm/Mg I 285.213-um as described by Mermet [7] Varying the nitric acid concentration from less than 4 % v/v to over 18 % v/v caused less than a 2 % change in signal intensity for the lines of the yytrium internal standard

The cleanliness of the laboratory becomes an even more germane issue as the sample size decreases Trace contamination becomes magnified when dilution factors are approximately 250

or more The sample size constraint of the high-pressure microwave digestion scheme tends to raise the minimum detection limits of the analytes in a sample Table 8 shows the result of digesting a certified Conostan standard The elements in the table represent several of those that may be found at relatively low concentrations in lubricating grease, either due to contamination

or wear For the standard with a certified concentration of 100 #g/g, the recovery is good for all listed elements with the relative difference within 4 % Uncertainty and recovery began to deteriorate for the standard with a certified concentration of 10 #g/g Replacing the pneumatic nebulizer with an ultrasonic nebulizer (USN) would likely result in generally lower minimum detection limits for analytes in the samples

TABLE 8 Elemental analysis of certi~ed standards after closed-vessel microwave digestion

or used grease samples Feng et al [8] have found that addition of a small quantity of HF enhances aluminum and silicon recoveries and may be analyzed without passivation by boric acid They also found that potential calcium precipitation in the presence of trace fluoride did not negatively impact their calcium recovery Since both boron and calcium are important additives, these findings are o f interest To date, we have not done much work to validate mixed acid, closed vessel microwave digestions with lubricating grease, but the application to accurate silica quantification in both clay-based and used grease is promising