SUPPLEMENT TO
Volume 31, Number S4b April 2013 www.chromatographyonline.com
ADVANCES IN ION CHROMATOGRAPHY 3PVUJOF$BQJMMBSZHP*$ Dedicated for routine analysis, this Reagent-Free™ High Pressure IC (HPIC™) system provides high pressure capability with multiple detectors in one integrated system. The Thermo Scientifi c Dionex ICS-4000 Capillary HPIC maximizes sample throughput by delivering TJNVMUBOFPVTTVQQSFTTFEDPOEVDUJWJUZBOEOFXDIBSHFEFUFDUJPO of compounds. This innovative HPIC system takes advantage of the new 4 μm particle columns for fast separations and high resolution. MFTTUJNF CFUUFSSFTVMUT
tLearn about the benefi ts of capillary IC at: thermoscientifi c.com/capIC © 2012 All trademarks are the property of Thermo Fisher Scientifi c Inc. and its subsidiaries. of Thermo Fisher Scientifi the property © 2012 All trademarks are Small Particles Maximize chromatographic resolution with unique Ion Chromatography (IC) column chemistries that are designed for high efficiency separations. Now using smaller resin particles, the Thermo Scientific™ Dionex™ IonPac™ 4 μm columns deliver higher resolution for better peak identification and reliable quantification without sacrificing speed. The high resolution IC columns separate ionic analytes from complex sample matrices with ease. A broad range of chemistries offered in three formats – standard bore, microbore, and capillary – enable users to take control of their separations and results. big benefits Thermo Scientific™ Dionex™ ICS-5000+ Reagent-Free™ HPIC™ system ! For detailed information, visit www.thermoscientific.com/4um On-request IC analysis 24/7 with no startup time
Thermo Scientific™ Dionex™ ICS-4000 Capillary HPIC system Dedicated Capillary High-Pressure Reagent-Free IC system
Thermo Scientific™ Dionex™ InGuard™ and OnGuard™ Cartridges Remove matrix interferences for better performance in IC applications
Thermo Scientific™ Dionex™ Chromeleon™ Chromatography Data System software
© 2013 Thermo Fisher Scientific Inc. All rights reserved. © 2013 Thermo Fisher Scientific Inc. All rights reserved. Shortest way from samples to results 4 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com
Advances in IonIon ChromatographyChromatography April 2013
Articles
Ion Chromatography: Yesterday, Today, and Tomorrow 7
Ron Majors A brief introduction to this special supplement by Guest Editor Ron Majors
Landmarks in the Evolution of Ion Chromatography 8
Hamish Small From the invention of eluent suppression to today’s “just add water” concept, pivotal developments over the last 40 years are chronologically highlighted from a chemical and instrumental viewpoint.
Recent Developments in Ion-Exchange Columns for Ion Chromatography 16
Chris Pohl A summary of the latest in new ion-exchange phases for ion chromatography, with a focus on general aspects of phase design and then a review of anion-exchange and cation-exchange columns introduced in the past few years
Ion Chromatography Yesterday and Today: Detection 23
Purnendu K. Dasgupta, Hongzhu Liao, and C. Phillip Shelor A look at the current status of detection in ion chromatography, focusing on the most popular detectors in IC: the conductivity detector and the charge detector
Practical and Popular Sample Preparation Techniques for Ion Chromatography 27
Rosanne W. Slingsby, Kassandra Oates, and Dong-bum Lee The authors present the most common and fundamental techniques that address common matrix issues and discuss the critical chemistry considerations.
Tools for Simulation and Optimization of Separations in Ion Chromatography 33
Boon K. Ng, Greg W. Dicinoski, Robert A. Shellie, and Paul R. Haddad Computer-based methods can simulate separations for a wide variety of analytes, columns, and eluents.
Advances in High-Speed and High-Resolution Ion Chromatography 38
Charles A. Lucy and M. Farooq Wahab The speed of IC has been limited by pressure and flow limits of the hardware. Recent advances are enabling faster IC separations.
Cover image courtesy of Getty Images. Speciation Analysis by Coupling IC with ICP/MS
Separate and Quantify Different Elemental Inductively Coupled Plasma Mass Spectrometry Oxidation States (ICP/MS) The need to distinguish between chemical forms of an element such as ICP/MS combines the ICP source with a mass spectrometer for fast Cr III/VI, As III/V, and organic v. inorganic mercury has become critical for mult-element analysis with extreme sensitivity (sub ppt detection). The food, environmental, and pharmaceutical analysis. In the past, measuring Thermo Scientific iCAP Q™ ICP/MS system represents a unique platform the total amount of an element was sufficient. Unfortunately, the effects of to determine total elemental concentration. The iCAP Q allows for high an element extend far beyond its absolute amount, since different forms sensitivity that can provide single figure ppt detection limits for low mass of an element can exhibit very distinct levels of toxicities. The process of elements. As a result, a full mass range analysis can be carried out for separation and quantification of different oxidation states of an element, routine samples in environmental or food applications. Additionally, the more specifically termed speciation analysis, can be utilized to determine an iCap Q series provides a low mass cut off so unwanted species do not element’s various chemical forms. Speciation analysis can be split into two pass through the quadrupole mass filter. components: separation of individual ionic species by ion chromatography Complete Inorganic Elemental Analysis followed by trace elemental detection and quantification using ICP/MS. The Solutions combined method is termed Ion Chromatography – Inductively Coupled When coupling a Dionex Ion Chromatography system with the iCAP Q, it Plasma Mass Spectroscopy (IC – ICP/MS). Using ICP/MS as a sensitive has essentially become feasible to measure every element in the periodic elemental detector in combination with a selective separation technique table. Both techniques successfully provide a complete picture when it such as Ion Chromatography provides complete information on the chemical comes to analyzing both total elemental concentration as well as the ionic form (species) of an element. state of the element of interest.
Elements Currently Speciated Using IC and ICP/MS Arsenic As+3, As+5 Chromium Cr+3 , Cr+6 Mercury Hg+2, EtHg+, and MeHg+ Selenium Se+4, Se+6
Ion Chromatography (IC) For ion analysis, the Thermo Scientific™ Dionex™ ion chromatography solutions deliver uniquely beneficial technology: Reagent-Free™ Ion Chromatography (RFIC™) systems. These RFIC systems use electrolytic technologies to generate eluents and regenerants from deionized water. The Thermo Scientific iCAP Q with the Thermo Scientific Dionex ICS-5000+ HPIC System The net effect decreases time spent on equilibration, calibration, method verification, and troubleshooting. Minimizing the unintentional variations in preparation of eluents and regenerants discards the need for required consistence checks.
Learn more at thermoscientific.com/Speciation 6 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com
PUBLISHING & SALES 485F US Highway One South, Suite 100, Iselin, NJ 08830 (732) 596-0276, Fax: (732) 647-1235
Michael J. Tessalone Science Group Publisher, [email protected] Edward Fantuzzi Publisher, [email protected] Stephanie Shaffer East Coast Sales Manager, [email protected] (508) 481-5885
EDITORIAL MANUSCRIPTS: To discuss possible article topics or obtain manuscript preparation guidelines, contact the editorial director at: (732) 346-3020, e-mail: [email protected]. Laura Bush Editorial Director, [email protected] Publishers assume no responsibility for safety of artwork, photographs, or manuscripts. Alasdair Matheson Every caution is taken to ensure accuracy, but publishers cannot accept responsibility for Editor-in-Chief, LCGC Europe, [email protected] the information supplied herein or for any opinion expressed. Megan Evans SUBSCRIPTIONS: For subscription information: LCGC, P.O. Box 6168, Duluth, MN 55806- Managing Editor, [email protected] 6196; (888) 527-7008, 7:00 a.m. to 6:00 p.m. CST. Outside the U.S., +1-218-740-6477. Kate Mosford Delivery of LCGC outside the U.S. is 3–14 days after printing. Single-copy price: Managing Editor, LCGC Europe, [email protected] Stephen A. Brown U.S., $10.00 + $7.00 postage and handling ($17.00 total); Canada and Mexico, $12.00 + Group Technical Editor, [email protected] $7.00 postage and handling ($19.00 total); Other international, $15.00 + $7.00 postage and Cindy Delonas handling ($22.00 total). Associate Editor, [email protected] CHANGE OF ADDRESS: Send change of address to LCGC, P.O. Box 6168, Duluth, MN Bethany Degg 55806-6196; provide old mailing label as well as new address; include ZIP or postal code. Assistant Editor, LCGC Europe, [email protected] Allow 4–6 weeks for change. Alternately, go to the following URL for address changes or Dan Ward Art Director, [email protected] subscription renewal: https://advanstar.replycentral.com/?PID=581 RETURN ALL UNDELIVERABLE CANADIAN ADDRESSES TO: Pitney Bowes, P.O. Russell Pratt Box 25542, London, ON N6C 6B2, CANADA. PUBLICATIONS MAIL AGREEMENT Vice President Sales, [email protected] No.40612608. Anne Young Marketing Manager, [email protected] REPRINTS: Reprints of all articles in this issue and past issues are available (500 minimum). Call Tamara Phillips 877-652-5295, ext. 121 or e-mail [email protected]. Outside US, UK, direct dial: 281-419-5725, ext. 121 Direct List Rentals, [email protected] DIRECT LIST RENTAL: Contact Tamara Phillips, (440) 891-2773; e-mail: Wright’s Media [email protected] Reprints, [email protected] Maureen Cannon INTERNATIONAL LICENSING: Maureen Cannon, (440) 891-2742, Permissions, [email protected] fax: (440) 891-2650; e-mail: [email protected]. Jesse Singer Production Manager, [email protected] ©2013 Advanstar Communications Inc. All rights reserved. No part of this publication Peggy Olson may be reproduced or transmitted in any form or by any means, electronic or mechanical Audience Development Manager, [email protected] including by photocopy, recording, or information storage and retrieval without permission Gail Mantay in writing from the publisher. Authorization to photocopy items for internal/educational or Audience Development Assistant Manager, [email protected] personal use, or the internal/educational or personal use of specific clients is granted by Ad- vanstar Communications Inc. for libraries and other users registered with the Copyright Clear- ance Center, 222 Rosewood Dr. Danvers, MA 01923, 978-750-8400 fax 978-646-8700 or visit http://www.copyright.com online. For uses beyond those listed above, please direct your written request to Permission Dept. fax 440-756-5255 or email: [email protected]. Joe Loggia Advanstar Communications Inc. provides certain customer contact data (such as cus- Chief Executive Officer tomers’ names, addresses, phone numbers, and e-mail addresses) to third parties who wish to promote relevant products, services, and other opportunities that may be of interest to Tom Ehardt you. If you do not want Advanstar Communications Inc. to make your contact information Executive Vice President, Chief Administrative Officer available to third parties for marketing purposes, simply call toll-free 866-529-2922 between Steve Sturm the hours of 7:30 a.m. and 5 p.m. CST and a customer service representative will assist you Executive Vice President, Chief Marketing Officer in removing your name from Advanstar’s lists. Outside the U.S., please phone 218-740-6477. Ted Alpert LCGC does not verify any claims or other information appearing in any of the advertise- Executive Vice President, Finance and Chief Financial Officer ments contained in the publication, and cannot take responsibility for any losses or other Andrew Pollard damages incurred by readers in reliance of such content. President LCGC welcomes unsolicited articles, manuscripts, photographs, illustrations and other ma- Georgiann Decenzo terials but cannot be held responsible for their safekeeping or return. Executive Vice President To subscribe, call toll-free 877-527-7008. Outside the U.S. call 218-740-6477. Chris Demoulin Executive Vice President Authorization to photocopy items for internal or personal use or for the internal or personal use of specific clients is granted by Advanstar Communications for libraries Danny Phillips and other users registered with the Copyright Clearance Center, 222 Rosewood Drive, Executive Vice President Danvers, MA 01923, tel. 978- 750-8400, fax 978-646-8700; call for copying beyond Ron Wall that permitted by Sections 107 and 108 of the U.S. Copyright law. LCGC and the Executive Vice President logos appearing on the cover of this magazine are registered trademarks of Advanstar Eric Lisman Communications, Inc. Executive Vice President Advanstar Communications Inc. (www.advanstar.com) is a leading worldwide media Francis Heid company providing integrated marketing solutions for the Fashion, Life Sciences and Vice President, Media Operations Powersports industries. Advanstar serves business professionals and consumers in these Ward D. Hewins industries with its portfolio of 91 events, 67 publications and directories, 150 electronic Vice President, General Counsel publications and Web sites, as well as educational and direct marketing products and services. Market leading brands and a commitment to delivering innovative, quality Nancy Nugent products and services enables Advanstar to “Connect Our Customers With Theirs.” Vice President, Human Resources Advanstar has approximately 1000 employees and currently operates from multiple offices J Vaughn in North America and Europe. Chief Information Officer www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 7
Ion Chromatography: Yesterday, Today, and Tomorrow
s a liquid chromatographer, one tends to lump ion-exchange chromatography (IEC) and ion Achromatography (IC) in the same bucket, but there are actually significant differences in the practice of these methodologies. Both techniques use columns with ionogenic functionality, mobile phases with various buffer compositions, and separate ionic compounds in simple to complex matrices. Early on, IEC had its biggest success in the separation of amino acids, helping Moore and Stein win the Nobel Prize in chemistry in 1972 and in the separation of transuranium elements dur- ing the development of the atomic bomb back during World War II. However, the biggest difference Ron Majors, between the two techniques is based on the needs of the early detection principle of conductivity — the Guest Editor of this the removal of the buffer from the mobile phase before detection. In the early 1970s, Hamish Small special issue, is a Senior and coworkers, while working at Dow Chemical in Midland, Michigan, envisioned a method with Scientist at Agilent a fast separation of non-chromophore-containing ionic compounds but in a nonconducting mobile Technologies and the phase, water. The concept of buffer removal (stripping), later termed eluent suppression, was a key ele- longtime Editor of LCGC’s “Column Watch” ment in the development and differentiation of IC from other ion-exchange separations. It spurred and “Sample Prep the development of specialized low specific exchange capacity packings, sophisticated suppressors Perspectives” columns. with regeneration capabilities, and new detection principles such as indirect photometric and pulsed [email protected] amperometric detection. Throughout the decades, alongside high performance liquid chromatography (HPLC), IC has seen parallel developments in separating ionic compounds in a variety of matrices.
IC has solved many practical problems where problematic cationic and anionic compounds at the parts-per-billion level were more easily measured such as trace chloride in water causing corrosion in power-generation turbines, trace electrolytes in semiconductor processing water, and in environmental water analysis. IC has progressed beyond simple ions in solution and can now measure ionized carbohydrates and alcohols at high pH using electrochemical detection in diverse and difficult samples such as encountered in the food and brewing industries.
In this special issue, a supplement to both LCGC North America and LCGC Europe, we have man- aged to assemble experts from the top IC research and development laboratories from across the world to provide state-of-the art reviews surrounding elements of this well established separation technique. Hamish Small, now a consultant, gives a fascinating historical prospective, from the development of IC at Dow through its commercialization, and explains why and how certain improvements were made along the way. In encountering difficult matrices, sample preparation is equally important in IC as it is in HPLC. Rosanne Slingsby and coworkers from Thermo Scientific (formerly Dionex) provide some practi- cal advice on sample preparation for IC. Chris Pohl, also of Thermo Scientific, who has contributed to many of the advances in the technology, especially from the commercial side, brings us up to date on the most widely used columns in modern IC. Chuck Lucy and Farooq Wahab from the University of Alberta in Edmonton, Alberta, Canada, have been instrumental in advancing high-speed and high-resolution IC and provide examples on the use of shorter, smaller particle columns. Sandy Dasgupta and students from the University of Texas in Arlington discuss the most widely used detectors — the conductivity detector and the charge detector — the latter developed in their own laboratory. Finally, Paul Haddad and his group from the University of Tasmania in Hobart, Australia, discuss their development of simulation software tools for the method development and optimization of separations in IC.
I sincerely hope you find this special issue of interest and think of IC the next time you encounter a challenging problem in the separation and measurement of ionic compounds. 8 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com
Landmarks in the Evolution of Ion Chromatography
8 A personal perspective on the milestones in the development of ion 3 2 chromatography. From the invention of eluent suppression to today’s 4 “just add water” concept, pivotal developments over the last 40 years 1 are chronologically highlighted from a chemical and instrumental viewpoint. Suggested key references on applications and detailed developments are provided. Response (nC)
n the late 1950s, a few of us at the Dow of interest from interfering species, frac- Chemical Company began thinking tion collectors made “cuts” of the effluent 3 I about a new form of chromatography from the ion-exchange column, and the 0 246810 that would challenge the old methods of cuts were analyzed by the classical meth- Time (min) inorganic ion analysis. Those ideas and ods. Because the analytes often had widely subsequent research were the precursors of diverse chemistries and required special what became known as ion chromatography analytical techniques, the task of analysis (IC). I have been privileged to be closely was handled by specialists. It was a slow linked to those early endeavors and to be business and days could go by between the at least an interested observer of the many separation and the results. We envisioned a other important events that followed. From method that would couple fast separation this long perspective I will attempt to iden- with prompt detection and measurement tify ideas, inventions, and innovations that of analytes by a monitor placed at the col- have altered the pace of development and umn outlet. Desirably, this detector should the trajectory of IC in the years since our be “universal” — that is, capable of quan- early imaginings. I call these events land- tifying ions of widely diverse chemistries. marks. But what might we use as a detector? Many In other publications (1,2) I have inorganic ions of interest such as alkali and described in some detail how quite unre- alkaline earth metal ions, ammonium, lated experiences in my early years at Dow halides, nitrate, sulfate, and phosphate influenced the first IC inventions, so for were notably “bland” in not having use- brevity’s sake I will jump directly to the ful chromophores or a general postsepara- first landmark, in 1971. tion means of generating chromophores, as Moore and Stein had done for amino 1971: The Invention acid analysis (4), so spectrophotometric of Eluent Suppression methods of detection seemed a no-go at By 1971, we had decided that what inor- that time. We recognized, however, that a ganic ion analysis needed was a chromato- universal property shared by aqueous elec- graphic technique that would supplant trolyte solutions was electrical conductance many of the tedious, time-consuming clas- and a notable landmark was our decision to sical wet-chemical methods. At this time, exploit that property. But therein we antici- ion-exchange chromatography had made pated a problem. significant contributions to inorganic With the considerable knowledge avail- Hamish Small analytical chemistry (3) but it was usually able to us (3), we were convinced that Direct correspondence to: as an adjunct to wet-chemical methods; ion-exchange chromatography could sepa- [email protected] chromatography segregated the analytes rate any mixture that challenged us. We Capillary Ion Chromatography— Bringing a new level of speed and resolution
Always On, Always Ready, Better Results Peaks: Capillary IC takes performance and ease-of-use to a whole new level, 1. Quinate 25 Dionex IonPac AG11-HC-4 m/AS11-HC-4 m 2. Fluoride 3. Lactate 3600 psi Acetate simplifying ion chromatography analyses while increasing throughput. By 21 22 4. 5. Propionate scaling down column size, injection volumes, and flow rates by a factor 6. Formate 20 7. Butyrate 8. Methylsulfonate of 10 to 100, Capillary IC provides excellent solvent economy, small 9. Pyruvate S 13 17 10. Valerate 8 16 24 26 6 11. Monochloroacetate sample requirements, and high mass sensitivity. Additionally, Capillary Ion 12 14 23 27 29 12. Bromate 9 19 11 25 13. Chloride Chromatography provides true on-request 24/7 uptime with no equilibration 2 15 14. Nitrite 5 7 4 10 18 15. Trifluoroacetate 1 3 28 16. Bromide necessary, thus the system is always on, always ready and delivers better 17. Nitrate 0 18. Carbonate results. Dionex IonPac AG11-HC/AS11-HC 19. Malonate Maleate 20. 2200 psi Sulfate S 21. 22. Oxalate Save Time and Resources 23. Tungstate 24. Phosphate The benefits of capillary IC are both unique and multifaceted. Capillary IC 25. Phthalate 26. Citrate -15 27. Chromate consumes only 15 mL of water a day, translating into 5.2 L a year due to the 0 6 12 18 24 30 36 28. cis-Aconitate Minutes 29. trans-Aconitate low flow rates of 10 μL/min. When a Reagent-free Ion Chromatography™ (RFIC™) Improved Peak Separations Using 4 μm particle size capillary columns system is used in capillary mode, the Eluent Generation Cartridge lasts for 18 months with continuous operation resulting in a system that provides Capillary Ion Chromatography Solutions 24/7 sample analysis. The nonexistent need to equilibrate helps virtually The Thermo Scientific™ Dionex™ ICS-5000+ and ICS-4000 HPIC™ decrease any waiting time, including preparation labor. Systems are the world’s first capillary ion chromatography (IC) systems Speed Up Your Analysis on the market. The Dionex ICS-5000+ operates at standard bore, Capillary IC allows for higher flow rates by optimizing column dimensions microbore and capillary flow rates at pressures up to 5000 psi enabling (i.d. 0.4 mm). The typical capillary flow rate is only 1/100th the eluent flow it to also perform high-resolution and fast IC analysis with the latest 4 rate of an analytical format which typically run at flow rates of 1 mL/min. μm particle-size columns. The Dionex ICS-4000 capillary HPIC system Due to the smaller cross-sectional area, the flow rate on a capillary format is solely dedicated to Capillary IC and provides all the benefits of high- can be scaled down to 10 μL/min enabling dramatically decreased run pressure capabilities for high chromatographic resolution. The Dionex + times while maintaining sufficient resolution. ICS-5000 and ICS-4000 systems are designed to take advantage of small particle IC columns for faster and improved separations. Improve Sensitivity and Selectivity The combination of a first-dimension standard bore column (i.d. 4 mm) with a second-dimension capillary column (i.d. 0.4 mm) permits lower detection limits in the low ng/L range. Combining columns with disparate stationary phase properties is especially important for high-ionic-strength samples. This precludes the need for equilibration and dramatically reduces waiting time, including sample preparation.
Learn more at thermoscientific.com/CapIC
Dionex ICS-5000+ and ICS-4000 Capillary HPIC Systems 10 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com and to conductometry as the detection and ciency. Clearly, this was not going to fly. measurement mode, but instead of placing But the problem suggested a solution HCL the conductivity cell right after the sepa- (5,6): Use separator columns with very NaCl,KCl rating column, we would insert a second low specific capacity, thus enabling elution
Cation-exchange resin column between the separator and the cell. with low concentrations of eluent; this in Initially we called this column the stripper turn would prolong the life of the strip- because its purpose was to strip the efflu- per and thus allow the analyses of many A bed of anion exchanger in the hydroxide form ent of the highly conducting eluent and samples before regeneration became neces- leave just the analytes in water, the ideal sary. And we envisioned that an effective Conductivity cell background for detection. For example, suppressor could be about the same size as using the workflow depicted in Figure 1, the separator, thus minimizing the prob- Resin-OH– + HCl Resin-Cl + H2O Resin-OH– + NaCl Resin-Cl + NaOH we would separate the alkali metals on a lems caused by the suppressor void volume. cation-exchange resin using hydrochloric These concepts were central to our inven- Figure 1: Workflow devised in 1971 us- acid as the eluent, then pass the effluent tion of IC with eluent suppression and con- ing a “stripper” column between the from the separating column through a bed ductometric detection (7). separator column and the conductivity cell, as applied to the separation of the of anion-exchange resin in the hydroxide alkali metals on a cation-exchange resin form, which would strip out the acid and Separation Media using hydrochloric acid as the eluent. present the alkali metals to the conductiv- and Eluents for Early IC ity cell as the metal hydroxides in a water The first successful separation by IC was background. Analogously, anions could the separation of a mixture of lithium, be separated on an anion exchanger using sodium, and potassium using a lightly sodium hydroxide as eluent, while a fol- sulfonated styrene–divinylbenzene (DVB) lowing bed of cation-exchange resin in the polymer as the separator, dilute hydrochlo- hydronium form would remove the sodium ric acid as eluent, and Dowex 1 (hydrox- hydroxide, thus presenting the anions (such ide form) as the stripper. I had used much as the halides) to the conductivity cell as more separator than necessary so elution their acids in a water background. times were unnecessarily long, but other- Thus, we would solve the problem of wise the separation looked excellent, and eluent conductance noise simply by remov- for that era, the detectability levels were ing the eluent and retaining the analytes in impressive (Figure 2). a water background. Because of the very basic eluents antici- Figure 2: The first chromatogram using eluent suppression and conductomet- As a first implementation of this new pated for anion analysis, styrene-based ric detection. The sample injected was idea, we proposed using a strong acid cat- anion exchangers with their great chemi- 0.1 mL of a mixture of LiCl, NaCl, and KCl, ion-exchange resin (Dowex 50, Dow Water cal stability were preferred over the silica- 0.01 M of each. The eluent was 0.02 M and Process Solutions) as separator and a based media that were being widely used HCl. From author’s laboratory notebook, dated November 9, 1971. weak-base resin (Dowex 30) to absorb the in high performance liquid chromatogra- HCl eluent. But that idea was never tried phy (HPLC). However, although surface were also aware that using conventional — for the following reason. sulfonation of styrene-based polymers high-capacity ion-exchange resins would In using the stripper, we were proposing gave a useful cation separator and surface in many cases require quite concentrated something radically new for chromatogra- quaternization of a styrene–DVB polymer (several molar) solutions to displace the phy, a component that would become par- seemed the obvious route to a low-capacity more intractable species and any conduc- tially depleted with each sample injected anion-exchange resin, I saw it as a route tance changes imparted to the effluent by and eventually would have to be regener- fraught with problems. If I used the same the appearance of the analyte ions could ated. We felt that for this technique to suc- synthetic steps as were used to produce easily be overwhelmed by the conductance ceed, this regeneration would have to be high-capacity resins, I anticipated serious “noise” of the eluent. So we decided — as unobtrusive as possible; users should be obstacles to creating, on a styrene-based before 1971 — to abandon ion-exchange able to run many samples before stripper substrate, a thin anion-exchanging shell chromatography and instead develop sta- regeneration became necessary. It became that would not have a diffuse boundary, tionary phases that would separate electro- clear that if we used our first proposal we and diffuse boundaries were anathema to lytes using water as eluent, because water could easily encounter cases where the elu- efficient chromatography. Instead, a prior was the perfect background for sensitive tion of even a single sample would exhaust experience had impressed on me the Vel- conductance measurement. Our early a stripper that at the same time had to be cro-like attachment that anion-exchange efforts to devise such media were spas- massively larger in volume than the sepa- resins formed with their cation-exchanging modic and mostly barren of success. rator to supply enough stripping capacity. counterparts (1,2) so I created the first use- In the fall of 1971, we made a number The large void volume of the stripper bed ful anion separator by treating a surface- of significant decisions that led to a break- would be the source of two serious defects: sulfonated styrene–DVB resin with a through: We would revert to ion-exchange It would greatly prolong elution times and suspension of a colloidal anion exchange chromatography as the separation mode seriously degrade chromatographic effi- resin (Figure 3). Because of the manifold www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 11
Cl- 18
- PO3 15 4
2- SO4 12
9
6 NO - Conductivity (μmhol) 3 F- 3
0 Figure 3: (Bottom) Colloidal anion-ex- 04 812162024 change particles (about 300 nm in diam- Time (min) eter) on surface-sulfonated styrene–DVB particles (about 50 μm in diameter) (top). Figure 4: Status of anion IC using carbonate eluent, circa 1974. advantages of preparing IC media in this pressor and eluent suppression seemed more Although the center of gravity of devel- way (1,2), particularly to the manufacturer, appropriate terms and we adopted them opment now moved to Dionex, our small stationary phases prepared by this means from then onwards. group at Dow stayed involved and had (8) would become, and for many years By 1975, we had established the foun- much still to contribute. remain, the workhorse separating media dations of IC. Particularly, we had devel- for anion analysis by IC. oped the procedures for producing anion 1979: IC without Suppression Improvements in cation IC came rapidly exchangers in colloidal form, a keystone of Although packed-bed suppressors had in our first year but progress in anion anal- stationary-phase production. enabled sensitive conductometric detection, ysis was slower. Although sodium hydrox- they had some drawbacks. Notably, there ide was an excellent choice in that the strip- 1975: IC Goes Commercial was the drifting of elution times for cer- per action produced the ideal background, In 1975, the Dow Chemical Company, tain analyte peaks (5,6) and a few analytes the hydroxide ion was an anion of low ion- which by this time had applied for several such as nitrite were degraded by interac- exchange affinity for anion-exchange resins patents on the new technology, established tions with the resin in the suppressor. Also, of that time and high concentrations were a licensing agreement with Durrum Chem- there were the interruptions for suppressor required to displace many analyte ions. ical, a small company whose main prod- regeneration, even though we had made This high concentration of sodium hydrox- uct was amino-acid analyzers. A separate the interruptions to regenerate less obtru- ide placed a heavy burden on the stripper. business unit was formed within Durrum sive by arranging for the suppressor to last This burden was greatly alleviated by using to pursue the commercialization of IC. about an 8-h day and regenerate overnight. the phenate ion as the eluting anion (5); This unit was later spun out of Durrum to The picture on suppressors changed in the phenate ion in the hydronium-form become Dionex Corp., surely a major land- 1979 when it was shown that ion chroma- stripper was converted to phenol, a very mark in the evolution of IC. In September tography could be accomplished using ion weak acid, which contributed little to the of 1975, Dionex signaled IC’s public avail- exchange and conductometric detection conductivity of the background. The most ability by demonstrating the first commer- without a suppressor (10–12). This new significant development in anion eluents, cial instrument at the fall meeting of the development turned the spotlight on the however, was the discovery of carbonate American Chemical Society. It was also at suppressor and the drawback of the inter- eluents that converted to the weak and this time that the term ion chromatography ruptions for its regeneration. Additionally, therefore feebly conducting carbonic acid was used for the first time. (In 1975, the this new version of IC was often promoted in the second column (9). Thus, carbonate term ion chromatography referred exclu- as avoiding the “complexity” that the sup- eluents became widely used for anion IC sively to the combination of ion-exchange pressor added. The interruptions argument for many years to come (Figure 4). separation, eluent suppression, and conduc- was a legitimate one; the complexity argu- With these developments we introduced tometric detection. In later years it came ment was much less so. It is true that for a new vocabulary to IC; we realized that to embrace a wide variety of techniques of samples sufficiently burdened with analyte, the second bed was really converting the separation and detection.) suppressorless IC is an adequate performer, eluent to a less conducting form rather than That same fall we published the first but it has been demonstrated in practice removing it entirely (“stripping”) so sup- article on the new technique (5). and in theory (13) that the so-called 12 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com Dionex’s 1991 introduction of a flat While these devices used an arrange- membrane continuous suppressor, the ment of membranes and electrodes similar 8 MicroMembrane Suppressor (MMS), to the Jansen device and to the MMS, the 3 2 was a landmark event. In this device, electrode compartments of the SRS were 4 using anion analysis as the example, flushed with deionized water. Using anion 1 the effluent from the separator passed analysis with sodium hydroxide eluent as through the narrow channel between the example, when the electrodes were DC- closely spaced, flat cation-exchange polarized, hydronium ions produced at the membranes in the hydronium form. The anode were driven by the applied field
Response (nC) outside of the membranes was bathed by across the cation-exchange membranes, a continuous stream of sulfuric acid that forcing sodium ions into the cathode com- supplied hydronium ions in exchange partment where they united with the cath- 3 for the sodium extracted from the inter- odically generated hydroxide ions and the 0 246810membrane channel. These devices were sodium hydroxide was flushed to waste. By Time (min) very rugged, could suppress higher eliminating the chemical regenerant, the Figure 5: Determination of carbohy- concentrations of eluent than their pre- SRS eliminated the problem of regenerant drates by IC: 1 = glucose, 2 = fructose, decessors, and, with their low-volume leakage. And in an improved embodiment 3 = lactose (internal standard), and 4 = intermembrane channels, they did not of the SRS, the water effluent from the sucrose. Adapted from reference 43 with permission. degrade the efficiency of the chromatog- IC operation was directed to the electrode raphy to an appreciable extent. compartments, thus eliminating the need complexity of adding a suppressor is the With these developments, the suppressor for an extra water pump (19). price of extracting the maximum sensitiv- had evolved from being a conspicuous part Electrochemistry also revived the ity from conductometric detection. (Please of IC, and something of a bother, to being packed-bed suppressor (20,21). In the note: In the context of this article, the term practically invisible to the user. Dionex Atlas suppressor, a small packed sensitivity describes a technique’s ability But further important developments in bed of ion-exchange resin, embraced by to detect low levels of analyte.) This new suppressors lay ahead. ion-exchange membranes, is continuously development did, however, ignite efforts to regenerated by polarizing the bed. devise better suppressors. Electrochemical Regeneration of Suppressors Revival of the Packed-Bed Continuous Suppression Although the flat membrane continuous Suppressor with Chemical When I performed the first anion separa- suppressor was a major advance, it still Regeneration tion in 1971, I used as suppressor a coil of had some limitations. In the first place, it By the 1990s, analytical chemistry was sulfonated polyethylene tubing immersed required a continuous supply of a chemical augmented by a powerful ally, the com- in a stirred suspension of a cation-exchange regenerant. Secondly, although the mem- puter. With the computer came the ability resin (Dowex 50) in the hydronium form. branes were preferentially permeable to the to automate many operations. Initially, we While the sodium hydroxide effluent suppressing ion, they did allow some leak- had introduced the “dogma” that a packed from the separator was passed through the age of its co-ion; in an anion suppressor, bed needed to suppress many samples lumen, sodium ions diffused across the wall for example, some regenerant sulfuric acid before regeneration, but we now realized of the tubing and exchanged with hydro- leaked into the mainstream, raising the that with automated valve switching, a nium ions from the resin as its particles background conductivity and compromis- small suppressor with just single-sample made bumping contacts with the exterior ing the measurement of analytes. capacity was viable and would require little wall of the tubular membrane. The hydro- As early as 1984, Jansen and others intervention from the user (22). This basic nium ions in turn diffused in the opposite had shown that electrochemistry could be idea was later implemented by manipulat- direction and united with hydroxide ions used in membrane devices to effect eluent ing three small suppressor beds in a clever to form water. These membrane devices suppression (17). By placing electrodes in three-compartment-revolver device (23) worked quite well as a continuous suppres- the regenerant compartments of a device and marketed by Metrohm. Two major sor but were fragile and prone to bursting, of MMS-like construction, ion trans- advantages of this small-bed approach over and because the bed suppressors were quite port across the intermembrane space was the earlier large suppressor beds are the robust and we had more pressing priorities, assisted by the electric potential applied to virtual elimination of peak drifting and we shelved the continuous tubular suppres- the electrodes. However, because electro- minimal degradation of chromatographic sor. When suppressorless IC emerged, we lyte was used in the electrode chambers, efficiency by peak spreading in the void revived the membrane concept and suc- these devices could be expected to show space of the small suppressor bed. ceeded in fabricating a number of more undesirable electrolyte leakage into the rugged devices (14,15) and they became the mainstream. Electrochemical first in a series of continuous, chemically Another landmark in the development Generation of Eluents for IC regenerated eluent suppressors. At about the of suppressors was the electrochemically While electrochemistry was a boon to sup- same time, Ban and others obtained a pat- regenerated suppressor or the Self-Regener- pression, it had another important role to ent on a similar continuous suppressor (16). ating Suppressor (SRS) of Dionex (18). play in IC. www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 13 Pioneering work by Dasgupta and oth- t5IFZ HSFBUMZ TJNQMJGJFE UIF DSFBUJPO PG ric chromatography (IPC) and the detec- ers (24,25) had demonstrated that while gradients. tion principle indirect photometric detec- membrane systems could remove eluent Concurrent with these developments tion (IPD). Although IPD had eliminated in IC they could also be used to intro- in hydroxide generators, new separa- the need for a suppressor and performed duce eluent in a controlled way, simply by tion media with much greater affinity for well in many applications (32), it lacked pumping water to a suitable electrically hydroxide (see a later article in this issue) the sensitivity reach of conductometric polarized membrane device. They also enhanced the impact of the electrochemi- detection and received little promotion recognized that the production of “pure” cal generators in anion analysis. in IC. However, the principle became sodium hydroxide by such a system could widely used in capillary electrophoresis provide major advantages for anion analy- Ion Reflux and Eluent Recycling and somewhat in HPLC. Our work was sis by IC. In the early years of IC, carbon- While the new electrochemically based elu- rewarded by recognition as a milestone in ate eluents were successful and widely used ent generators have made a major change to analytical chemistry (33). but they had a few significant shortcom- the trajectory of IC, they do exhaust and Carbohydrates and alcohols are not ings. One was that carbonate suppressed have to be replaced at a cost to the user. In usually thought of as ionics, but Rocklin to carbonic acid, which has low conduc- the late 1990s, we invented a hybrid of sup- and Pohl discovered how their intrinsic tivity but orders of magnitude higher pression and regeneration where the efflu- ionicity could be expressed and used to conductivity than pure water. Another ent from the suppressor is not discarded to chromatographically separate them (34). was that the carbonate background con- waste but instead is recaptured and used Because carbohydrates are extremely ductivity was lowered by the presence of again as eluent. In principle, these systems weak acids they express their ionicity only analyte; this was not a big issue when the could work perpetually simply by pumping at very high pH, but Pohl and coworkers analyte was abundant, but caused non- water to the IC system. In one embodiment, saw the opportunity of exploiting this and linear responses at lower analyte levels. called ion reflux, the three components of separated carbohydrates by ion exchange A third drawback was that gradient elu- an IC operation — eluent generation, sepa- using strongly basic eluents (Figure 5). tion, as it became more of a requirement, ration, and suppression — were performed The anion-exchange resins of IC, with was complicated by the ramping baseline continuously within a single column, with their great stability in high-pH environ- conductivity of the carbonic acid back- just water as the pumped phase. In another ments, were important facilitators of this ground. Sodium hydroxide (or potassium embodiment of ion reflux the separator new technique. Of course, suppressed hydroxide) as eluent had always been a phase was uncoupled from the other two conductometric detection was a non- sort of holy grail for IC because it could functions, allowing any stationary phase to starter because the carbohydrate anions be suppressed to the ideal background, be used (27,28). would revert to their non-conducting water, but two issues delayed its adoption: Another invention, called eluent recycling form in the typical suppressor, so ampero- Hydroxide ion was a relatively ineffective (29), also enabled reuse of eluent. metric detection methods, particularly displacing ion, thus demanding high sup- Reference 30 provides a comprehensive pulsed amperometric detection (PAD) pression capacities, and it was notoriously and detailed review of these developments (35,36), became important adjuncts to IC difficult to prevent its contamination by in electrochemistry as applied to suppres- and enabled IC separation of a wide vari- omnipresent carbon dioxide that altered its sion and eluent generation in IC. ety of analytes (37). eluting power in unpredictable ways and This application of IC to nonionics was led to unstable backgrounds. Although the Other Detection Methods in IC a distinctly new and different trajectory new MMS suppressors had much greater Although conductometric detection was for the method and certainly a landmark suppression capacity, thus diminishing the method that launched IC, it became event. the first issue, the carbonate-in-the-eluent obvious that other detectors could be used problem remained. The membrane-based with the new separation media. UV detec- IC with Water as Eluent electrochemical generator alleviated the tors were effective when the analytes were Although in the early years we made little contamination problem to a great extent UV-absorbing. And of course, UV detec- progress using simply water as the eluent, because the generator could be provided tors did not require a suppressor, although others have been notably more successful. with ultrapure water and the generator was it should be added that conductometric A significant landmark in the evolution of intrinsically a generator of pure carbonate- detection was often the preferred method IC is the work of Lamb and others using free base (26). when the analyte mixture contained target macrocylic species that form selective and The new eluent generators had these ions, some of which were UV-absorbing reversible complexes with electrolytes in positive features:: and others not. pure water (38). t5IF PQFSBUPS XBT GSFFE GSPN UIF UBTL It had been a sort of dogma in detec- of frequently preparing eluents with its tion that UV detectors were usable only Applications of IC attendant problems of contamination if the analytes were UV-absorbing. We It is impossible, in an article of this length, and occasional operator error. showed, however, that using ion-exchange to do justice to the pioneer users of IC t&MVFOU DPODFOUSBUJPO XBT DPOUSPMMBCMF separation coupled to UV detectors could and the many landmark applications that simply by controlling the current in the indeed be used to detect and sensitively expanded the method’s usefulness. There- generator or the flow rate of the water measure UV-transparent ions (31). We fore, I will restrict my choice to two: the stream, or both. called this combination indirect photomet- first symposium on IC and the series of 14 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com In our earliest embodiments of IC we were proud of our ability to measure at the parts-per-million level; all the efforts Pump Standard that have extended detectability of ions by Electrolytic 10-μL loop water purifier a million-fold are truly landmarks in the evolution of IC. EG Waste
10-mL loop Speculations on the Future of IC Suppressor It is to be expected that IC will continue its penetration into ion analysis and it is Electrolytic water purifier likely that special cases will emerge, like sulfate in the early days or perchlorate Column more recently, where IC will display its Sample unique ability to solve urgent analytical problems. Waste As to the future of IC technology, the indicators already point in the direction of smaller-scale instruments, with their many benefits (40). From our earliest 0.300 days in IC we were aware that the con- F- 1 ppt ductivity cell was unique among chro- 5 ppt SO 2- matographic detectors in its amenability 0.200 4 10 ppt to miniaturization and recent develop-
3- PO4 ments in detectors, where electrodes Cl- - 0.100 NO3 make capacitive contact with the con- NO - Br- 2 tents of capillaries (41), are a significant Response (μS-cm) step in this direction. IC and other forms of liquid chroma- -0.050 tography use elution times of analytes as 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 the defining measure of their chromato- Time (min) graphic behavior. The elution time of an analyte is not a fundamental property in Figure 6: A dual-loop injection valve and concentrator column are used to auto- mate calibration and sample loading. Electrolytic deionization of the conductivity chromatography, but time has the great cell waste produces ionically pure water for in-line standard dilution, sample load- advantages of ease of measurement and of ing, and loop rinsing. This system results in extremely low detection limits (parts per almost indefinite subdivision. However, trillion), low blanks, and high precision. The chromatogram was generated using a for elution time to be a reliable definer of Dionex ICS3000 system with a KOH generator and ASRS300 suppressor (2 mm). Col- umns: IonPac AG17 and AS17 (2 mm); flow rate: 0.5 mL/min; loop volumes: 10 μL and the chromatogram, the chromatographic 10 mL. Concentrator column: UTAC-ULP2 low pressure anion concentrator. (Adapted pump must deliver very stable and pre- with permission from Trovion Singapore Pte Ltd.) cisely controlled flow. The pump is thus a flowmeter as well as a mover of eluent. developments that have produced astound- purity: the power-generating industry and As a result, the pump is often the most ing advances in the detectability limits the electronics industry. The first must pro- highly engineered and costly compo- of IC. (References 32 and 37 are recom- tect its critically important boilers from the nent in the system. Further, reductions mended for accounts of the myriad applica- corrosive effects of ions, notably chloride, in pump size do not seem to be keeping tions of IC.) while in the second, sensitive electronic pace with the miniaturization of station- After the introduction of IC in 1975, components can be irreparably damaged ary phases and detectors. it took a certain boldness to embrace this by traces of electrolytes in processing water. A more fundamental property than an brand-new technology; there were many Careful monitoring of water streams is vital analyte’s elution time is its elution volume, who said it would not prosper, or as one to both these industries so there is great and development of volume flow meters as guru expressed it — so I’ve been told — demand to extend the sensitivity of IC as a separate chromatographic device would “IC was fatally flawed” by the suppressor. much as possible. By adhering to scrupu- relieve the pump of this task. This should, But many did embrace it, and some of their lously clean procedures, such as the in situ in turn, lead to a significant reduction early creative applications are recorded in production of ultrapure water by electro- in pump size and complexity. We have the proceedings of that first symposium at chemical deionizers, adventitious contami- recently taken a step toward reducing Gatlinburg organized by the U.S. Environ- nation of samples and eluents can be avoided pump size by inventing an electrochemi- mental Protection Agency (39). and techniques have been developed that cally driven pump (42). There are at least two industries that enable the detection and measurement of In IC there is a property of analyte elu- depend on water of the highest available ions at the parts-per-trillion level (Figure 6). tion that is even more fundamental than www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 15 elution volume. Assuming that the eluent (9) H. Small and J. Solc, “Ion Chromatography (32) P.R. Haddad and P.E. Jackson, Ion Chroma- and the stationary phase have been estab- – Principles and Applications” Proceedings of tography: Principles and Applications (Elsevier, lished, then, with one important caveat, An International Conference on “The Theory Amsterdam, 1990). the number of equivalents of displacing and Practice of Ion Exchange”, Society of (33) Milestones in Analytical Chemistry (American ion required to elute an analyte is fixed; Chemical Industry, London (1976). Chemical Society, Washington DC, 1994). or the number of coulombs to elute is (10) K. Harrison and D. Burge, Pittsburgh Con- (34) R.D. Rocklin and C.A. Pohl, J. Liq. Chrom. fixed if an eluent generator is being used. ference on Analytical Chemistry, Abstract 6, 1577 (1983). (This is true only if the analyte and the #301 (1979). (35) S. Hughes and D.C. Johnson, Anal. Chim. displacing ion are of the same valence; (11) D.T Gjerde, J. Fritz, and G. Schmuckler, J. Acta 149, 1–10 (1983) it is more complicated if they are not, Chrom. 186, 509–519 (1979). (36) T.R. Cataldi, C. Campa, and G.E. Bene- but there are solutions for this compli- (12) J.S. Fritz, D.T. Gjerde, and C. Pohlandt, Ion detto Fresenius, J. Anal. Chem. 368, 739–758 cation.) So IC chromatograms might be Chromatography (Dr. Alfred Huthig, Heidel- (2000) defined not by time or volume but by the berg, 1982). (37) J. Weiss, Ion Chromatography, second edition number of coulombs applied. Does this (13) H. Small, Ion Chromatography (Plenum (VCH, Weinheim, 1995). mean not only that the pump need not Publishing, New York, 1989), pp. 180–186. (38) J.D. Lamb and J.S. Gardner, Macrocyclic be supremely stable but that we might (14) T.S. Stevens, J.C. Davis, and H. Small, Anal. Chemistry: Current Trends and Future Perspec- also in some cases even dispense with a Chem. 53, 1488–1492 (1981). tives, Karsten Gloe, Ed. (Kluwer Academic flow meter? In IC, the field is open for (15) T.S. Stevens, J.C. Davis, and H. Small, U.S. Publishers, Dordrecht, the Netherlands, innovation in how eluent is delivered and Patent 4,474,664 (1984). 2005), pp. 349–363. its flow measured. (16) T. Ban, T. Muryama, S. Muramota, (39) E. Sawicki, J.D. Mulik, and E. Wittgenstein, IC eluent generators include a reservoir and Y. Hanaoka, U.S. Patent 4,403,039 Eds., Chromatographic Analysis of Environ- of concentrated acid or base as the source (1983). mental Pollutants (Ann Arbor Science, Ann of the eluent and the membrane or mem- (17) K.H. Jansen, K.H. Fisher, and B. Wolf, U.S. Arbor, Michigan, 1978). branes that separate this reservoir from Patent 4,459,357 (1984). (40) Y. Liu, V.M.B. Barreto, C.A. Pohl, and N. the mainstream must be of substantial (18) C. Pohl, R. Slingsby, J. Stillian, and R. Gajek, Avdalovic, Patent Application # 20120228227 thickness to prevent leakage of the base (or U.S. Patent 4,999,098 (1991). (2012). acid) into the mainstream. A reservoir of (19) J.R. Stillian, V.B. Barreto, K.A. Friedman, (41) J.A.F. da Silva and C.L. do Lago. Anal. Chem. ion-exchange resin, the perfectly nondif- S.B. Rabin, and M. Toofan, U.S. Patent 70, 4339–4343 (1998). fusible electrolyte, would offer a remedy 5,248,426 (1993). (42) H. Small, Y. Liu, and C.A. Pohl, U.S. Patent for this problem and is worth examining. (20) H. Small, Y. Liu, J.M. Riviello, N. Avdalovic, 8,133,373 (2012). And the neglected area of eluent recycling and K. Srinivasan, U.S. Patent 6,325,976 (43) Dionex Application note #92 (ThermoFisher (27–29) is likely to be re-examined in the (2001). Scientific, Sunnyvale, California). years ahead. (21) J.M. Anderson, R. Saari-Nordhaus, B.C. In the next several years, the landscape Benedict, C. Sims, Y. Gurner, and H.A. Hamish Small of IC will change as dominant patents Pham, U.S. Patent 6,468,804 (2002) was born and received expire and other players enter the field. (22) H. Small, J. Riviello, and C. Pohl, U.S. Patent his early education Where this might lead is beyond specula- 5,597,734 (1997) in Northern Ireland tion, but we can be certain that new inven- (23) H. Schafer, M. Laubli, and P. Zahner, U.S. where he received B.Sc. and M.Sc. tions and innovations will emerge to chal- Patent 6,153,101 (2000). degrees from the lenge the status quo, just as they have since (24) D.L. Strong, P.K. Dasgupta, K. Friedman, Queen’s University of the beginning of IC. and J.R. Stillian, Anal. Chem. 63, 480–486 Belfast. He worked (1991). from 1949 to 1955 for the United Kingdom Atomic Energy Authority at Harwell in Eng- References (25) P.K. Dasgupta, D.L. Strong, J.R. Stillian, land before immigrating to the United States (1) H. Small, Chem. Heritage 21(3), 10–11 and and K.A. Friedman, U.S. Patent 5,045,204 and joining the Dow Chemical Company in 38–41 (2003). (1991). Midland, Michigan. He remained at Dow until (2) H. Small, J. Chem. Educ. 81(9), 1277 (2004). (26) Y. Liu, H. Small, and N. Avdalovic, U.S. Pat- 1983, before retiring to pursue his career in independent research and invention. He has (3) J. Inczedy, Analytical Applications of Ion ent 6,225,129 (2001). maintained a close association with Dionex Exchangers (Pergamon Press, Oxford, 1966). (27) H. Small and J. Riviello, Anal. Chem. 70(11), since its inception. Small is credited with 49 (4) S. Moore and W.H. Stein, J. Biol. Chem. 192, 2205–2212 (1998). U.S. patents, several of which cover key inven- 663–681 (1951). (28) H.Small, U.S. Patent 5,914,025 (1999) tions in ion chromatography. Direct corre- spondence to: [email protected] ◾ (5) H. Small, T.S. Stevens, and W.C. Bauman (29) H. Small, Y. Liu, and N. Avdalovic, Anal. Anal. Chem. 75, 1801–1809 (1975). Chem. 70(17), 3629–3635 (1998). (6) H. Small, Ion Chromatography (Plenum Pub- (30) Y. Liu, K. Srinivasan, C. Pohl, and N. lishing, New York, 1989). Avdalovic, J. Biochem. Biophys. Methods 60, For more information on this topic, (7) H. Small and W.C. Bauman, US Patent 205–232 (2004). please visit 3,920,397 (1975). (31) H. Small and T.E. Miller, Anal. Chem. 54, www.chromatographyonline.com (8) H. Small and T.S. Stevens, U.S. Patent 462–469 (1982); U.S. Patent 4,414,842 4,101,460 (1978). (1983). 16 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com
Recent Developments in Ion-Exchange Columns for Ion Chromatography
10 1 Ion-exchange chromatography is a relatively mature area of
4 chromatographic separation yet advances in this technique continue 3 11 unabated. This article provides a summary of the latest in new ion- 12 9 10 6
Conductivity (μS) 8 13 exchange phases for ion chromatography. It starts by focusing on 2 5
7 general aspects of phase design and then reviews anion-exchange 0 -1 and cation-exchange columns introduced in the past few years. 0510 15 20 25 30 Time (min)
on chromatography (IC) contin- Stationary-Phase Architecture I ues to be the chromatographic Stationary-phase construction for IC technique most widely used for columns comprises nine basic archi- the separation of ionic and ioniz- tectures: silane-based modification of able compounds, with a special focus porous silica substrates, electrostatic- on the analysis of inorganic anions, agglomerated films on nonporous inorganic cations, small hydro- substrates, electrostatic-agglomerated philic organic acids, and aliphatic films on ultrawide-pore substrates, amines. Although a number of sepa- polymer-grafted films on porous ration modes are included under the substrates, chemically derivatized umbrella term of ion chromatography, polymeric substrates, polymer-encap- ion exchange is by far the most widely sulated substrates, ionic molecules used technique in IC. Although use adsorbed onto chromatographic sub- of ion-pair techniques in conjunction strates, step-growth polymers on poly- with reversed-phase columns remains meric substrates, and hybrid materials a viable alternative to ion exchange based on a combination of a silane- for IC applications, ion exchange modified silica substrate with a poly- continues to be the focus of develop- meric exterior surface coating. Five of ment when it comes to new stationary these — electrostatic-agglomerated phases designed for specific applica- films on ultrawide-pore substrates, tions involving the separation of ionic polymer-grafted films on porous compounds. There are a number of substrates, chemically derivatized reasons why ion exchange has proven polymeric substrates, polymer-encap- to be the preferred separation tech- sulated substrates, and step-growth nique in IC. These include a broad polymers on polymeric substrates range of available selectivities, the — represent the architectures most ability to tailor selectivity for spe- widely used in recently introduced cific applications, the exceptional phases. Hence it’s worth taking a chemical stability of polymeric ion- deeper look at these five architectures exchange materials, the ability to sep- to better understand their relative Chris Pohl arate ions of similar size, and rapid strengths and weaknesses. Thermo Fisher Scientific, Sunnyvale, Califor- nia. Direct correspondence to: chris.pohl@ equilibration when operated in the Electrostatic agglomerated films thermofisher.com gradient mode. on ultrawide-pore substrates: For www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 17
Figure 1: Ion chromatography stationary-phase architectures most widely used in recently introduced phases: (a) electrostatic agglomerated ultrawide-pore substrates, (b) polymer-grafted film on porous substrates, (c) chemically derivatized polymeric substrates, (d) polymer-encapsulated substrates, and (e) step-growth polymers on polymeric substrates. the most part, electrostatic agglom- to construct materials with substan- substrate. With the optimal ratio of erated films on nonporous substrates tially higher capacity (1). The pore substrate pore size to colloidal particle have been largely supplanted by higher size of the ultrawide-pore substrate size, the resulting material can exhibit capacity versions utilizing ultrawide- and the particle size of the colloidal 6–8 times the capacity achievable on pore substrates (Figure 1a). By using ion-exchange material are chosen an identical particle size nonporous an architecture similar to that based such that the pore size is large enough substrate (that is, 30–150 μEq/mL on nonporous substrates, but making to accommodate a coating of ion- for materials using an ultrawide pore use of substrates with pore sizes in exchange colloid on both the interior format compared to 5–30 μEq/mL for the 100–300 nm range, it is possible and the exterior surfaces of the porous materials using a nonporous format). 18 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com Given the increasing importance of capacity suppressor devices, this sta- Polymer-grafted films on porous high capacity chromatographic mate- tionary phase architecture has seen substrates: This type of material rials in IC and the availability of high wide application in recent years. (Figure 1b) is widely used to prepare high capacity packings where cross- linking is not required for selectivity control. Chromatographic materials of 10 this sort are prepared through attach- 1 ment of polymer strands to the surface of a substrate (2,3). To prepare such 4 materials, the substrate is either pre- 3 pared with polymerizable groups on 11 the surface, the surface is modified to introduce polymerizable groups, or 12 the surface is modified to introduce 9 10 6 an initiator species. Resin, monomer
Conductivity (μS) 8 13 2 5 (or monomers), and initiator are then allowed to react to produce a compos- 7 ite polymer graft with polymer strands 0 projecting from the substrate sur- face. Because including a crosslink- -1 ing monomer into the reaction mix- 0510 15 20 25 30 Time (min) ture will cause the reaction mixture to form a gel with substrate particles Figure 2: Isocratic separation of sulfur species and inorganic anions. Column: IonPac suspended in the gel, this synthesis AS25, 4 mm; eluent: 37 mM potassium hydroxide; eluent source: EGC III KOH; flow approach precludes the use of cross- rate: 1 mL/min; injection volume: 25 μL; temperature: 30 °C; detection: suppressed linking monomers. The fact that no conductivity, ASRS 300 4 mm, AutoSuppression recycle mode, 92 mA. Peaks: 1 = fluo- ride, 2 = bromate, 3 = chloride, 4 = nitrite, 5 = bromide, 6 = nitrate, 7 = carbonate, crosslinker can be used in grafted poly- 8 = sulfite, 9 = sulfate, 10 = iodide, 11 = thiocyanate, 12 = perchlorate, 13 = thiosulfate. mer films limits the ability to control
Autosampler First dimension Second dimension
Pump AS diverter valve Waste Waste CD 2 EG Injection valve 1 1 Large loop Waste Rinse External Water CRD 2 Load 2 6 Inject 3 5 4 3 Suppressor 2 5 Diverter valve 6 2 4 1 Waste 4 Injection valve 2 First-dimension 5 column (4 mm) CD 1 3 Second dimension column (0.4 mm) 6 2
1 Concentrator Suppressor 1 column Pump EG (lonSwift MAC-200)
External base Load concentrator Waste CRD 300 Transfer to 2D Rinse concentrator
Figure 3: Two-dimensional ion chromatography instrumental setup. www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 19 take place beneath the surface in the dense polymer matrix of the substrate (a) will exhibit sluggish mass transport 1 and relatively poor chromatographic 3 65.00 performance. Early examples of this stationary-phase architecture exhib- ited relatively poor performance but 6 4 newer materials such as the IC SI-52 4E column (Showa Denko) illustrate 7 5 that high performance materials can 8 9 indeed be constructed in this manner. 1 2 Polymer-encapsulated substrates: Conductivity (μS) 18.00 Professor Gerard Schomburg of the Max Planck Institute in Mulheim- Concentration: 7.00 mM Ruhr, Germany, pioneered this type 0 of material (Figure 1d) as a means of 0 10 20 30 40 50 60 preparing materials for reversed-phase (b) chromatography using alumina as 10 the base material. Synthesis of poly- 160.00 mer-encapsulated materials is accom- plished by combining the substrate, 1 130.00 a preformed polymer with residual
6 double bonds, and a suitable free radi- cal initiator dissolved in an appropri- ate solvent, stripping off the solvent 7 to leave a polymer film on the surface 2 of the substrate, and then curing the 3 8 Conductivity (μS) 4 film at elevated temperature to yield 9 5 a crosslinked film permanently encap- Concentration: 6.00 mM sulating the substrate. The advantage 0 of this architecture is that chemi- 0 10 20 30 40 50 60 70 cal attachment to the surface of the Time (min) substrate is not required, allowing it to be used with inorganic substrates Figure 4: Two-dimensional ion chromatography analysis of haloacetic acids: (a) not amenable to covalent modifica- D1 columns: Dionex IonPac AG24A, AS24A, 4 mm; flow rate: 1.0 mL/min; eluent: tion. Although initially developed as potassium hydroxide, 7 mM (0–12 min), 7–18 mM (12–32 min), step to 65 mM at 32.1 min; suppressor: Dionex ASRS 300, 4 mm; current: 161 mA; loop: 500 μL; oven: a means of producing a reversed-phase 15 °C. (b) D2 columns: Dionex IonPac AG26, AS26, 0.4 mm; flow rate: 0.012 mL/ material based on alumina, the tech- min; eluent: potassium hydroxide, 6 mM (0–50 min), step to 160 mM at 50 min, nique was later adapted by Schom- step to 130 mM at 57 min; suppressor: Thermo Scientific Dionex ACES anion capil- burg’s group as a means of preparing lary electrolytic suppressor; current: 25 mA; concentrator: MAC-200; oven: 14 °C. a weak-cation-exchange phase using Matrix: A. HIW (250 ppm Cl, 250 ppm SO4, 150 ppm HCO3, 10 ppm NH4Cl). Peaks: 1 = MCAA, 2 = MBAA, 3 = DCAA, 4 = BCAA, 5 = DBAA, 6 = TCAA, 7 = BDCAA, a preformed butadiene-maleic acid 8 = CDBAA, 9 = TBAA. copolymer as the encapsulating poly- mer (4). The first commercial intro- selectivity in such grafted films. This tion reaction is generally unknown duction of stationary phases based on architecture is mainly used in applica- in commercial products. In general, this approach brought about a major tions that require a stationary phase chromatographic materials of this shift in stationary-phase design as with relatively high capacity and high sort have substantial capacity because applied for the separation of inor- water content. Such materials can be functional groups are not necessarily ganic cations. Before the introduction prepared from either polymer-based or limited to the surface of the substrate. of this new synthesis method, nearly silica-based substrates, but in practice Such materials have become popular in all separation products were based on nearly all such materials are produced recent years as column capacities have strong-acid cation-exchange station- using polymeric substrates. shifted higher. The critical difficulty ary phases. Since that time, nearly Chemically derivatized polymeric with this stationary-phase synthesis all stationary phases utilized for the substrates: This type of material (Fig- methodology is the requirement that separation of inorganic cations have ure 1c) tends to involve proprietary the derivatization be constrained to used weak-cation-exchange carboxylic synthesis techniques, so the actual the surface to achieve good chromato- acid–based stationary phases. A disad- chemistry used for the derivatiza- graphic performance. Reactions that vantage of this synthetic approach is 20 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com can be controlled. Figure 2 shows an example application of this column for the analysis of sulfur species, an 1 application not possible on previous columns using this architecture. 600 3 In addition, two new specialty high capacity columns, the IonPac AS24A 500 4 (4 mm i.d.) and IonPac AS26 (0.4 mm 400 i.d.) columns (Thermo Fisher Scien- tific), were introduced. These columns 300 5 are specifically designed for analysis of
Response (mV) 200 2 haloacetic acids via two-dimensional 6 7 (2D) IC. The technique of 2D chro- 100 matography uses two columns in a switching arrangement and is useful to 02468 101214161820222426 achieve improved separations of com- Time (min) plex mixtures by selecting (heart-cut- ting) unresolved components from the Figure 5: Ion chromatography separation of transition metal cations. Column: 150 primary column and further separating mm X 4.6 mm Metrosep C5; eluent: 6 mM oxalic acid, 3 mM citric acid adjusted to pH them on the secondary column that has 4.2 (KOH); flow rate: 1 mL/min; PCR reagent: 0.15 mM 4-pyridylazoresorcinol, 0.4 M ammonium hydroxide, 80 mM nitric acid; PCR flow rate: 0.5 mL/min; detection: ab- a different selectivity than the primary sorbance (vis) at 530 nm. Peaks: 1 = copper (5.00 mg/L), 2 = nickel (3.00 mg/L), 3 = zinc column. The technique is also useful (4.00 mg/L), 4 = cobalt (5.00 mg/L), 5 = lead (30.0 mg/L), 6 = manganese (4.00 mg/L), for samples where there is a mismatch 7 = cadmium (7.00 mg/L). in concentration levels such that quan- titation of the smaller concentration the possibility of swelling and shrink- mary amine or ammonia or a trifunc- solutes is jeopardized. Figure 3 shows a ing of the phase during gradients or tional epoxy monomer, it is possible to schematic representation of the instru- temperature programming depending introduce branch sites. The resulting mental setup that allows these two col- on the cure conditions of the film. In surface composite can be exception- umns to be used in combination for addition, even if the coating is free ally hydrophilic because the epoxy the analysis of trace levels of haloace- of surface defects, alkaline reagents monomer and the amines used in its tic acids in the presence of high levels can still attack the underlying silica construction contain only aliphatic of common inorganic anions such as by penetrating to the surface coating, substituents. And yet, such materials chloride and sulfate. On the left side resulting in bed collapse. are completely compatible with high- of the schematic is shown the config- Step-growth polymers on poly- pH mobile phases that tend to damage uration for the 4-mm columns. Ana- meric substrates: This simple yet most hydrophilic stationary phases. lytes of interest are separated at least versatile synthesis method has seen partially on the 4-mm column after wide use in recent years (Figure 1e). New Anion-Exchange which they pass through suppressor Over the past decade, more than 10 Chromatography Columns 1, a carbonate-removal device (CRD), anion-exchange columns have been A number of new anion-exchange a conductivity cell, the diverter valve, introduced using this stationary- columns were introduced in the last and finally onto a concentrator col- phase architecture. This synthesis few years. Thermo Fisher Scientific umn located in injection valve 2 where approach is a hybrid of the first and extended its Dionex IonPac stationary the analyte bands are refocused before second architectures described above. phases with the introduction of Ion- reinjection onto a capillary column. Stationary-phase preparation begins Pac AS25 polymeric anion-exchange Because the second-dimension column with functionalization of a wide-pore columns with alkanol quaternary has a 100-fold smaller cross-sectional substrate to introduce anionic surface ammonium functionality. This col- area, a 100-fold increase in detec- charges (5,6). Then, an epoxy-amine umn uses the type 5 architecture tion sensitivity is achieved compared copolymer is formed in the presence described above but unlike prior ver- to using columns of identical inter- of this material, producing an amine sions of this chemistry it uses alternat- nal diameter for both portions of the rich “basement” polymer that is elec- ing reactions with a di-epoxide and a separation. Figure 4 shows an example trostatically bound to the resin sur- di-tertiary amine to produce linear separation using these two columns face. Finally, in a repetitive series of projecting strands. Earlier versions of in this 2D IC application. The two reactions, this polymer-coated sub- this architecture had relatively poor columns were developed in collabora- strate is allowed to react with first an selectivity for sulfate and sulfite. By tion with a team of scientists from the epoxy monomer containing at least controlling the gap between the ter- United States Environmental Protection two epoxy functional groups and then tiary amine sites in the polymer chain, Agency (USEPA). Because drinking an amine or ammonia. By using a pri- the selectivity for these two anions water samples vary widely in terms of www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 21 are diverted to waste rather than being passed through the second column. Both columns use type 5 architecture to achieve good selectivity and excel- 794.0 lent peak shape for haloacetic acids, 796.0 250 mm X 4 mm which are highly polarizable anions and 1 2 3 798.0 6 tend to exhibit poor peak shape on most 800.0 4 5 polymeric anion-exchange columns. 802.0
804.0 1 2 3 New Cation-Exchange 6 Chromatography Columns 806.0 4 5 150 mm X 4 mm In the area of cation-exchange col- 808.0 umns, Metrohm recently introduced Response (μS/cm) 810.0 2 two new cation-exchange columns: 3 812.0 1 6 the Metrosep C5 column and more 4 814.0 5 100 mm X 4 mm recently, the Metrosep C6 column. 816.0 The Metrosep C5 is a strong-acid cat- 818.0 ion column with a polystyrene–divi- 820.0 nylbenzene substrate and is based on 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 26.0 28.0 30.0 32.0 34.0 36.0 38.0 type 3 stationary-phase architecture. Time (min) The column allows the separation of Figure 6: Ion chromatography separation of alkali metals and alkaline earths. Column: transition metal cations when using Metrosep C6, 4 mm; eluent: 1.7 mM nitric acid, 1.7 mM dipicolinic acid; flow rate: 0.9 chelating mobile-phase components, mL/min; temperature: ambient; detection: conductivity. Peaks: 1 = lithium, 2 = sodium, as shown in Figure 5. This column 3 = ammonium, 4 = potassium, 5 = magnesium, 6 = calcium. is designed to be used in conjunction with postcolumn addition of colori- metric metal chelating agents such as 4-pyridylazoresorcinol (PAR), allowing 6 4 low parts-per-billion detection limits. In addition, Metrohm also recently 3 5 introduced the Metrosep C6 column, 5 7 which is based on type 4 stationary- phase architecture. The column is well- 4 6 1 suited to samples with analytes exhibit- ing extreme concentration differences. 3 The utility of the column can be dem- onstrated, for example, by environmen- 2 tal water samples, where low levels of 2 ammonium can be quantified in the
Conductivity (μS) presence of 12,500-fold higher concen- 1 trations of sodium. With the Metrosep C6 column, all of the more common alkali metals and alkaline earths can 0 be determined in a single run (see Fig- ure 6). –1 Another new cation-exchange col- 0 1 2 3 4 umn targeting suppressed IC appli- Time (min) cations is the IonPac CS19 column (Thermo Fisher Scientific). This weak- Figure 7: Fast ion chromatography analysis of common inorganic anions. Column: cation-exchange column makes use of 100 mm X 4.6 mm TSKgel SuperIC-Anion HS; eluent: 3.8 mM sodium bicarbonate and 3 mM sodium carbonate; flow rate: 1.5 mL/min; injection volume: 25 μL; temperature: an ultrawide-pore substrate with type 30 °C; detection: suppressed conductivity, ASRS 300 4 mm, AutoSuppression, recycle 2 stationary-phase architecture. It is mode, 77 mA. Peaks: 1 = fluoride (1 ppm), 2 = chloride (3 ppm), 3 = nitrite (5 ppm), 4 = the first column in the IonPac cation- bromide (10 ppm), 5 = nitrate (10 ppm), 6 = phosphate (15 ppm), 7 = sulfate (15 ppm). exchange column family to use ultra- wide-pore substrate morphology. The ionic composition and ionic strength, in a high-capacity format. The second substrate enables exceptionally good the first of these two columns used in column is of somewhat lower capac- chromatographic performance for inor- the 2D configuration was developed ity as most of the matrix components ganic cations while providing excellent 22 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com peak shape for common aliphatic ence new column introductions in IC as ion-exchange columns tend to have amines, which is more difficult to well. Smaller particles tend to provide relatively high ion-exchange capaci- achieve with higher surface area media improved separation efficiencies com- ties. Increased ion-exchange capacity is with small pore size. pared to larger particles and allow the important for challenging applications use of shorter columns to achieve simi- where analytes span a wide concentra- Columns with a lar separations. For example, Tosoh Bio- tion range. A trend toward the use of Smaller Internal Diameter science recently introduced the TSKgel columns with smaller internal diam- In addition to the new columns men- SuperIC-Anion HS column, which it eters is clearly apparent with the 2-mm tioned above, there is an increas- describes as a “hypervelocity anion anal- i.d. columns now widely available and ing trend toward the use of smaller ysis column.” Based on 3.5-μm particle capillary column formats available in internal diameter columns in IC. size media, the new column is available most common column chemistries. In Microbore (1–2 mm i.d.) and capil- in a 100 mm × 4.6 mm format, opti- addition, particle sizes for IC columns lary columns (<1 mm i.d.) have two mized to take advantage of the highest have decreased in recent times, follow- main advantages. The first is higher plate count possible with small particle ing the trend in HPLC and UHPLC. sensitivity in analyses with a limited media for fast analysis of common inor- amount of sample. If the same mass ganic anions (see Figure 7). In addition, References of sample is injected onto a column Shodex has introduced IC SI-35 4D col- (1) T. Stevens, U. S. Patent 4,351,909, May with a smaller internal diameter, the umn, which is based on 3.5-μm particle 21, 1981. peak height will increase, perhaps size polyvinylalcohol media in a 150 mm (2) D. Jensen, J. Weiss, M.A. Rey, and C.A. allowing one to measure smaller con- × 4 mm column format. Taking advan- Pohl, J Chromatogr. 640(1–2), 65–71 centrations of substance. Second, to tage of the small particle size and asso- (1993). maintain the same separation time, ciated high chromatographic efficiency, (3) J. Riviello, M. Rey, J. Jagodzinski, and the linear velocity must be the same. this phase enables fast analysis of the C.A. Pohl, U. S. Patent 5,865,994, Feb- For a column with a smaller internal common anions and oxyhalide disinfec- ruary 2, 1999. diameter, this means that the flow rate tion by-products in less than 14 min. (4) P. Kolla, J. Köhler, and G. Schomburg, must be decreased proportionally to The same column is useful for the anal- Chromatographia 23, 465 (1987). the inverse radius-ratio squared. The ysis of a number of common aliphatic (5) C. Pohl and C. Saini, J Chromatogr. A result is lower usage of solvent. carboxylic acids. Following the same 1213(1), 37–44 (2008). Over the past few years there has trend, over the past 12 months, Thermo (6) C.A. Pohl and C. Saini, U. S. Patent been a significant expansion in the Fisher Scientific has made three popu- 7,291,395, November 6, 2007. range of column chemistries available lar column chemistries available with in 2-mm i.d. columns. Metrohm added 4-μm particle size substrates: the Ion- four new chemistries to its 2-mm col- Pac AS18-4μm, IonPac AS11-HC-4μm Christopher Pohl is the Vice President umn portfolio: the Metrosep A Supp (both using type 1 architecture), and of Chromatography 10, Metrosep A Supp 15, Metrosep A IonPac CS19-4μm chemistries. Each Chemistry in the Supp 16, and Metrosep C4 column of these three columns was separately Chromatography & chemistries. Thermo Fisher Scien- optimized targeting specific applica- Mass Spectrometry Division of Thermo tific added the IonPac AS24A, IonPac tion areas. The IonPac AS18-4μm col- Fisher Scientific. Chris- AS25, and IonPac 26 column chemis- umn was optimized for fast analysis of topher joined Dionex, tries to its already extensive range of common inorganic anions; the IonPac now part of Thermo Fisher Scientific, in 1979 2-mm column chemistries. In addi- AS11-HC-4μm column was optimized where the focus of his work has been new stationary phase design. He is an author tion, over the past few years, Thermo for high-resolution separations of inor- or coauthor of 50 US patents in a number Fisher Scientific has expanded its col- ganic anions and carboxylic acids; and of areas, including separation methods, umn portfolio to include more than the IonPac CS19-4μm column was opti- stationary-phase design, suppressor tech- 20 different column chemistries in mized for high-resolution separations of nology, solid-phase extraction, capillary electrophoresis techniques, and accelerated capillary (400 μm i.d.) column for- inorganic cations and aliphatic amines. solvent extraction (ASE). He is the author or mats. coauthor of four book chapters and more Conclusions than 95 scientific papers published in peer Columns Based on New ion-exchange columns for IC reviewed scientific journals. He received his BS in Analytical Chemistry from the Univer- Reduced-Particle-Size Media continue to be introduced each year, sity of Washington in 1973. ◾ With the advance of ultrahigh-pressure as improvements in column selectivity liquid chromatography (UHPLC), the progress. The growth in ion exchange high performance liquid chromatog- is spurred by new environmental regu- raphy (HPLC) community has seen lations and growing concerns over pos- For more information on this topic, a significant increase in the number sible food contamination. Most new please visit of small-particle-size columns avail- columns have been packed with rugged www.chromatographyonline.com able for HPLC applications. In recent polymeric-based materials, and this years, this trend has begun to influ- trend will undoubtedly continue. Most www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 23
Ion Chromatography Yesterday and Today: Detection
300 Here, we look at the current status of detection in ion chromatography (IC), focusing on the most popular detectors in IC: the conductivity detector and the charge detector. Conductivity detection, including the 200 capacitively coupled contactless conductivity detection, is discussed. Charge detectors developed in the author’s laboratory are shown to complement conductivity detectors. 100 in article title (Google Scholar)
Documents with “ion chromatography” he great power of ion-exchange chro- Conductivity detectors, on the other hand, 0 matography to bring about complex can sense all ions, but there was no simple 1980 1990 2000 2010 T Year separations was aptly demonstrated way to use them. For analyte ions to be by Moore and Stein (1); this work eventually eluted from available (high-capacity) ion- led to their receiving the 1972 Nobel lau- exchange columns in a reasonable period, rel. They were able to separate all 50 amino other (eluent) ions are needed in significant acids in 175 h (slightly more than a week) concentrations. With a high conductivity using a 100 cm × 0.9 cm column packed background, minor conductivity changes with 25–37 μm cation-exchange resin, at accompanying the elution of analyte ions a flow rate of 67 μL/min. They recognized would have been impossible to detect. It that not only pH affects the retention of was the genius of Small, Stevens, and Bau- amino acids by controlling their ionization man (2) that solved this problem through a but that temperature can also profoundly unique solid-phase postcolumn reactor that affect such equilibria. They used a pH gra- literally made the separation visible to the dient from 4.25 to 11.0, and temperatures of conductivity detector, and a unique electro- 25–75 °C at various points during the pro- statically agglomerated stationary phase that file. Interestingly, in this paper Moore and was both efficient and of sufficiently low Stein thanked William Bauman of the Dow capacity to be used in that configuration. Chemical Company for supplying them The stationary phase was akin to pellicular with an ultrafine cation-exchange resin (not ion-exchanger stationary phases proposed then commonly available). earlier by Horvath and colleagues (3) but far Detection in the Moore and Stein work more robust and functionally no different in was off-line with fractions collected and its attributes than present-day superficially reacted with ninhydrin before colorimetric porous particles. measurement. Two decades later, amino The original discovery and the principles acid analyzers with postcolumn ninhydrin of suppressed conductometric ion chroma- reaction and flow-through colorimeters and tography (IC) are discussed in detail by high performance liquid chromatography Small in the present issue (4). The solid- (HPLC) with flow-through UV absorbance phase postcolumn reactor survives in the detectors were in existence. For the analy- form of a unique three-position alternating Purnendu K. Dasgupta, sis of simple inorganic ions, however, many revolving device (5) but has largely been Hongzhu Liao, and major anions of interest (such as sulfate) supplanted by electrodialytically regenerated C. Phillip Shelor have no useful optical absorption or, in the membrane devices (6). Here, we focus on Chemistry and Biochemistry Department, case of chloride and other anions, absorb the evolution and the current status of the University of Texas at Arlington, Arlington, Texas. Direct correspondence to very poorly except at very low wavelengths two detectors unique to IC: the conductivity [email protected] where eluents are also likely to absorb. detector and the charge detector. 24 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com illary ~1 mm apart. An excitation voltage is applied to one electrode; this voltage usually has a frequency of several hundred kilohertz, but some operating at a frequency as low as 200 Hz have been reported (9). This excita- 7.00 tion signal is capacitively coupled through the capillary walls to the solution inside
Acetate and travels to the other electrode. There are many different approaches to measure Sulfate (goes off scale) Sulfate (goes off Nitrite (goes off scale) Nitrite (goes off the conductance. In the simplest approach,
3.00 Sulfite a current-to-voltage converter is connected between the pickup electrode and ground. The resulting signal is amplified and recti- Sulfide fied and is directly proportional to the solu- tion conductance. Kuban and Hauser have Conductance (μS/cm)
-1.00 Carbonate
Cyanide repeatedly reviewed design and application Arsenite Borate Silicate of C4D techniques; the latest appeared in 2013 (10). A readily available inexpensive capacitance-to-voltage high resolution digi- -5.00 tal converter behaves very much like a C4D 0.00 10.00 20.00 30.00 40.00 (11,12) but saturates at specific conductance values of ~100 μS/cm. As yet, C4D has not Time (min) been much used in conventional IC. But it Figure 1: Two-dimensional gradient hydroxide elution ion chromatogram. Red trace: sup- has many virtues and it is merely a matter of pressed conductivity detector; blue trace: second-dimension detector (after baseline off- time before it is used more extensively in IC. set, background ~25 μS/cm). Column: AS11HC; mobile-phase gradient: 0–3.0 mM KOH in In galvanic conductivity measurements, 10 min, hold at 3.0 mM until 15 min, ramp to 10 mM 15–20 min, ramp to 20 mM 20–30 min, ramp to 30 mM 30–40 min. Analyte amounts: 6.25 nmol for nitrate, borate, acetate, and there are two basic arrangements: a four- sulfate; 12.5 nmol for the others, except carbonate (deliberately not added). Adapted from electrode and a two electrode arrangement. reference 17 with permission. The four-electrode arrangement uses an outer pair of electrodes to apply a constant current through the system and then the voltage drop across the inner pair is moni- − tored; this voltage drop is reciprocally related H2O Cation exchanger to the conductance. Although the four- electrode technique is the gold standard for From suppressor H+ X- resistance measurements in the solid state Anion exchanger and used in some CE instruments, it has + H O seen little use in IC. 2 A common geometry for a flow-through Figure 2: Basic schematic for the charge detector. electrical conductivity detection cell consists of two disk-shaped stainless steel electrodes Conductivity Detection bore), whether packed or wall-coated. With (~1–1.5 mm in diameter) that are spaced Conductivity detection has been and contin- such dimensions, it is not practical to make ~1 mm apart, or two ring-shaped electrodes ues to be the mainstay of IC. In the analyte measurements in a separate cell, external spaced 4–5 mm apart. A temperature sen- concentration range of interest, conductiv- to the separation system, because of exces- sor, usually a low thermal mass thermistor, ity is linearly related to ionic concentration: sive dispersion from connecting tubing and is also placed in the thermal block contain- The slope is dependent on the specific ion or other flow-path components. The measure- ing the cell to measure and correct for the ions. At high concentrations, the response is ment must be made directly on-capillary. temperature dependence of the measured less than linear, but this area typically is not Although fine wire electrodes in galvanic conductance. A typical correction coef- of interest in IC. contact with the solution inserted through ficient assumes that conductivity increases Conductivity detectors are fundamen- holes drilled through the capillary walls (7) 1.7% per Celsius degree, but most detectors tally of two types, differentiated by whether or at the end of the capillary (8) have been allow user-selectable temperature compensa- or not the electrodes are in galvanic contact demonstrated for capillary electrophoresis tion values to be input. (Two major vendors with the solution. It is possible to make mea- (CE) systems, a more elegant and generally of IC equipment both specify temperature surements from outside a glass or polymeric applicable solution is capacitively coupled constancy better than 0.001 °C. We remain tube without electrodes contacting the liq- contactless conductivity detection (C4D). somewhat skeptical about how well these uid. Currently available IC equipment does In C4D, a pair of ring-shaped electrodes dry cell block temperature specifications not yet use very small capillaries (<100 μm is put on the separation or measurement cap- relate to the temperature constancy of the www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 25 (17) and then detected with a second detec- tor. The detection principle of this second 35 detector will be the same as that in a nonsup- (a) 21 22 pressed hydroxide eluent IC. The results of 20 such a two-dimensional detection are shown 1314 16 in Figure 1. This is a powerful adjunct to 24 26 29 8 12 1719 23 27 9 11 25 conventional suppressed IC. While not 2 6 15 34 7 commercially available, commercial systems 1 5 10 18 28
Conductivity (μS) are now available that have two independent -5.0 IC systems, each equipped with an eluent 11 generator. Such systems can be configured (b) 20 22 21 24 to carry out such dual detection. Note that 26 29 the ratio of the peaks in the two detectors 25 14 8 1113 19 27 is indicative of the pK of the acid and thus 9 16 a 4 7 12 17 23 23 5 6 10 15 serves as an independent identifier.
Current (μA) 1 28 18 Charge Detection 3.0 The charge detector (18) is the most recently 0.0 10 20 30 42 introduced detector for IC. It is generically Time (min) responsive to ions but the detection prin- ciples differ from those of a conductivity Figure 3: Gradient hydroxide elution on a 250 mm X 0.4 mm, 4-μm AS11HC column with a flow rate of 15 μL/min and temperature set at 30 °C. (a) conductivity detector, detector and hence a charge detector is a (b) charge detector. Injection volume: 0.4 μL. Peaks: 1 = quinate (10 mg/L), 2 = fluo- good adjunct to the latter. The basic con- ride (3 mg/L), 3 = lactate (10 mg/L), 4 = acetate (10 mg/L), 5 = propionate (10 mg/L), figuration of the charge detector is the 6 = formate (10 mg/L), 7 = butyrate (10 mg/L), 8 = methanesulfonate (10 mg/L), 9 = same as that of an electrodialytic suppres- pyruvate (10 mg/L), 10 = valerate (10 mg/L), 11 = monochloroacetate (10 mg/L), 12 = bromate (10 mg/L), 13 = chloride (5 mg/L), 14 = nitrite (10 mg/L), 15 = trifluoroacetate sor except that it uses a cation-exchange (10 mg/L), 16 = bromide (10 mg/L), 17 = nitrate (10 mg/L), 18 = carbonate (20 mg/L), membrane (CEM), which is held negative, 19 = malonate (15 mg/L), 20 = maleate (15 mg/L), 21 = sulfate (15 mg/L), 22 = oxa- and an anion-exchange membrane (AEM), late (15 mg/L), 23 = tungstate (20 mg/L), 24 = phosphate (20 mg/L), 25 = phthalate which is held positive, while the suppressor (20 mg/L), 26 = citrate (20 mg/L), 27 = chromate (20 mg/L), 28 = cis-aconitate (20 mg/L), 29 = trans-aconitate (20 mg/L). (Courtesy of Thermo Scientific.) effluent flows between the two membranes. The current between the electrodes (applied flowing liquid. If thermal noise were the erally not used, to avoid electrochemical voltage is 2–12 V DC, typically 6 V DC) only noise, in a measured background con- processes at the electrodes. Conductometry is the analytical signal. There is a small ductance of 1 μS/cm, the noise level will be is not an electrochemical technique — no background current from the residual ionic 20 pS/cm, an order of magnitude less than chemistry occurs at the electrodes. However, impurities in the water and also from the the 100–300 pS/cm observed in practice. As when the background conductance is low, as autodissociation of water itself. a reference point, the best refractive index with hydroxide eluents or a carbonate eluent When an electrolyte (for suppressed detectors in HPLC claim a temperature with a carbon dioxide removal device, DC anion IC this is necessarily an acid, but in constancy of 0.005 °C.) In the simplest measurements have been shown to provide principle it can be any electrolyte) moves arrangement, an alternating voltage, 1–20 a very simple inexpensive alternative (13). into the detector, H+ and X-, proceed kHz in frequency, is applied across the elec- In the stopped flow mode, a DC detector through the CEM and AEM, respectively, trodes and the resulting current is measured, serves as a chronoamperometric device. The to the oppositely charged electrodes; it is rectified, and converted to a voltage signal. temporal signal profile reflects the ionic the charge carried by these ions that is mea- The ratio of the current to applied voltage is mobility of the anion in the cell; this helps sured (Figure 2). This allows universal cali- the conductance. identify a peak beyond retention time (14). bration for fully ionized electrolytes given 2- Both the above measurements are, how- A major shortcoming of suppressed con- that 1 μM SO4 produces the same signal ever, affected by the capacitance at the elec- ductometric IC is its reduced response to as 2 μM Cl-, and so on. The charge detec- trode–solution interface. A bipolar pulse weak acids. For acids with a pKa ≤ 7, practi- tor is a destructive detector: It is essentially conductance measurement technique is cally no response is observed. On the other a deionizer; the charge transport occurring not affected by the presence of capacitance, hand, hydroxide eluent nonsuppressed IC during deionization is what is measured. either serially or in parallel to the resistive can provide a response (15); however, chro- Interestingly, compared to the conductiv- element. Two successive voltage pulses of matographically such eluents are essentially ity detector, the charge detector has a rela- equal magnitude but opposite polarity are unusable with real samples. Ideally, the best tively higher response for weak electrolytes. applied to the measurement cell. The cur- of both worlds can be attained if, after the Whereas a conductivity detector responds rent passing through the cell at the end of first (conventional) suppressed conductivity only to the ionized portion of a weak elec- the second pulse in measured. This is the detector, a small amount of hydroxide (such trolyte, as deionization occurs in the charge technique typically used in high-end con- as sodium hydroxide) is introduced electro- detector, more of the undissociated material ductivity detectors. DC voltages are gen- dialytically (16) or by Donnan breakdown ionizes to maintain equilibrium. To what 26 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com
(10) P. Kuban and P.C. Hauser, Electrophoresis 34, 55–69 (2013). 300 (11) M. Takeuchi, Q.Y. Li, B. Yang, P.K. Dasgupta, and V. E. Wilde, Talanta 76, 617–620 (2008). (12) I.K. Kiplagat, P. Kubáň, P. Pelcová, and V. Kubáň, J. Chromatogr. A 1217, 5116–5123 (2010). (13) D. Qi, T. Okada, and P.K. Dasgupta, Anal. 200 Chem. 61, 1383–1387 (1989). (14) T. Okada, P.K. Dasgupta, and D. Qi, Anal. Chem. 61, 1387–1392 (1989). (15) T. Okada and T. Kuwamoto, Anal. Chem. 57, 829–833 (1985). 100 (16) A. Sjögren and P.K. Dasgupta, Anal. Chim. Acta 384, 135–141 (1999). (17) R. Al-Horr, P.K. Dasgupta, and R.L. Adams, in article title (Google Scholar) Anal. Chem. 73, 4694–4703 (2001). (18) B.C. Yang, Y.J. Chen, M. Mori, S.I. Ohira, Documents with “ion chromatography” A.K. Azad, P.K. Dasgupta, and K. Srinivasan, 0 Anal. Chem. 82, 951–958 (2010). 1980 1990 2000 2010 Year Purnendu K. Figure 4: Number of documents retrieved by Google Scholar with the phrase “Ion (Sandy) Dasgupta Chromatography” in the title. The solid line indicates the best fit to an asymptotic is the Jenkins Garrett approach to a plateau. Data from 2012 are incomplete and were not used in the fit. Professor of Chemistry at the University of Texas at Arlington. He degree it is ultimately deionized depends on Srinivasan, Yan Liu, Christopher Pohl, has authored over 400 the applied electric field and the residence and Andrea Willie for their generous and journal articles and time in the detector. As such, the effect of prompt help with sought-after informa- book chapters and is the eluent flow rate (reciprocally related to tion. We apologize that space constraints the inventor on 23 US patents, including sev- eral related to IC. He is the recipient of the the detector residence time) is far greater for precluded much else that should otherwise 2011 ACS Award in Chromatography. Direct a weak acid anion than a strong acid anion. have been included. correspondence to: [email protected] Even at normal chromatographic flow rates, the charge detector has enhanced response References Charles Phillip for weak electrolytes like organic acids, as (1) S. Moore and W.H. Stein, J. Biol. Chem. 192, Shelor is a senior PhD shown in Figure 3. 663–681 (1951). student who defacto manages the Dasgupta (2) H. Small, T.S. Stevens, and W.C. Bauman, Laboratory. Phillip is Conclusions Anal. Chem. 47, 1801–1809 (1975). working on sensitive Undoubtedly, as has been proven in the (3) C.G. Horvath, B.A. Preiss, and S.R. Lipsky, methods of urinary past, IC and its modes of detection will con- Anal. Chem. 39, 1422–1428 (1967). iodine analysis. He has authored three publi- tinue. In Figure 4, we plot the number of (4) H. Small, LCGC North Amer. 31(S4b), 8–15 cations and is the recipient of the 2012 Ameri- documents Google Scholar retrieves as hav- (2013). can Chemical Society Division of Environmen- ing the exact phrase “Ion Chromatography” (5) Metrohm Corporation web site, http://www. tal Chemistry Graduate Student award. in the title. The best fit to an asymptotical youtube.com/watch?v=8KQktLNIHf8&nore approach to a plateau model is indicated by direct=1, accessed March 2013. Hongzhu Liao the solid line. The predicted plateau ordi- (6) Thermo Scientific. Advances in chemical sup- is graduate student nate value is 420, which would suggest that pression. Http://www.dionex.com/en-us/ in chemistry at the University of Texas we are only two-thirds of the way to the webdocs/61381-Bro-SuppressorHistory- at Arlington. Hong- plateau; when we reach that, the time for 22Dec08-LPN1855-01.pdf, accessed March zhu focuses on new the next paradigm-shifting technology will 2013. methods in ion chro- have arrived. (7) X. Huang, T.K. Pang, M.J. Gordon, and matography, especially the detection of weak R.N. Zare, Anal. Chem. 59, 2747–2749 acids. ◾ Acknowledgments (1987). We thank Hamish Small first of all for (8) X. Huang and R. N. Zare, Anal. Chem. 63, planting the seed from which this tree has 2193–2196 (1991). For more information on this topic, grown. We also thank him for sharing a (9) S. Pencharee, P.A. Faber, P.S. Ellis, P. Cook, J. please visit draft version of his article that also appears Intaraprasert, K. Grudpan, and I.D. McKelvie, www.chromatographyonline.com in this issue. We are indebted to Kannan Anal. Meth. 4, 1278–1283 (2012). www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 27
Practical and Popular Sample Preparation Techniques for Ion Chromatography
Techniques that could be included in a discussion of sample preparation for ion chromatography are as diverse as the panorama of sample types (a) containing charged or ionizable targets. In this article, we present the
(b) 25 most common and fundamental techniques that address common matrix 9 23
18 22 24 2 19 20 21 3 11 12 13 14 151617 (c) 1 468 issues, and discuss the critical chemistry considerations. We show the 02 46 8 1012141618202224262830323436384042 Time (min) expected improved data quality by using sample preparation.
on chromatography (IC) is first ruggedness can be unacceptable with- I and foremost a chromatographic out it. The good news is that there analytical technique used for the are quite a number of separation col- analysis of ionizable as well as perma- umns available for IC, each developed nently ionized target analytes. Com- to meet a key analytical need. These ponents of the sample can interfere columns may have higher capac- directly with the targets or with the ity or optimized selectivity, or both. eluent components, thus altering the For some samples, choosing the right elution behavior. They also can foul column that can separate analytes the analytical column or compromise of interest from matrix interferences analyte detection. The interferents can eliminate the need for sample can be charged or neutral, so knowl- preparation. Likewise, some detection edge of the chemistries involved is principles such as inductively coupled important to minimize time spent plasma–mass spectrometry (ICP-MS) optimizing a method. The interfer- can eliminate or minimize sample ents can be identified (known) or preparation through the technique’s be unknowns characterized by their inherent selectivity. behaviors. Many of the most common The capabilities available in a well application problems in IC and their stocked “toolbox” of columns can sample preparation solutions have be invaluable in time and money been described (1,2). savings. For example there are a number of “high-low” applications Have the Right where ratios as high as 10,000:1 Toolbox of Columns sodium:ammonium can be managed The goal of sample preparation, in isocratically simply by choosing the any form, is to achieve good quantifi- separation column designed for the Rosanne W. Slingsby, Kassandra cation, reproducibility, accuracy, and analysis (3). In such a case, when the Oates, and Dong-bum Lee long system life. Sample preparation best column is selected for the job, Thermo Fisher Scientific, Sunnyvale, California. Direct correspondence to: is a cost component in an analysis and there is no additional cost for sample [email protected] is used because data quality or system preparation. As most any tradesperson 28 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com
Table I: Cartridge-type sample preparation products
Ions Removed Chemistry Example of Commercial Products
OnGuard II H and InGuard (Thermo Cations and neutralization Sulfonic acid, H-form Scientific), IC-H (Alltech), IC-H (Metrohm)
OnGuard II Na and InGuard Na (Thermo Cations without pH change Sulfonic acid, Na-form Scientific), IC-Na (Alltech & Metrohm)
OnGuard II M (Thermo Scientific), Transition metals without other cations Iminodiacetate IC-Chelate (Alltech)
Chloride and other halides by OnGuard II Ag and InGuard Ag (Thermo Sulfonic acid, Ag-form precipitation Scientific), IC-Ag (Alltech & Metrohm)
OnGuard II Ba (Thermo Scientific), Sulfate Sulfonic acid, Ba-form IC-Ba (Alltech)
Halides and cations Sulfonic acid, Ag- and H-forms, two-layer OnGuard II Ag/H (Thermo Scientific)
Anions and neutralization Quaternary ammonium, OH-form IC-OH (Alltech)
Anions and neutralization Quaternary ammonium, bicarbonate form OnGuard II A (Thermo Scientific)
Halides, sulfate, and cations Sulfonic acid, Ag-, Ba-, and H-forms OnGuard II Ba/Ag/H (Thermo Scientific)
OnGuard II RP (Thermo Scientific), Hydrophobic species Styrene–divinylbenzene IC-RP (Alltech and Metrohm)
Hydrophobic species Styrene–divinylbenzene, hydrophilic InGuard HRP (Thermo Scientific)
Hydrophobic species Octadecylsilane IC-C18 (Metrohm)
Phenolics, azo dyes, humic- and Polyvinylpyrrolidone OnGuard II P (Thermo Scientific) tannic-acids
will say, having the right tool for the Sample Preparation ary phases are chosen to extract ionic job makes all the difference. Phase Chemistries and organic compounds while let- Added concepts that expand the In IC, the most common sample ting target components pass through column selection option are the use preparation concept (after filtration) the cartridge. After the samples are of heart-cutting and two-dimensional is matrix elimination, which is the “prepared” they can be analyzed by heart-cutting. Heart-cutting takes concept of selectively removing matrix a variety of techniques including IC, advantage of column selectivity to species while flowing the target ana- high performance liquid chromatog- direct only the targets to a detector lytes into a vial or to a concentra- raphy (HPLC), ICP, gas chromatogra- such as a mass spectrometer. Two- tor column. The sample preparation phy (GC), nuclear magnetic resonance dimensional heart-cutting combines chemistries take advantage of differ- (NMR) spectroscopy, and hyphenated (usually) two column formats of two ences in chemical properties between techniques such as IC–MS and others. orthogonal chemistries to eliminate target species and matrix species to Matrix elimination products are the matrix while also concentrating accomplish matrix removal. The sta- available in several formats that allow the target analytes in one method (4). tionary phase of the sample prepara- off-line or in-line matrix removal of The present discussion picks up at tion device is designed for maximally species including chloride (halides), that point: What can one do when the selective retention for the matrix spe- sulfate, azo dyes, fats, transition need is beyond column selection? cies while exhibiting no retention metals (by several different selectiv- The most commonly used sample for the species passing through the ity phases), common anions, humic preparation technique for IC other resin bed. Table I includes the selec- acids (polyphenolics), general neu- than filtration is matrix elimination, tive retention principles for sample tral hydrophobic species, and others. with and without off-line or in-line preparation phases used in ion chro- These devices can be used alone or in concentration. In this article we review matography. The most common for- combination and can have other uses, two of the most commonly used tech- mat is an off-line cartridge with Luer including pH adjustment. Most of the niques for sample preparation in IC connectors for use with a syringe or available chemistries were developed and discuss the various formats of a vacuum manifold. The technique to solve specific application problems, their use. Exemplary applications are using this format is often referred such as the determination of nitrate used to illustrate the key concepts of to as solid-phase extraction, in this in a 1.6% NaCl brine solution or the data quality and ruggedness. case, of the matrix. Various station- removal of fat from milk to eliminate www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 29 Given: The sample contains 1% NaCl as matrix. The target is nitrite, at about 1 mg/L. Issue: Nitrite is eluted close to chloride, so at high concentration, the chloride peak swamps the nitrite peak, reducing its recovery and ren- dering quantification impossible. Solution: Nitrite does not precipi- tate with silver. A silver-form resin cartridge, such as the OnGuard II Ag cartridge (Thermo Scientific), can remove chloride from a sample matrix by precipitation of silver chloride in the resin bed. Calculations: 1% NaCl is 1 g per 100 g or 10 g/L NaCl. The chlo- ride portion of this is 35.45/58.45, 6.06 g/L, or about 0.17 mEq/mL. One OnGuard II Ag cartridge (1-cc) contains about 2.5 mEq of capac- ity. Dividing the cartridge capacity by the number of milliequivalents per milliliter of the sample, one can Figure 1: Determination of nitrite in brine with the aid of matrix elimination sample theoretically treat about 14 mL of preparation. Column: IonPac AG/AS15, 4 mm; eluent: 23 mM KOH; flow rate: 1 mL/ 1% NaCl. The cleanest sample is min; sample preparation: InGuard Ag and Na; injection volume: 100 μL; detection: suppressed conductivity, ASRS 300. Chromatogram 1: 1.6% NaCl brine after sample generally obtained by using about preparation; chromatogram 2: water blank; chromatogram 3: standard of 2 ppm ni- 20% less than the maximum capac- trite, sulfate, nitrate in 1.6% brine after sample preparation. Peaks: 1 = chloride, ity. A white precipitate is visible in 2 = nitrite, 3 = carbonate, 4 = nitrate, 5 = sulfate. the resin bed and can guide the ana- lyst in the determination of break- through volume. This cartridge is used in series with a sodium-form resin cartridge (such as the OnGuard V M II Na cartridge, Thermo Scientific) X so that any silver ion from dissolved AgCl can be trapped on the sodium- form resin without any pH change. As a comparison, if the silver-form resin is followed by a hydrogen-form V B resin, then any trapped silver ion dis-
Concentration X places hydronium ion, thus lowering pH and causing oxidation of nitrite to nitrate. Figure 1 shows an overlay of chromatograms showing 2 mg/L nitrite, carbonate, nitrate, and sulfate Sample volume in 1.6% NaCl brine, the standard, and the blank. Figure 2: A typical breakthrough curve for a sample preparation cartridge. An alternative experimental approach to measure capacity is to detector electrode fouling. Table I is treat the required sample, in terms of determine the “breakthrough vol- a summary of the commercially avail- concentration and volume. The overall ume” of a stationary phase–sample able chemistries and the off-line prod- goal is to determine if the sample prepa- combination. A solution of sample ucts that use them. ration cartridge has enough capacity to of known concentration is passed treat the required volume of sample so through the cartridge and the output Practical Calculations that the matrix is actually removed from of the cartridge is collected in frac- Use of matrix elimination requires a few the sample. Following is an example of a tions for analysis or monitored with calculations to ensure that the device can possible sample: a detector that can measure matrix 30 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com line until the original concentration of sample is being eluted continu-
ously (VM). By knowing the volume of sample passed through the car- tridge or the flow rate and the initial (a) sample–matrix concentration, one can determine the loading capacity of the cartridge resin and thus make sure that one doesn’t inject a sample (b) concentration that will exceed this 25 9 23 value (maximum sampling volume or
18 22 24 calculated mass). Similarly to the cal- 2 19 20 21 3 11 12 13 14 151617 (c) 1 468 culation approximation, one should back off on the matrix mass by about 02 46 8 1012141618202224262830323436384042 20% to take into account matrix Time (min) variability and to obtain the cleanest sample. Figure 3: Chromatograms of a dried distillers grains with solubles (DDGS) sample showing improvements in peak shapes and recovery of inositol phosphates. Off-Line and In-Line Chromatograms: (a) without treatment, (b) with OnGuard II RP only, and (c) with OnGuard II RP and Ag/H. Columns: Dionex CarboPac PA100 Guard and Analytical Matrix Elimination (250 mm × 4 mm); eluent: 25–500 mM HCl; flow rate: 1 mL/min; temperature: The most common need for matrix 30 °C; postcolumn derivatization; ferric nitrate in perchloric acid; detection: UV removal arises from the presence of absorbance at 290 nm. Peaks: 1–4 = InsP2, 5–13 = InsP3, 14–20 = InsP4, 21–24 = sample matrix components that inter- InsP5, 25 = InsP6. fere with the elution of the targets. Such interference can mean that matrix components are co-eluted with the target or otherwise change the retention, elution, or recovery of the target. Off-line matrix removal sample preparation has been applied to a wide variety of sample types (5–21). Figure 3 demonstrates a recent study on the determination of inositol phosphates in dried distillers grains with solubles (DDGS). Because the eluent is HCl, chloride needed to be removed from the sample so that the early elutions could occur only from the eluent, and not be affected by the chloride in the sample matrix. The off-line, two-layer silver–hydro- nium cartridge (OnGuard II Ag/H, Thermo Fisher Scientific) contains silver-form resin to remove chloride and also has a layer of hydronium- form cation-exchange resin at the outlet to trap any breakthrough Ag+ ion that might redissolve. The sam- ple also contains hydrophobic species Figure 4: Example analytical setup for use of in-line sample preparation fol- lowed by preconcentration of the target analytes. The water for the load- that are co-eluted with the targets. A ing pump should be cleaned using an electrolytic water purifier for minimal reversed-phase resin (OnGuard II RP, background. Thermo Scientific) removed those species (22). The cartridges were components. Figure 2 shows a typi- retained by the resin until all sites are used in series with the reversed-phase cal breakthrough curve. Initially, occupied. Then, the sample matrix cartridge followed by the two-layer as the sample is passed through the components begin to be eluted (VB), silver–hydronium cartridge. Recov- cartridge, matrix components are resulting in an increase in the base- ery of phytate was monitored at every www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 31
Table II: Retention time of polyphosphates in shrimp at the first and 300th injection, with and without using a sample preparation cartridge
Retention Time Without Cartridge (min) Retention Time With Cartridge (min) Analyte First Sample Injection 300th Sample Injection First Sample Injection 300th Sample Injection
Orthophosphate 2.74 2.50 3.34 3.32
Citrate 3.55 3.12 4.18 4.14
Pyrophosphate 6.10 5.78 6.66 6.63
Trimetaphosphate 7.10 6.66 7.61 7.55
Triphosphate 9.77 9.13 10.32 10.28
of in-line sample preparation. It is important to note that for low level determinations of common ions, the loading pump is usually supplied with electrolytically purified water (Electrolytic Water Purifier, Thermo Scientific or CIRA, Trovion Pte Ltd. Singapore) to minimize ions in the blanks from trace contamination in the loading water. Figure 5 shows the results of 200 injections of 10 ppb bromate in a matrix of 300 ppm chlo- ride using in-line sample preparation. The relative standard deviation for retention time on 200 injections was 0.062%, the bromate peak area was 0.309%, and the amount (parts per billion), was 0.309%. A general application of in-line matrix removal in the food industry is the use of a hydrophilic styrene– divinylbenzene cartridge (InGuard HRP, Thermo Scientific) for the in- line removal of hydrophobic matrix components. Hydrophobic matrix components are usually unidentified Figure 5: Comparison of lifetime chromatograms for 10 ppb bromate in 300 but can be retained on reversed-phase ppm chloride using sample preparation. Columns: IonPac AG24/AS24 (250 mm × 2 mm); eluent: 15–50 mM KOH gradient; flow rate: 0.3 mL/min; suppressor: stationary phases. One example is the ASRS 300 (2 mm); concentrator: TAC-ULP1; injection volume: 500 μL; sample determination of polyphosphates in preparation: InGuard Ag/InGuard H; sample loading: deionized water, 0.5 mL/ shrimp (24). The goal was to prolong min, 6 min. column life as evidenced by stable retention times over hundreds of step in the method development, the advantage of requiring much less injections. Table II shows the reten- and the overall loss on recovery was sample, as only the injection vol- tion time data for the target poly- 4.5% using this order of cartridges. ume is treated, and can therefore be phosphate species found in shrimp, Interestingly, the loss on recovery used repeatedly until the capacity comparing the first and 300th injec- using the reverse order of cartridges is depleted. A concentrator column tion, with and without the use of in- was 22%. is used to “recollect” the targets for line matrix removal. An in-line version of sample prepa- injection onto the analytical column. ration chemistries was introduced in This arrangement minimizes opera- Concentrators the last couple of years, and its accep- tor time and saves money. Use of concentrators can lower detec- tance is growing (23). Use of in-line Figure 4 shows a flow diagram for tion limits by several orders of magni- sample preparation chemistries has one configuration (of many) for use tude depending on the sample volume. 32 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com
Critical parameters include sample (5) Higher Resolution Separation of Organic Rosanne Slingsby volume, capacity of the concentrator Acids and Common Inorganic Anions in is a Senior Princi- pal Scientist in the stationary phase, and composition Wine, Application Note 273 (Thermo Research and Devel- of the sample matrix. Concentrators Scientific, Sunnyvale, California). opment Department are used when the concentration of (6) Determination of Melamine in Milk by at Thermo Fisher Sci- the target in the sample is too low Ion Chromatography with UV Detec- entific in Sunnyvale, California. She has for good quantification or when the tion, Application Note 231 (Thermo Sci- worked in R&D for sample band entering the analytical entific, Sunnyvale, California). Dionex Corporation, now part of Thermo column is diffused, as in the case of (7) T. Phesatacha, S. Tukkeeree, and J. Fisher, for 31 years, developing a wide vari- some applications of in-line sample Rohrer, Determination of Phytic Acid ety of products including the OnGuard and InGuard sample preparation cartridges. She preparation. Some of the most com- in Soybeans and Black Sesame Seeds, is a co-author on two U.S. EPA methods for mon (and successful) uses of concen- Application Note 295 (Thermo Scien- the determination of perchlorate and halo- trators are in the determination of tific, Sunnyvale, California). acetic acids by IC-MS and IC-MS-MS. Direct ultratrace targets in ultrapure water. (8) Y.S Ding and S.F. Mou, Chinese J. correspondence to: Rosanne.Slingsby@ther- mofisher.com In this case, the matrix to be elimi- Chrom. (Zhongguo hua xue hui) 20(3), nated is the water. Sometimes guard 262–264 (2002). columns can be used as concentrators (9) Z.-X Guo et al., J. Anal. Atomic Spec. because the selectivity matches the 18(11), 1396–1399 (2003). Kassandra Oates analytical column. However, specifi- (10) M.-L Jao et al., J. Food Drug Anal. 4(4) received a B.S. in Biochemistry and cally designed concentrator columns 349–358 (1996). Molecular Biology are available with important features (11) S.D. Kumar, B. Maiti, and P.K. Mathur, from the University including ultralow back pressure for Talanta 53(4), 701–705 (2001). of California in Santa compatibility with an autosampler. (12) X. Liu et al., Lizi Jiaohuan Yu Xifu/ Cruz, and has over 10 years of experience as Some concentrators have ultralow Ion Exch. Adsorption 23(6), 559–563 an organic and ana- background of some analytes, such (2007). lytical chemist. Prior to working at Thermo as sulfate (25). So once again, choices (13) Y. Liu and S. Mou. Talanta 60(6), 1205– Fisher Scientific, Kassandra held positions at save time and money. 1213 (2003). Elan Pharmaceuticals, Pickering Laborato- ries, Theravance Pharmaceuticals, and the (14) Y. Liu, S. Mou, and D. Chen, J. Chrom. Drug Enforcement Administration. Through Summary A 1039(1-2) 89–95 (2004). these experiences she gained considerable In this short review we have tried to (15) Y. Liu, S. Mou, and S. Heberling. J. experience using a variety of analytical cover the most commonly used sample Chrom. A 956(1–2), 85–91 (2002). instrumentations, including HPLC, NMR, GC, CE, IR, MS, and other related techniques. As preparation techniques for ion chro- (16) B. Savary et al. Rev. Sci. Eau 13(2), 139– an Applications Chemist at Thermo Fisher matography. The reference list pro- 154 (2000). Scientific, Kassandra specializes in producing vides many details that were outside (17) W. Shotyk, J. Chrom. 640(1-2), 309–316 high quality applications documents to IC the scope of this article. We began (1993). customers. with the concept of choosing the best (18) W. Shotyk, I. Immenhauser-Potthast, analytical column for the job; then, and H.A. Vogel, J. Chrom. A 706(1-2), by understanding the chemistry issues 209–213 (1995). Dong-bum Lee associated with the sample, chose the (19) J.E. Van Nuwenborg, D. Stöckl, and is an assistant techni- best off-line or in-line matrix elimina- L.M. Thienpont, J. Chrom. A 770(1–2), cal support manager tion solution. 137–141 (1997). of the LC–IC–SP team (20) F. Wang et al., Aust. J. Chem. 57(10), in the Chromatog- raphy & Mass Spec- References 1005–1010 (2004). trometry Division of (1) R. Slingsby and R. Kiser, Trends Anal. (21) F.F. Zhang et al., Chinese J. Chrom./ Thermo Fisher Sci- Chem. 20, 288–295 (2001). Zhongguo hua xue hui 20(5), 452–455 entific. He obtained (2) A. Seubert, W. Frenzel, H. Schafer, G. (2002). a Master’s degree at Yong-in University in 2000 in Air Pollution. Dongbum worked Bogenschutz, and J. Schaffer, Sample (22) Application note, Thermo Scientific for four years at the Korean Association of Preparation Techniques for Ion Chro- 2013, in press. Industrial Health and joined Dionex Korea in matography, Metrohm Monograph, (23) W. He et al., Chinese J. Chrom. (Se Pu) 2004 as a chemist. ◾ www.metrohm.com 30(4), 340–344 (2012). (3) L. Bhattacharyya and J.S. Rohrer, Eds., (24) C. Chantarasukon, S. Tukkeeree, and J. Applications of Ion Chromatography in the Rohrer, Determination of Mono-, Di- Analysis of Pharmaceutical and Biological and Triphosphates and Citrate in Shrimp Products (John Wiley & Sons, February by Ion Chromatography, Application For more information on this topic, 2012), p. 207. Note 1007, (Thermo Scientific, Sunny- please visit (4) R. Lin, K. Srinivasan, and C. Pohl, vale, California). www.chromatographyonline.com LCGC Europe 24(6), 322–332 (25) Concentrator and Trap Columns, (2011). Thermo Fisher Dionex LPN 0913-16. www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 33
Tools for Simulation and Optimization of Separations in Ion Chromatography
Users of ion chromatography (IC) are faced with lengthy method Sulfate Chloride Fluoride Phosphate development times because of the relatively slow equilibration of IC Nitrite Bromate Chlorite Nitrate Bromide Chlorate columns to each new eluent. Method development becomes even more Simulated Chrbonate demanding when multistep elution profiles (containing sequential isocratic and gradient steps) are used. Given that modern IC instruments
Experimental are designed to generate these multistep elution profiles, computer-
0 3 6 9 12 15 18 based methods to simulate separations for a wide range of analytes, Retention time (min) columns, and eluents are highly desirable. The tools available for computer-assisted method development are discussed here.
on chromatography (IC) is used widely the use of computational tools to facilitate Ifor the detection and determination the method development process. of ionic species in various samples. Considerable work has been under- With the advent of eluent generators and taken to understand the retention pro- electrolytic suppressors, IC is now oper- cesses that apply in IC. Over the years, ated routinely using both isocratic elution there have been numerous attempts to and multistep eluent profiles comprising produce mathematical retention models sequential isocratic and gradient steps. for isocratic elution (1–5), gradient elution Frequently, an eluent profile consisting (6), and multistep elution profiles (com- of multiple steps (commonly up to five) prising sequential isocratic and gradient is applied, typically when the separation steps) (7–10) in IC. These models aim is difficult or the number of analytes to to provide a mathematical relationship be separated is large. The practical imple- between the analyte retention factor and mentation of these multistep eluent pro- measurable properties of the analyte (such files is simplified greatly by the use of an as its charge), the eluent driving ion (such eluent generator in which the eluent is as its concentration and charge), and the generated electrolytically from an input stationary phase (such as its ion-exchange stream of water. The desired eluent con- capacity and phase ratio). Many of the centration profile can be created simply published models are remarkably accurate by applying a suitable profile of electrical in their prediction and in theory, they current to the eluent generator. could be used for the a priori calculation The development of a new IC separa- of retention factor for any analyte without tion method involves many decisions. the need for experimentation, provided Which column should be used? Which that values for all of the above properties eluent type should be used and what is are known. This situation rarely occurs Boon K. Ng, Greg W. Dicinoski, the optimal eluent composition? As with in practice because the process of deter- Robert A. Shellie, and Paul R. most chromatographic techniques, the mining values for all the necessary ana- Haddad development of a new IC method can be a lyte, eluent, and stationary-phase prop- Australian Centre for Research on time-consuming process. Trial-and-error erties would normally take longer than Separation Science (ACROSS), School of optimization of the separation is therefore a straightforward trial-and-error manual Chemistry, University of Tasmania, Australia Direct correspondence to: paul.haddad@ both tedious and challenging. For these optimization. For this reason, computer- utas.edu.au reasons, there has been a strong interest in assisted method development in IC 34 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com as constants and thus the model can be simplified to Prediction of retention times for entire search area log k = a – b log[E y–] [2]
where a and b are both constants. Limited Analyte, eluent, Nonlinear A plot of log k versus log[Ey-] will experimental stationary phase Model regression give rise to a linear relationship with the properties data effective charge of the analyte relative to the competing ion as the slope, b, and a, Figure 1: Schematic overview of computer-assisted optimization in IC. the intercept, indicating the degree of interaction between analyte and sta- tionary phase. This linear relationship is illustrated in Figure 2 for a univalent 1.4 eluent ion (OH-) and a series of inorganic and organic univalent analyte ions.
- Retention modeling for gradient elu- SO3 0.9 tion IC or for separations involving multistep elution profiles is understand-
- SO3 ably more difficult than is the case for 0.4 isocratic separations. Generally, two
k approaches can be used. In the first, a NO - 3 mathematical gradient retention model Log is derived, but this generally involves -0.1 quite complex derivations, requires accu- Br- COO- rate gradient retention data to solve the model, and can also be challenging com- -0.6 - putationally when calculations of reten- NO2 - BrO3 tion factors under numerous eluent con- IO - ditions are undertaken. The second, and -1.1 3 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 preferred, approach is to break up the Log [OH-] gradient or multistep elution profile into a series of small, successive isocratic seg- Figure 2: Plot of log k versus log [OH-] for inorganic and organic anions. ments, to then apply an isocratic reten- tion model to each segment, and finally almost invariably involves a combination it is pertinent to show which models are to integrate the individual segments into of experimentation and computer predic- most useful for the process outlined in Fig- the overall elution profile. This process tion. The experimentation step is done to ure 1. It is also often the case that the most is illustrated schematically in Figure 3. acquire some limited retention data using, useful models are surprisingly simple. The chief advantages of this approach for example, three eluent conditions, and Isocratic elution in IC is described are that only isocratic retention data are then applying the chosen mathemati- accurately by the following relationship: required (because an isocratic retention cal retention model to calculate analyte ( model is used), the same model can be 1 x Q retention for all possible eluent conditions. log k = log (K )+ log used for both isocratic and gradient steps
A y A,E y ( y This process is represented schematically w ( x of an elution profile, and computational E y– in Figure 1. This process shows that, pro- + log V + y log [ ] [1] demands are less for isocratic models. vided the retention model is accurate and ( m k k sufficient experimental data are available where A is the retention factor, A,E is ion- Simulation and Optimization to permit reliable implementation of the exchange selectivity coefficient between of Isocratic Separations in IC model, it should be possible to quickly the analyte and the eluent competing It is clear from Figure 2 that the isocratic produce a retention map for all analytes ion, x is the charge of the analyte, y is the retention model shown in equation 2 gives over all eluent combinations (the “search charge on the eluent, Q is the effective ion- a very accurate description of retention area”), leading to straightforward simula- exchange capacity of the stationary phase, behavior. It is also evident from Figure 2 w V tion and optimization of separations. is the mass of the stationary phase, m is that the retention data for any of the ana- the volume of the eluent species, and [Ey-] lytes shown could be predicted if a mini- Retention Models in IC is the concentration of the eluent. mum of two experimental data points It is not the intention of this article to If this model is employed for isocratic were used to calculate the intercept (a in discuss in any detail the theoretical basis separations consisting of a single compet- equation 2) and slope (b in equation 2) of k Q, w V underlying IC retention models. However, ing ion, A,E, , and m can be treated the plots shown in Figure 2. Thus, when www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 35 assign a numerical value to an entire chro- matogram, according to its alignment 30 30 with a desired goal. For example, this goal 25 25 might be to maximize the separation of
20 20 all peaks, to maximize the separation of a problematic pair of peaks, or to achieve 15 Segmented to 15 the separation in the fastest possible time. 10 10 Such an algorithm is referred to as an opti- 5 5 mization criterion, BOENBOZTVDIBMHP- 0 0 Operating concentration (mM) 0.00 0.25 0.50 0.75 1.00 Operating concentration (mM) 0.00 0.25 0.50 0.75 1.00 rithms have been proposed. Time (min) Time (min) It is clear from the above comments that reliable isocratic retention data are essential 30 30 for the successful implementation of simu- 25 25 lation and optimization of isocratic separa- 20 20 tions in IC. Of course, these retention data 15 Segmented to 15 can be acquired as needed for particular 10 10 DPNCJOBUJPOTPGBOBMZUFT FMVFOUT BOETUB- 5 5 UJPOBSZQIBTFT"OBMUFSOBUJWFBQQSPBDIJT 0 0
Operating concentration (mM) to compile an extensive set of reliable and Operating concentration (mM) 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 Retention time (min) Retention time (min) accurate isocratic retention data for a wide SBOHF PG BOBMZUFT FMVFOUT BOE DPMVNOT Figure 3: Schematic overview of segmenting gradients and multistep profiles to isocratic segments. BOEUPTUPSFUIFTFEBUBJOBTVJUBCMZBDDFT- TJCMF EBUBCBTF *O UIJT XBZ VTFST DPVME TJNVMBUF BOE PQUJNJ[F TFQBSBUJPOT CZ accessing the database, without the need UP VOEFSUBLF BOZ FYQFSJNFOUBUJPO 5IF BVUIPST IBWF XPSLFE XJUI 5IFSNP 4DJ- entific Dionex (hereinafter referred to as %JPOFY PWFSUIFQBTUZFBSTUPDPNQJMF such a database and this has formed the basis of the Virtual Column software for IC simulation and optimization, marketed CZ%JPOFY5IFEBUBCBTFVOEFSMZJOHUIJT software contains retention data for more UIBOBOBMZUFT BOJPOTBOEDBUJPOT PO more than 20 Dionex IC columns using a range of eluents, column diameters, BOEUFNQFSBUVSFT5IFVTFSTJNQMZOFFET UPTFMFDUUIFBOBMZUFTUIBUBSFUPCFTFQB- rated and the software can then be used to simulate all possible chromatograms on BMM DPMVNOT BOE UP JEFOUJGZ UIF PQUJNBM combination of eluent and column for the desired separation. One example of the output (in its sim- QMFTUGPSN HFOFSBUFECZUIF7JSUVBM$PM- Figure 4: Screen display from Virtual Column software for isocratic separations. umn software is shown in Figure 4. In this case, fluoride, chloride, bromide, sulfate, applied to the simulation of isocratic IC t 5IFa and b values can then be used iodide, and phosphate are being separated separations, the process depicted in Figure to calculate the retention factors for POB%JPOFY"4DPMVNOVTJOHIZESPY- 1 would have the following steps: FBDI BOBMZUF BU BOZ JTPDSBUJD FMVFOU JEFFMVFOUTJOUIFSBOHFoN.5IF t 3FUFOUJPOGBDUPSTGPSBMMBOBMZUFTJOUIF DPODFOUSBUJPO*OUIJTXBZ UIFDISP- top panel shows a plot of the optimiza- test mixture would be determined using NBUPHSBNTPCUBJOBCMFGPSBOZJTPDSBUJD tion criterion (in this case, the normal- at least two eluent concentrations. elution composition can be simulated. ized resolution product, which reaches t 5IFTFSFUFOUJPOEBUBXPVMECFVTFE "GUFSUIFBCJMJUZUPQSFEJDUSFUFOUJPOIBT a maximum value of 1.0 when all peaks UPTPMWFFRVBUJPOGPSFBDIBOBMZUFUP been achieved, it is then a simple further BSFTQBDFEFWFOMZPWFSUIFDISPNBUPHSBN obtain a and b values for the retention TUFQUPJODMVEFPQUJNJ[BUJPO5IJTJOWPMWFT BOE IBT B WBMVF PG [FSP XIFO BOZ UXP QMPUGPSFBDIBOBMZUF using a mathematical algorithm that can peaks are coeluted). It can be seen that 36 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com isocratic and gradient steps can be gener- ated, and the versatility that they offer in developing new separations, such profiles are now the preferred mode of operation in Sulfate modern IC. However, the selection of an Chloride
Fluoride optimal eluent profile is a very challenging Phosphate and time-consuming task if trial-and-error Nitrite Bromate Chlorite
Nitrate procedures are used. For this reason, there Bromide Chlorate
Simulated Chrbonate has been very strong interest in extending the isocratic approaches described above for use with multistep elution profiles. The earlier discussion of retention mod- els highlighted the potential use of iso- cratic retention data for gradient elution by dividing the gradient into successive small isocratic segments. If this approach Experimental is taken, only isocratic retention data are needed for calculation of retention fac- 0 3 6 9 12 15 18 tors, so the same isocratic database used Retention time (min) for the Virtual Column isocratic software can be applied. Simulation performed in Figure 5: Simulated and experimental separations on a Dionex AS19 column under a three-step gradient profile consisting of: 3.75 mM KOH for 0.8 min, 3.75 to 33.75 mM this way is highly successful, as evidenced KOH for 10 min, followed by 33.75 to 99.75 mM KOH for 4 min at 1 mL/min at 30 °C. by Figure 5, which shows the simulated and experimental chromatograms for the separation of 11 anions on a Dionex AS19 column using a three-step elution 35 profile. Figure 6 shows the overall correla- y = 0.9866x tion between experimental and simulated 30 R2 = 0.991 retention times for 37 analytes (anions and cations) on several columns using 25 four different five-step elution profiles. Figures 5 and 6 show that accurate 20 simulation of retention under multistep 15 elution profiles is possible. However, optimizing the separation by finding the 10 optimal elution profile is an exceedingly difficult task because of the large number of parameters that need to be optimized. Observed retention time (min) 5 For example, in a simple three-step elu- 0 tion profile comprising an isocratic step, 01020 30 40 a gradient step, and then a final isocratic Simulated retention time (min) step, it is necessary to optimize the initial eluent concentration, the duration of the Figure 6: Correlation of experimental and simulated retention times for 24 anions first step, the slope and duration of the and 13 cations on Dionex AS11 HC, AS16, AS19, CS12A, and CS16 columns for four 5-step complex eluent profiles. gradient step, and the duration of the final isocratic step. The complexity increases the maximum value of this criterion is optimization criteria. In addition, two- when more steps are included in the eluent reached when the eluent concentration is component eluents (such as mixtures of profile. Currently, there are no published 7.65 mM. The bottom panel displays the carbonate and bicarbonate) can be simu- methods for this type of optimization. chromatogram obtained with this eluent lated, in which case the top panel reverts However, it is possible to simulate the concentration. There is a slider bar at the to a response surface plotted against the separation using the approach depicted in bottom of the top panel and the user can concentrations of the two eluent compo- Figure 7. Here, a four-step elution profile slide this to any desired eluent concentra- nents on the horizontal and vertical axes. is depicted, with an initial isocratic step, a tion and the chromatogram obtainable at shallow gradient, a steeper gradient, and a this concentration will appear in the bot- Simulation and Optimization of final isocratic step. The square boxes rep- tom panel. The software also allows other Multistep Elution Profiles in IC resent points where the shape of the profile columns, diameters, and temperatures to Because of the ease with which multi- can be adjusted on a computer screen. For be investigated, as well as the use of other step eluent profiles containing sequential example, the length of the first isocratic www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 37
Robert A. Shellie is an associate profes- 60 sor at the University of Tasmania, where he is a member of the 50 Australian Research Centre on Separation 40 Science (ACROSS). Hav- ing worked in the area 30 of hyphenated techniques in chromatog- raphy for more than a decade, his research 20 has largely focused on the exploration of approaches for separation and characteriza-
Eluent concentration (mM) 10 tion of complex multicomponent samples. He has been actively engaged in the develop- 0 ment and application of retention models for 0105 15 20 25 rapid method development since 2006. Time (min) Boon K. Ng Figure 7: Screen display for manipulation of a multistep elution profile. is a Science and Endow- ment Industry Funded step can be adjusted by moving the first isocratic data. The computational simplic- Postdoctoral Fellow at box left or right. Any change made to the ity of these calculations means that real- the University of New profile will lead to the instant simulation time simulations can be performed. His- South Wales, Australia. His PhD focused on of a new chromatogram by the software. torical isocratic retention databases can the development of Users can thereby simulate any desired be easily updated with a small amount of computer-assisted eluent profile and can optimize the sepa- experimentation. These new tools are not optimization tools for method development ration manually without experimentation. yet available in the commercial Virtual in ion chromatography. His current research interests are in ion chromatography and coor- Column software, but work is currently dination chemistry. Updating the Retention Database under way to update this software. An obvious potential limitation of the Greg W. Dicinoski above approaches to simulation and References is an associate profes- sor and the Head of optimization is that they rely on reten- (1) J. Havel, J.E. Madden, and P.R. Haddad, the School of Chemistry Chromatographia tion data stored in the Virtual Column 49, 481–488 (1999). and the Deputy Direc- database, which may become dated. For (2) P.R. Haddad and R.C. Foley, J. Chromatogr. tor of the Australian example, a new column might not behave 500, 301–312 (1990). Centre for Research on Separation Science. He in the same manner as the column used (3) J.E. Madden and P.R. Haddad, J. Chro- leads projects relating matogr. A to acquire the data some years previously. 829, 65–80 (1998). to counter-terrorism and pharmaceutical analy- We have addressed this issue by devising a (4) J.E. Madden and P.R. Haddad, J. Chro- sis. Specific focus is given to theoretical aspects, system where the database can be updated matogr. A 850, 29–41 (1999). including the modelling and optimization of retention in chromatographic systems and (or “ported”) to a new column by making (5) P. Hajós, O. Horváth, and V. Denke, Anal. mobility simulation in electrophoresis, and spe- Chem. adjustments to the stored data, based on 67, 434–441 (1995). cialist applications for the separation of target some actual measurements obtained on (6) J.E. Madden, N. Avdalovic, P.R. Haddad, chemical species using reversed-phase and ion the new column on which the IC method and J. Havel, J. Chromatogr. A 910, 173– chromatography, and capillary electrophoresis. is to be used (11). In this way, the database 179 (2001). Paul R. Haddad can be continually adapted for use with (7) R.A. Shellie, B.K. Ng, G.W. Dicinoski, is a Distinguished Pro- newer versions of the columns on which S.D.H. Poynter, J.W. O’Reilly, C.A. Pohl, fessor at the University the original data were obtained. and P.R. Haddad, Anal. Chem. 80, 2474– of Tasmania, Australia, 2482 (2008). and he is the Direc- tor of the Australian Conclusions (8) T. Bolanča, Š. Cerjan-Stefanović, M. Luša, Centre for Research The ability of modern IC instruments to M. Rogošić, and Š. Ukić, J. Chromatogr. A on Separation Science apply multistep elution profiles has placed 1121, 228–235 (2006). (ACROSS). He has very severe demands on method develop- (9) P.J. Zakaria, G.W. Dicinoski, M. Hanna- diverse interests spanning separation science and is the author of over 500 publications in J. Chromatogr. A ment procedures designed to exploit the Brown, and P.R. Haddad, this field. Method development in ion chro- full advantages of these elution profiles. 1217, 6069–6076 (2010). matography has been one of his strongest Retention models derived for simple iso- (10) P.J. Zakaria, G.W. Dicinoski, B.K. Ng, R.A. research interests. Please direct correspon- ◾ cratic IC separations can be applied to Shellie, and M. Hanna-Brown, P.R. Had- dence to: [email protected] multistep eluent profiles by segmenting dad, J. Chromatogr. A 1216, 6600–6610 these profiles into small isocratic steps. (2009). For more information on this topic, This approach simplifies the retention (11)B.K. Ng, R.A. Shellie, G.W. Dicinoski, C. please visit modeling process and enables rapid cal- Bloomfield, Y. Liu, C.A. Pohl, and P.R. Had- www.chromatographyonline.com culation of retention factors based only on dad, J. Chromatogr. A 1218, 5512–5519 (2011). 38 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com
Advances in High-Speed and High-Resolution Ion Chromatography
10 Cl- Compared to modern high performance liquid chromatography (HPLC) and 8 ultrahigh-pressure liquid chromatography (UHPLC), separation speeds in ion
- NO2 6 chromatography (IC) have lagged behind. In recent years, dramatic strides
2- SO4 4 have been made by the development of PEEK flow components, eluent - F NO - - 3 3-
Conductivity (μS) PO Br 4 generators and membrane suppressors with increased pressure and lower 2 volumes, smaller particle size packings that permit shorter columns and 0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Time (min) higher flow rates, and capillary columns with higher efficiencies. Here, we review these developments and show examples of faster IC separations. For simple systems, separation speeds rival those of modern LC. For complex samples, separation times are still slow, but the use of 2D techniques allows more rapid analysis of complex samples in high ionic-strength matrices.
ust four decades ago, the analysis of an cles have enabled faster analysis by allowing Jindividual ion required that the analyst shorter column lengths while maintaining perform a time-consuming titration or high efficiency and resolution (5). Introduc- gravimetric measurement. Each additional tion of sub-2-μm particles required ultra- ion to be analyzed magnified the effort high-pressure (up to 15,000 psi) metallic required. In 1975, Small and coworkers (1) pumps because of the dramatically greater introduced ion chromatography (IC). This back pressure generated by these smaller technique enabled the simultaneous analysis particle diameters (5,6). of multiple ions with low detection limits. Why has IC lagged behind this trend? Compared to the hours of labor needed for First, IC was limited by the hardware (7). classical methods, IC yielded rapid measure- The stainless steel pumps, tubing, and ments. For instance, Li+, Na+, and K+ could fittings used in high performance and be baseline-resolved in 134 min (2). Simi- ultrahigh-pressure liquid chromatography larly common anions could be analyzed in (HPLC and UHPLC) are not compatible 20 min (1,2). with the highly corrosive acidic or alkaline Today, IC is widely used for the analy- eluents required in IC. Early IC systems sis of inorganic ions and many ionizable were constructed with glass columns and organic compounds. More than 70 com- polyoxymethylene (Delrin) components (8). mercial IC columns are available (3). Until The introduction of polyether ether ketone recently, however, developments by IC col- (PEEK) pumps and tubing increased the umn manufacturers focused on increasing pressure capabilities of IC to near that of the ion-exchange capacity, reliability, and conventional HPLC systems. Second, severe selectivity of stationary phases (4) rather baseline drift and ghost peaks were experi- than the speed of analysis. Thus, up to just enced with gradient elution in IC because Charles A. Lucy and a few years ago the typical separation time of the inherent background conductivity M. Farooq Wahab of anions differed little from that demon- of carbonate eluents and impurities in the Department of Chemistry, University of strated by Small and coworkers in 1975 (1,2) carbonate or hydroxide eluents. The gen- Alberta, Gunning/Lemieux Chemistry In contrast, in the last decade, rapid eration of ultrahigh-purity eluents such Centre, Edmonton, Alberta, Canada Direct correspondence to: charles.lucy@ual- developments have taken place in reversed- as carbonate-free sodium hydroxide was berta.ca and [email protected] phase liquid chromatography. Smaller parti- made possible through membrane-based www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 39 The efficiency is reduced from 9600 to 5000 because of the shorter column length, but it is still sufficient for baseline resolu- (a) tion. The lower back pressure of the shorter Cl- - 2- column also allows use of higher flow rates - NO SO F Br- 3 4 - while staying within the pressure limits of NO2 the IC instrumentation. Figure 1c shows a PO 3- 4 4.5-min separation of seven common inor- ganic anions using a 15-cm AS22-Fast at 2 mL/min. Using even shorter columns while keeping the particle size the same can further reduce the analysis time. For 0 1 2 3 4 5 6 7 8 9 10 11 12 instance, Haddad’s group used 3–5 cm col- (b) - - SO 2- umns packed with 7.5-μm resins to demon- Cl NO3 4 Br- strate 15-min separations of highly retained - F - NO2 perchlorate and thiocyanate ions (12). Such ions are typically retained for more than PO 3- 4 1 h on standard 25-cm IC columns. The next evolution in the speed of IC analysis was associated with the redesign of the commercial hardware for capillary col- umns (~0.4 mm i.d.). Such columns allow 0 1 2 3 4 5 6 7 8 9 10 11 12 the same linear velocity (u, cm/min) as a
- normal 4-mm i.d. column at a 100-fold - Br 2- Cl SO4 NO - lower volumetric flow rate (13), enabling (c) 3 - - F NO2 extended operation with a single bottle of eluent. An immediate consequence of the PO 3- 4 lower volumetric flow rates needed for cap- illary IC was a decrease in the surface area of the electrodialytic membrane in the elu- ent generator (14,15), which yielded a higher burst pressure for the membrane. Thus, a capillary IC system such as the Thermo 012 34 567 8 Scientific Dionex ICS-5000 has a pres- Time (min) sure limit of 5000 psi (compared to a limit of 3000 psi for a conventional IC system). Figure 1: Comparison of separations obtained using 25- and 15-cm columns with the Capillary separation systems also produce same particle size (6.5 μm) and column diameter: (a) 25 cm × 0.4 cm AS22 (flow rate: greater column efficiencies (reduced plate × × 1.2 mL/min), (b) 15 cm 0.4 cm AS22-Fast (flow rate 1.2 mL/min), (c) 15 cm 0.4 cm heights <2) than large-bore counterparts AS22-Fast (2.0 mL/min). The typical working pressure of AS22 and AS22-Fast is ~1900 psi. The excessive resolution is shown in blue boxes. Eluent: 4.5 mM sodium carbonate, (13). Figure 2 shows the separation of seven 1.4 mM sodium bicarbonate; injection volume: 10 μL; suppressed conductivity detec- common anions in less than 2 min using a tion. (Courtesy of Thermo Fisher Scientific.) 15-cm capillary column packed with 4-μm particles (16). At 25 μL/min, the back pres- electrodialytic eluent generators (9,10). The with 6.5-μm particles. Seven common sure was only 3480 psi. For the sake of com- fragility of the eluent generators restricted anions are separated in ~12 min with 9600 parison, this capillary separation in Figure the upper pressure limit to 3000 psi (11). plates at a column pressure of <2000 psi. 2 takes 17 s/ion compared to 98 s/ion for The maximum flow rate in IC was further However, there is actually excessive reso- the AS22-Fast column in Figure 1c. Thus, limited by the suppressor to 1–3 mL/min lution in Figure 1a, leading to “wasted” capillary systems have the potential of high- (for 2 mm to 4 mm columns, respectively) time, which is highlighted in blue in the speed IC separations by employing smaller to protect the membranes from leaking. figure. Clearly, some of this excess reso- particles in 15-cm capillaries. A number of Thus the standard IC hardware was pres- lution could be traded for analysis time such “fast” IC capillaries are commercially sure- and flow rate–limited as late as 2011. by shortening the column while keep- available for cations and anions. In January Because of these pressure and flow rate ing the particle size and other parameters 2013, a redesign of the eluent generators for limits, IC columns continued to be char- the same. The first such column was the conventional flow IC yielded the Thermo acterized by large particle sizes (7–13 μm) 15-cm IonPac AS22-Fast column, which Scientific Dionex ICS-5000+ system, which and long (20–25 cm) columns. For instance, was introduced at Pittcon 2010. Figure 1b also has a 5000-psi pressure limit (17). Figure 1a shows a typical ion chromatogram shows a 7-min separation on the AS22-Fast Given the ubiquitous presence of using a 25-cm IonPac AS22 column packed column — a 40% reduction in run time. UHPLC systems in the industry, 40 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com of sub-2-μm particles is not a trivial task. Even 4.4-μm charged polymeric particles 10 display unexpected behavior in the packing - Cl process (22). Third, the mechanical strength of PEEK limits the pressure of current IC 8 systems to <5000 psi. IC certainly has room to catch up with the efficiencies and speed - NO2 6 of UHPLC. In the above discussion, we provided 2- SO4 examples of fast separations of a few ions 4 in relatively simple samples by employing - F NO - smaller particles or very short columns, - 3 3-
Conductivity (μS) PO Br 4 or both. Speed becomes a secondary fac- 2 tor, however, when analyzing difficult and ill-characterized samples such as food, bio- 0 logical samples, or environmental waste. In 0.0 0.5 1.0 1.5 2.0 2.5 3.0 these cases, the target is to fully resolve all Time (min) the ions of interest. The factors that affect the resolution (R) of critical pair of analytes Figure 2: Fast separation of seven common anions on an IonPac AS18 capillary col- in a given separation are described by umn (150 mm × 0.4 mm) packed with 4-μm particles. Eluent: 35 mM OH- with a flow rate of 25 μL/min; operating pressure = 3480 psi. Suppressed conductivity detection. √N k α−1 R =( )( )( ) Adapted from reference 16. 4 1+k α [1] The plate number, N, is easily increased by using longer columns or smaller particles. The selectivity factor, α, and retention fac- SO 2- k, Cl- 4 tor, become important for high resolu- tion. Both of these variables are dependent IO - Br- 3 on the stationary phase chemistry and the eluent (4,23). Software packages for the NO - - 3 simulation and optimization of separations NO2 2- HPO4 are available (24). For instance, the Virtual Column Separation Simulator (Thermo Fisher Scientific) uses a database of experi- 0.2 μS mental retention data to predict retention using the linear solvent strength model– empirical approach. Resolution maps are also generated by the software for a critical pair of ions. Retention times have been pre- 0 10 20 30 40 dicted within 3% of the observed behavior Time (s) for 33 anionic and cationic analytes across Figure 3: Ultrafast separation of seven common anions on very short column (1.3 cm × a variety of column formats and complex 0.46 cm) with suppressed conductivity detection. Column: Cationic surfactant (didodecy- gradients (25). ldimethylammonium bromide) coated Extend-C18 column; flow rate: 2.0 mL/min; particle size: 1.8 μm; eluent: 2.5 mM 4-hydroxybenzoic acid (pH 10.1). Adapted from reference 20. Thus, by using specially designed high capacity columns and small particles, one chromatographers have begun to ask “Can achieving fast separations in commercial IC can achieve very high resolution. Figure 4 IC be as fast as UHPLC?” Although the columns, we have yet to see subminute IC illustrates a high-resolution separation of terms “fast” and “ultrafast” separations are separations in commercial columns. Tech- 44 inorganic and organic ions in 40 min relative, a survey of the literature and com- nology exists to synthesize 2-μm highly using a long 25-cm capillary and 4-μm mercial catalogs shows that these terms crosslinked polymer particles that can with- particles. This corresponds to a separation imply subminute separations. Subminute stand pressures of >5000 psi (21). However, speed of 53 s/ion. IC separations were pioneered by the groups there are still significant challenges in using However, in most real samples not all ions of Paull (18) and Lucy (19). For instance, smaller particles. First, as columns become are in similar concentration. For example, Figure 3 shows the ultrafast separation of more efficient, extracolumn band broaden- bromate is a human carcinogen whose con- seven common anions in just 40 s using ing effects come into play (20), and those centration is regulated by various govern- an ultrashort 1.3-cm column packed with effects are particularly challenging in IC ment environmental agencies around the 1.8-μm C18 silica particles (20). Although where postcolumn devices such as suppres- world (26). In IC, bromate is eluted early, significant developments have taken place in sors are essential. Second, optimum packing close to the chloride peak. In clean, low www.chromatographyonline.com APRIL 2013 ADVANCES IN ION CHROMATOGRAPHY 41 the upper trace of Figure 5 (27). The high concentration of the matrix ions overloads the analytical column. The resulting wide fronted or tailed peak broadens and shifts 7 the retention time of the trace (~10 ppb) bro- 15 40 mate peak (28). Eventually, the broadening 35 2 compromises accurate quantification and 5 31 22 29 degrades the resolution of bromate from the overloaded peak. 3334 36 32 One ingenious way to obtain high- 34 7 39 38 43 3 10 21 resolution separation in complex matrices 6 20 41 8 14 24 37 Conductivity (μS) 12 2728 is matrix elimination IC (27,29,30) based 9 11 26 5 17 18 23 30 44 1 13 19 16 on high-resolution two-dimensional IC 25 1 42 (2D IC). In this heart-cutting approach for bromate analysis (EPA Method 302.1), the bromate ion is separated in the first dimension using a high-capacity column 0 10 20 30 40 capable of handling the high-ionic-strength Time (min) matrix. A ~2 mL portion of the suppressed Figure 4: High-resolution gradient separation of 44 anions on a specially tailored high- eluent (the heart-cut portion) containing capacity capillary column (prototype IonPac AS11-HC, 250 mm × 0.4 mm, 4-μm particle t t the bromate ( 1 to 2 in Figure 5) is passed size). Eluent: KOH: 1–14 mM KOH in 16 min, 14–55 mM KOH in 24 min, 15 μL/min. Injec- through a concentrator column positioned tion volume: 0.4 μL. Peaks: 1 = quinate, 2 = fluoride, 3 = lactate, 4 = acetate, 5 = 2-hy- droxybutyrate, 6 = propionate, 7 = formate, 8 = butyrate, 9 = 2-hydroxyvalerate, 10 = in place of the sample loop of the injection pyruvate, 11 = isovalerate, 12 = chlorite, 13 = valerate, 14 = bromate, 15 = chloride, 16 = valve for the second-dimension separation. 2-oxovalerate, 17 = nitrite, 18 = ethylphosphonate, 19 = trifluoroacetate, 20 = azide, 21 The trapped analyte is eluted off the con- = bromide, 22 = nitrate, 23 = citramalate, 24 = malate, 25 = carbonate, 26 = malonate, centrator column into a second-dimension 27 = citraconitate, 28 = maleate, 29 = sulfate, 30 = alpha-ketoglutarate, 31 = oxalate, 32 = fumarate, 33 = oxaloacetate 34 = tungstate, 35 = molybdate, 36 = phosphate, 37 = IC column of different selectivity than the phthalate, 38 = arsenate, 39 = citrate, 40 = chromate, 41 = iso-citrate, 42 = cis-aconitate, first-dimension column. For high-resolution 43 = trans-aconitate, 44 = iodide. (Courtesy of Thermo Fisher Scientific.) applications, a second-dimension column with a lower cross-sectional area is used to enhance sensitivity (31). Figure 6 shows the second-dimension high-resolution separa- tion of trace levels (15 ppb) of bromate in both reagent and a high-ionic-strength High ionic strength water sample. This heart-cutting matrix t water sample elimination approach is generic and can eas- R ily be applied on IC samples that give poor resolution of trace components as a result of - matrix overload. For instance, perchlorate BrO3 peak has similarly been resolved in highly com- Reagent water plex water samples (31). Research is ongo- sample
Conductivity (μS) ing on comprehensive 2D IC for complete analysis of complex ionic samples (32).
Conclusions t t 1 2 Until recently, the speed of IC was limited Heart-cut portion by the pressure and flow rate limits of the IC 0246810 12 14 16 18 20 22 24 26 28 30 32 hardware. These conditions required the use Time (min) of large particles and long columns. Numer- ous recent advances have enabled faster IC Figure 5: First-dimension analysis of a bromate peak in reagent water and high ionic strength water. The bromate concentration is 15 μg/L. The bromate retention time separations by reducing the column length
(tR) is 9.2 min. The start of the cut window (t1) is 8.0 min and the end of the cut win- while using large particles, or by reducing × dow (t2) is 9.8 min. Column: 25 cm 0.4 cm IonPac AS19; eluent: 10 mM KOH, step the particle size to 4 μm. However, com- changed to 65 mM KOH following the elution of bromate at 1.0 mL/min; injection mercial IC has not yet achieved the ultra- volume: 1.0 mL. (Courtesy of Thermo Fisher Scientific.) fast subminute separations that have been ionic strength samples, baseline resolution samples, such as seawater or wastewater, demonstrated by sub-2-μm particles in very is satisfactory, as shown in the lower trace the resolution of bromate with the adjacent short columns. Thus, further advances in IC in Figure 5. In high-ionic-strength water peak is severely compromised, as shown in speed for simple samples are anticipated. 42 ADVANCES IN ION CHROMATOGRAPHY APRIL 2013 www.chromatographyonline.com
(21) K. Li and H.D.H. Stover, J. Poly. Sci. Part A: Polym. Chem. 31, 3257–3263 (1993). (22) M.F. Wahab, C.A. Pohl, and C.A. Lucy, J. t R Chromatogr. A 1270, 139-146 (2012). - High-resolution BrO3 (23) C. Liang and C.A. Lucy, J. Chromatogr. A separation in the second 1217, 8154–8160 (2010). dimension by the heart-cut approach (24) J.E. Madden, M.J. Shaw, G.W. Dicinoski, N. Avdalovic, and P.R. Haddad, Anal. Chem. 74, 6023–6030 (2002). (25) B.K. Ng, R.A. Shellie, G.W. Dicinoski, C. Bloomfield, Y. Liu, C.A. Pohl, and P.R. Had- dad, J. Chromatogr. A 1218, 5512–5519 (2011). High ionic strength (26) R.J. Joyce and H.S. Dhillon, J. Chromatogr. A Conductivity (μS) water sample 671, 165–171 (1994). (27) R. Lin, K. Srinivasan, and C. Pohl, LCGC Reagent water Europe sample 24(6), 322–332 (2011). (28) H.P. Wagner, B.V. Pepich, K.Srinivasan, C. Pohl, B. De Borba, R. Lin, and D.J. Munch, 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 EPA Method 302.0 (2009). Time (min) (29) Ayushi, S.D. Kumar, and A.V.R. Reddy, J. Figure 6: High-resolution separation of bromate ion in second-dimensional analysis of Chromatogr. Sci. 50, 81–83 (2012). 15 μg/L bromate in reagent water and high ionic water matrix. Column: lonPac AS24 ana- (30) P. Zakaria, C. Bloomfield, R.A. Shellie, P.R. × × lytical column (25 cm 0.2 cm) and AG24 guard column (5 cm 0.2 cm) were used for the Haddad, and G.W. Dicinoski, J. Chromatogr. A second dimension analysis; eluent: isocratic 10 mM KOH changed to 65 mM KOH following the elution of bromate for approximately 10 min and re-equilibrate at 10 mM hydroxide 1218, 9080–9085 (2011). before injection; eluent flow: 0.25 mL/min. (Courtesy of Thermo Fisher Scientific.) (31) H.P. Wagner, B.V. Pepich, C. Pohl, D. Later, R. Joyce, K. Srinivasan, B. De Borba, D. Thomas, For complex matrices, the enhanced pres- (8) C.A. Lucy, J. Chromatogr. A 1000, 711–724 A. Woodruff, and D.J. Munch, EPA Method sure capabilities of modern IC are enabling (2003). 314.1 (2005). greater one-dimensional peak capacities. (9) D.L. Strong, P.K. Dasgupta, K. Friedman, and (32) S.S. Brudin, R.A. Shellie, P.R. Haddad, and Commercial equipment has made targeted J.R. Stillian, Anal. Chem. 63, 480–486 (1991). P.J. Schoenmakers, J. Chromatogr. A 1217, two-dimensional IC separations routine. (10) D.L. Strong, C.U. Joung, and P.K. Dasgupta, 6742–6746 (2010). J. Chromatogr. 546, 159–173 (1991). Acknowledgments (11) Y. Liu, K. Srinivasan, C. Pohl, and N. Charles A. Lucy is a professor of J. Biochem. Biophys. Methods The authors are grateful to Chris Pohl, Avdalovic, 60, analytical chemistry Kannan Srinivasan, and Rong Lin (Dionex, 205–232 (2004). at the University of Thermo Fisher Scientific, California) for (12) É. Tyrrell, R.A. Shellie, E.F. Hilder, C.A. Pohl, Alberta and is on the providing data for this review. Funding from and P.R. Haddad, J. Chromatogr. A 1216, Scientific Committee of the International Ion Natural Sciences and Engineering Research 8512–8517 (2009). Chromatography Sym- Council of Canada (NSERC) and Thermo (13) H. Eghbali, C. Bruggink, Y. Agroskin, C.A. posium. He is on the Fisher Scientific is gratefully acknowledged. Pohl, and S. Eeltink, J. Sep. Sci. 35, 3461–3468 editorial advisory boards of Analytica Chimica (2012). Acta and Journal of Chromatography A. Direct correspondence to: [email protected] References (14) B. Yang, F. Zhang, and X. Liang, Central Eur. (1) H. Small, T.S. Stevens, and W.C. Bauman, J. Chem. 10, 472–479 (2012). M. Farooq Wahab Anal. Chem. 47, 1801–1809 (1975). (15) B. Yang, M. Takeuchi, and P.K. Dasgupta, is a senior doctoral (2) H. Small, J. Chem. Educ. 81, 1277–1284 Anal. Chem. 80, 40–47 (2008). student in Dr. Lucy’s (2004). (16) B. Wouters, C. Bruggink, G. Desmet, Y. group at the University of Alberta. His research (3) Thermo Fisher website (http://www.dionex. Agroskin, C.A. Pohl, and S. Eeltink, Anal. focuses on developing com/en-us/products/columns/ic-rfic/lp-71774. Chem. 84, 7212–7217 (2012). new stationary phases html) and the Metrohm website (http://www. (17) Dionex website (http://www.dionex.com/ and high efficiency metrohm.com/com/Produkte2/IC/Columns. en-us/products/ion-chromatography/ic-rfic- small particle ion chro- matography. Direct correspondence to: html), accessed February 2013. systems/ics-5000/lp-72594.html?ca=ics5000+), [email protected] ◾ (4) R.W. Slingsby and C.A. Pohl, J. Chromatogr. accessed February 2013. 458, 241–253 (1988). (18) D. Connolly and B. Paull, J. Chromatogr. A (5) P.W. Carr, X. Wang, and D.R. Stoll, Anal. 953, 299–303 (2002). For more information on this topic, Chem. 81, 5342–5353 (2009). (19) P. Hatsis and C.A. Lucy, Analyst 127, 451–454 please visit (6) M.E. Swartz, J. Liq. Chromatogr. Related Tech- (2002). www.chromatographyonline.com nol. 28, 1253-1263 (2005). (20) S. Pelletier and C.A. Lucy, J. Chromatogr. A (7) C.A. Lucy, J. Chromatogr. A 739, 3–13 (1996). 1125, 189–194 (2006). new! HP " Continuous high-pressure operation for capillary and analytical scale — true ™ " is here. The high-pressure capability of the ! + " enables fast separations and high resolution with the new 4 μm particle IC columns. This Reagent-Free™ HPIC system simplifies your analysis and increases reproducibility by removing the inconsistencies of manually prepared eluents.
#Learn about the ! + at thermoscientific.com/HPIC
Forty inorganic ions and organic acids resolved in 40 minutes © 2012 Thermo Fisher Scientific Inc. All rights reserved. All trademarks © 2012 Thermo Fisher Scientific Inc. All rights reserved. of Thermo Fisher Scientific Inc. and its subsidiaries. the property are Is your water safe? As a scientist in a municipal laboratory you have been entrusted with safety of your community. We understand the importance of your responsibility and as the innovation leader in Ion Chromatography we continually design new technologies like eluent generation, Capillary IC, and High-Pressure IC to ensure you are successful. The Thermo Scientific™ Dionex™ ICS-2100 Reagent-Free™ ion chromatograph has been specially designed to provide high throughput with confident results for water
analysis applications in municipal laboratories. The integrated system streamlines the The Dionex ICS-2100 is the system of choice for water analytical separation process to deliver maximum ease of use and reproducibility. analysis. Rest assured in the safety of your water with a line-up of unmatched performance preferred by regulatory agencies developing compliance and safety standards. Be confident.
For detailed information go to thermoscientific.com/ICS-2100 The complete portfolio of Thermo Scientific Dionex Ion Chromatography Systems © 2013 All trademarks are the property of Thermo Fisher Scientific Inc. and its subsidiaries. Copyrights in to the water safety images are owned by a third party and licensed for limited use only to Thermo Fisher Scientific by Shutterstock.