volume 32 number 3 march 2017 pages 423–694 jaas

JAASJournal of Analytical Atomic Spectrometryrsc.li/jaas

ISSN 0267-9477

PAPERFrank Vanhaecke et al.Evaluation of the use of cold plasma conditions for Fe isotopic analysis via multi-collector ICP-mass spectrometry: eff ect on spectral interferences and instrumental mass discrimination

Volume 32 Number 3 March 2017 Pages 423–694

JAAS

PAPER

Publ

ishe

d on

25

Janu

ary

2017

. Dow

nloa

ded

by G

hent

Uni

vers

ity L

ibra

ry o

n 08

/03/

2017

19:

37:3

7.

View Article OnlineView Journal | View Issue

Evaluation of the

aDepartment of Analytical Chemistry, Ghen

281-S12, 9000 Ghent, Belgium. E-mail: franbVrije Universiteit Brussel, Earth System Scie

† Both authors contributed equally to thi

Cite this: J. Anal. At. Spectrom., 2017,32, 538

Received 29th November 2016Accepted 25th January 2017

DOI: 10.1039/c6ja00428h

rsc.li/jaas

538 | J. Anal. At. Spectrom., 2017, 32, 5

use of cold plasma conditions forFe isotopic analysis via multi-collector ICP-massspectrometry: effect on spectral interferences andinstrumental mass discrimination

Stepan M. Chernonozhkin,†ab Marta Costas-Rodrıguez,†a Philippe Claeysb

and Frank Vanhaecke*a

The advantages and disadvantages of using cold plasma conditions in combination with both the standard

and the high-transmission (‘jet’) plasma interface and under dry and wet plasma conditions were evaluated

in the context of high-precision isotopic analysis of Fe viamulti-collector inductively coupled plasma-mass

spectrometry (MC-ICP-MS). When using the standard interface and wet plasma, cold plasma conditions

suppressed the occurrence of the polyatomic ions 40Ar14N+ and 40Ar16O+, but not that of 40Ar16O1H+.

Dry plasma conditions efficiently removed the (hydr)oxide ions, but at the cost of a dramatic increase in

the prevalence of 40Ar14N+. Thus, interference-free 56Fe/57Fe measurement can be performed at low

mass resolution using cold plasma conditions and the standard plasma interface. No alleviation of the

spectral interferences due to Ar-based ions was observed with the ‘jet’ interface with cold plasma

conditions. With cold plasma conditions, the instrumental mass discrimination affecting the Fe isotope

ratios was mitigated compared to that obtained with hot plasma when the standard interface was used

(1.4% per amu versus 3.2% per amu, respectively). This mitigation in instrumental mass discrimination

with cold plasma conditions was not observed when the high-transmission interface was used.

1. Introduction

In many contexts, predominantly within geo- and cosmochem-istry, but also to an increasing amount in the context ofbiomedicine, the ability to measure natural variations in theisotopic composition of Fe with high-precision is of crucialimportance.1–4 Fe isotope ratios can be measured with highprecision using thermal ionization mass spectrometry (TIMS);however, its use is challenged by the low ionization efficiency (forconversion into M+ ions) and the necessity of laborious isotopedilution.5 Fe isotopic analysis can be accomplished using nega-tive TIMS via the monitoring of molecular ions,6 but it was theintroduction ofmulti-collector ICP-mass spectrometry with its farhigher ionization efficiency and sample throughput that causeda breakthrough in this eld. However, the occurrence of spectralinterferences, predominantly caused by the presence of 40Ar16O+,40Ar14N+ and 40Ar16O1H+, and that of instrumental massdiscrimination are inherent disadvantages of MC-ICP-MS. Bothissues require proper attention, as they otherwise jeopardizeaccurate Fe isotope ratio measurements. Many experimental andcomputational studies focused on the formation of polyatomic

t University, Campus Sterre, Krijgslaan

[email protected]

nces, Pleinlaan 2, 1050 Brussels, Belgium

s work.

38–547

ions in ICP-MS.7–9 To suppress the occurrence of Ar-based ionsand thus, to alleviate the corresponding spectral interferences,high mass resolution, cold plasma conditions (i.e. operation ofthe ICP with a low forward power and higher nebulizer gas owrate), collision/reaction cells and to some extent, aerosol des-olvation providing dry plasma conditions, can be used. Multi-collector (MC) ICP-mass spectrometers equipped with a colli-sion/reaction cell did not become widespread, although prom-ising results were obtained for Fe isotope ratios.10,11 Also coldplasma conditions can be relied on to avoid or at least stronglysuppress the interference due to overlap of the signals of 40Ar16O+

and 56Fe+.12,13 However, as a result of the widespread applicationof sector-eld instruments capable of working at higher massresolution, cold plasma conditions became nearly forgotten.13

Also the drawbacks accompanying the use of cold plasmaconditions contributed to this waning success: deterioration ofthe sensitivity for elements characterized by a high rst ioniza-tion energy, more pronounced matrix effects and an increasedformation of oxide ions. As a result, high mass resolution1 is thetypical approach used to avoid spectral overlap in MC-ICP-MSisotopic analysis of Fe nowadays, although cold plasma condi-tions were successfully used for isotope ratio measurements ofFe,14 Li,15Mg16 and Cr17 byMC-ICP-MS. Thus, themain aim of thiswork was to re-evaluate the advantages and disadvantages of coldplasma conditions for high-precision Fe isotope ratio measure-ments. The effect on the occurrence of interfering ions and the

This journal is © The Royal Society of Chemistry 2017

http://crossmark.crossref.org/dialog/?doi=10.1039/c6ja00428h&domain=pdf&date_stamp=2017-03-06

http://dx.doi.org/10.1039/c6ja00428h

http://pubs.rsc.org/en/journals/journal/JA

http://pubs.rsc.org/en/journals/journal/JA?issueid=JA032003

Paper JAAS

Publ

ishe

d on

25

Janu

ary

2017

. Dow

nloa

ded

by G

hent

Uni

vers

ity L

ibra

ry o

n 08

/03/

2017

19:

37:3

7.

View Article Online

extent of instrumental mass discrimination were paid specialattention to.

One of the most interesting recent advances in MC-ICP-MS isthe introduction of high-transmission plasma interfaces (alsotermed ‘jet’ interface in the case of one manufacturer, Thermo-Scientic), which provide a signicantly higher ion extractionefficiency from the plasma.18 A signal enhancement with a factorof two or higher can be obtained for both heavy (e.g., Sr, Nd, Hf,Pb and U) and light (e.g., Li) elements using the ‘jet’ interface andenhanced pumping of the plasma interface, compared to thestandard arrangement in MC-ICP-MS instruments.19 The ‘jet’interface in combination with dry plasma conditions, achievedvia sample aerosol desolvation (either viamembrane desolvationor solvent evaporation techniques),20 enabled precise measure-ments of isotope ratios at element concentrations down to 0.5 ngg�1.21 Overall, an enhanced and thus closer to stoichiometrictransmission of ions through the mass spectrometer might beexpected to reduce the instrumental mass discrimination.22 Onthe other hand, high-sensitivity cone geometries are associatedwith increased space-charge repulsion within the ion beam andthus, a larger contribution to instrumental mass discriminationin the interface region. Also effects not seen with the standardinterface have been reported on. For instance, mass-independentinstrumentalmass discrimination, associated with the formationof NdO+, was observed for Nd when using the ‘jet’ interface.23

These benets and limitations of the ‘jet’ interface for high-precision isotopic analysis are yet to be fully addressed.

In view of the above, the potential benets of cool plasmaconditions for high-precision isotopic analysis of Fe were alsotested with the high-transmission ‘jet’ interface and withmembrane aerosol desolvation.

2. Experimental2.1. Reagents and materials

Only high-purity reagents and acids were used throughout theexperiments. Pro analysis grade HNO3 (65%, Chem-Lab, Bel-gium) was further puried through sub-boiling in PFA equip-ment. Optima-grade HCl (37%, Seastar Chemicals Inc.,Canada), trace-metal grade HF (47–51%, Seastar ChemicalsInc., Canada) and ultrapure 9.8 M H2O2 (Sigma Aldrich, Bel-gium) were used as such. Water was puried (to a resistivity $18.2 MU cm) in a Milli-Q Element water purication system(Millipore, France). All the Teon® recipients, disposableplastic tubes and pipette tips used in the sample preparationwere pre-cleaned by soaking in 10% pro-analysis HCl underclass-10 clean lab conditions for 48 h.

IRMM-014 (Institute for Reference Measurements andMaterials, Belgium) and NIST SRM 986 (National Institute forScience and Technology, USA) certied isotopic referencematerials of Fe and Ni, respectively, were used throughout thisstudy. A standard solution of Fe (Inorganic Ventures, TheNetherlands) with isotope ratios known from 5 years of in-housemeasurements was used for the Fe isotope ratio measurements.NIST SRM 997 Tl, NIST SRM 981 Pb, NIST SRM 987 Sr, IRMM-16Li and NIST SRM 951 B certied isotopic reference materialswere used to study the instrumental mass discrimination under


different plasma conditions. Reference materials of terrestrialmac rock were used to validate the MC-ICP-MS measurementprotocol using cold plasma conditions. These geological refer-ence materials (GRMs) included PCC-1 peridotite and BHVO-2Gbasalt (United States Geological Survey, CO, USA), and ironformation IF-G from the Centre de Recherches Petrographiqueset Geochimiques (CRPG, France).

2.2. Acid digestion and Fe chromatographic isolation

Powdered GRMs were digested via two-step microwave-assistedacid digestion in an Ethos One Microwave digestion system(Milestone, Italy). For details of the digestion procedure, thereader is referred to ref. 24. Fe was subsequently chromato-graphically isolated from the GRM digests using AG MP-1strong anion exchange resin in HCl medium, as rstdescribed in ref. 25 and 26. The puried Fe solutions were thensubjected to MC-ICP-MS isotopic analysis.

2.3. Instrumentation

Iron isotope ratios were measured using a Thermo Scientic(Germany) Neptune multi-collector ICP-MS unit, equippedwith a Peffer (Germany) OnTool™ Booster 150 dry interfacepump (130 m3 h�1 pumping speed). Fe isotope ratios weremeasured using 8 different instrument setups, realized viaa combination of cold or hot plasma conditions, dry or wetplasma conditions and standard or ‘jet’ interface. Coldplasma conditions were obtained by a gradual decrease of theplasma RF power, optimization of the ICP torch position, andcorresponding reduction of the nebulizer gas ow rate. Thelatter is needed to optimize the position of the zone with thehighest analyte ion density to that of the sampling cone ofthe interface so as to achieve maximum sensitivity for Fe.27

Either high-transmission ‘jet’ Ni interface cones or ‘stan-dard’ Ni interface cones were used to test the effect of anenhanced interface transmission efficiency on spectralinterferences and the extent of instrumental mass discrimi-nation. A 100 ml min�1 concentric nebulizer with a doublespray chamber with Scott-type and cyclonic sub-units orwith an Aridus II desolvating nebulizer system (TeledyneCETAC Technologies Inc., USA) was used to study the effect ofwet and dry plasma conditions. Detailed instrument settingsare shown in Table 1.

2.4. MC-ICP-MS measurements and correction forinstrumental mass discrimination

All sample solutions were prepared in 3%HNO3. Depending onthe instrument conguration, the Fe solutions were diluted toa nal concentration between 50 ng ml�1 and 1 mg ml�1 inorder to obtain similar intensities with the different congu-rations. The NIST SRM 986 Ni isotopic reference material wasadded to all samples and standards in a concentration equal tothat of Fe to serve as an internal standard relied on for internalmass discrimination correction. Concentrations of samplesand bracketing standards were tted to within �3%. Themeasurements of Fe isotope ratios were carried out ina sample-standard bracketing (SSB) sequence with the IRMM-

J. Anal. At. Spectrom., 2017, 32, 538–547 | 539


Table 1 Instrument settings, data acquisition parameters and multi-collector detector configuration for the Thermo Scientific Neptune MC-ICP-MS instrument used for Fe isotopic analysis at different instrument settings

Instrument settings

Hot plasma conditions Cold plasma conditions

Resolution mode Medium Medium, lowRF power, W 1260a 600Gas ow ratesa, l min�1 Cool 15 15

Auxiliary 0.8 0.8Nebulizer 0.997 0.570

Dry plasma (Teledyne Cetac Technologies Aridus II)

Gas ow ratesa, l min-1 Nebulizer 0.680 0.580Sweep Ar 4.85 4.85N2 0.008 0

Spray chamber temperature, �C 110 110Membrane desolvator temperature, �C 160 160

Wet plasma (pneumatic nebulization)

Sample uptake Pumped via a peristaltic pump, 0.1 ml min�1

Nebulizer Concentric, 100 ml min�1

Spray chamber Double, cyclonic and Scott-type sub-units

Interface

‘Standard’ interface ‘Jet’ interfaceInterface pump High-efficiency dry interface pump (130 m3 h�1 pumping speed)Sampling cone Standard Ni sampling

cone (1.1 mm actualorice ø)

Jet type Ni samplingcone (1.0 mm actualorice ø)

Skimmer cone H-type Ni skimmer (0.8mm actual orice ø)

X-type Ni skimmer (0.7mm actual orice ø)

Data acquisition

Mode Static; multi-collectionIdle time, s 3Integration time, s 4.194Number of integrations 1Number of blocks 9Number of cycles perblock

5

Cup conguration for Fe isotope ratio measurements

Cup L4 L2 L1 C H1 H3

Amplier 1011 U 1011 U 1011 U 1011 U 1011 U 1011 UNuclide 54Fe 56Fe 57Fe 58Fe, 58Ni 60Ni 62Ni

a Parameters optimized on a daily basis for the highest sensitivity and stability.

JAAS Paper

Publ

ishe

d on

25

Janu

ary

2017

. Dow

nloa

ded

by G

hent

Uni

vers

ity L

ibra

ry o

n 08

/03/

2017

19:

37:3

7.

View Article Online

014 Fe isotopic reference material as an external standard.Correction for instrumental mass discrimination was per-formed according to the model proposed by Baxter et al.28 Themodel of Baxter corrects for both the mass- and time-dependent contributions to the instrumental mass discrimi-nation. Aer this correction, the Fe isotope ratios are convertedinto delta notations, calculated following eqn (1) and relative to

540 | J. Anal. At. Spectrom., 2017, 32, 538–547

the bracketing isotopic reference material IRMM-014; here xdenotes 56 or 57.

dx=54Fe ¼

264

�xFe

.54Fe

�Ni corr;smp�

xFe

.54Fe

�Ni corr;IRMM-014

� 1

375$1000 (1)



Table 2 Analytical performances observed with cold and hot plasma conditions and using different instrument configurations (wet/dry plasmaand standard/‘jet’ interface cones). MR is medium mass resolution; LR is low mass resolution

Hot plasma Cold plasma

Pneumatic nebulization (wet plasma)Standard interface (a) Sensitivity for Fe is 20 V/ppm. (b) Fe isotope ratio

measurements in MR. (c) Precision, expressed as 2SD, is0.07& for d56/54Fe and 0.15& for d57/54Fe (n ¼ 17, 1 year)

(a) 2-fold loss in sensitivity compared to hot plasma(in MR). (b) Application of LR is possible for the56Fe/54Fe ratio: 40Ar16O+ and 40Ar14N+ are removed.40Ar16O1H+ is not entirely suppressed. (c) The extentof instrumental mass discrimination issignicantly smaller than with hot plasmaconditions. (d) Precision, expressed as SD ford56/54Fe (n¼ 5), is 0.020 and 0.034& for LR andMR,respectively

‘Jet’ interface (a) 2-fold improvement in sensitivity compared to thestandard interface. (b) Increased levels of 40Ar16O+,40Ar14N+ and 40Ar16O1H+ compared to the standardinterface. (c) Precision, expressed as 2SD, is 0.03& for d56/54Fe (n ¼ 17, 1 year)

(a) 3-fold loss in sensitivity compared to hot plasma.(b) Application of low mass resolution mode is notpossible: 40Ar16O+, 40Ar14N+ and 40Ar16O1H+ are notsuppressed. (c) The extent of instrumental massdiscrimination is similar to that with hot plasmaconditions

Membrane desolvation nebulization (dry plasma)Standard interface (a) 7-fold improvement in sensitivity compared to wet

plasma. (b) 40Ar16O+ and 40Ar16O1H+ signals are removedwhen using the desolvation unit. (c) 40Ar14N+ is muchhigher than in the case of wet plasma. (d) Precision,expressed as SD, is 0.016& for d56/54Fe (n ¼ 5)

(a) 1.5-fold loss in sensitivity compared to hotplasma. (b) Application of low mass resolution isnot possible: 40Ar14N+ is not suppressed. (c) Theextent of instrumental mass discrimination issmaller than with hot plasma conditions. (d)Precision, expressed as SD, is 0.038& for d56/54Fe (n¼ 5)

‘Jet’ interface (a) 2-fold improvement in sensitivity compared to thestandard interface. (b) 40Ar16O+ and 40Ar16O1H+ signals areremoved when using the desolvation unit. (c) Applicationof MR is not sufficient: enhanced levels of 40Ar14N+

compared to the standard interface

(a) 10-fold lower sensitivity than with hot plasmaconditions. (b) Application of low mass resolutionis not possible: 40Ar16O+ is present (when N2 gasow is switched off). 40Ar14N+not fully removed. (c)Precision is deteriorated in comparison to hotplasma

Paper JAAS

Publ

ishe

d on

25

Janu

ary

2017

. Dow

nloa

ded

by G

hent

Uni

vers

ity L

ibra

ry o

n 08

/03/

2017

19:

37:3

7.

View Article Online

3. Results and discussion

An overview of the results obtained using the instrumentcongurations: cold/hot plasma conditions, wet/dry plasma andstandard/‘jet’ interface cones (8 different conditions) is pre-sented in Table 2. These results are discussed in more detail inthe next paragraphs.

3.1. Wet plasma conditions

The peak proles at them/z ratios of 54, 56 and 57 obtained withwet plasma conditions and for different instrument settings areshown in Fig. 1. Fig. 1(A and B) presents peak proles at the m/zratios of 54, 56 and 57 measured in medium mass resolutionmode (MR, m/Dm ¼ 4500 with Dm being the difference betweenthe masses where the signal intensity is 5% and is 95% of theplateau intensity, respectively29) with standard interface conesand with hot (A) and cold (B) plasma conditions, respectively.With hot plasma conditions and at low mass resolution (LR, m/Dmz 300), the peaks of 54Fe+, 56Fe+ and 57Fe+ suffer from overlapwith those of the Ar-based polyatomic ions 40Ar14N+, 40Ar16O+ and40Ar16O1H+, respectively. MR allows the separation of the Fe+

signals (separate plateau) from the polyatomic ion, enablinginterference-free Fe isotope ratio measurements (A). A reduction


of the plasma RF power from 1260 W down to 600 W efficientlysuppresses the signals of 40Ar14N+ and 40Ar16O+ with this setup.Complete absence of these interferences with cold plasmaconditions is conrmed by accurate 56Fe/54Fe isotope ratiomeasurement at LR. The 56Fe/54Fe isotope ratios measured forour in-house isotopic standard with cold plasma conditions in LR(d56/54Fe ¼ 0.452 � 0.020&, SD, n ¼ 5) and MR (d56/54Fe ¼ 0.450� 0.034&, SD, n ¼ 5) are in agreement with the corresponding 2year average (d56/54Fe ¼ 0.450 � 0.030&, 2SD, n ¼ 200).

The results of d56/54Fe measurements with cold plasmaconditions at LR and with the standard interface in thegeological reference materials PCC-1 (3 replicate sample prep-arations, 0.029� 0.085&, 0.015� 0.048& and 0.016 � 0.080&,SD, n ¼ 4), BHVO-2G (2 replicate sample preparations, 0.076 �0.100 and 0.071 � 0.063, SD, n ¼ 4) and IF-G (3 replicate samplepreparations, 0.597 � 0.018&, 0.608 � 0.054& and 0.660 �0.041&, SD, n ¼ 4) are in good agreement with the corre-sponding literature values (see Fig. 2). The results obtained forthe in-house isotopic standard and the geological referencematerials clearly indicate that application of cold plasmaconditions at LR can be an alternative to the traditionallyapplied MR with a hot plasma when using the standard inter-face and pneumatic nebulization for sample introduction.



Fig. 1 Peak profiles atm/z ¼ 54, 56, and 57 in medium resolution mode (m/Dm ¼ 4500) obtained with different instrument configurations. Thepeak at m/z ¼ 54 is typically a combination of signals of 54Fe+ and 40Ar14N+; the peak at m/z ¼ 56 consists of signals of 56Fe+ and 40Ar16O+; thepeak at m/z ¼ 57 consists of the signals of 57Fe+ and 40Ar16O1H+. (A) hot plasma conditions (1260 W), pneumatic nebulization and standardinterface; (B) cold plasma conditions (600 W), pneumatic nebulization and standard interface; (C) hot plasma, pneumatic nebulization and ‘jet’interface; (D) cold plasma, pneumatic nebulization and ‘jet’ interface; (E) hot plasma, membrane desolvation nebulization and standard interface;(F) cold plasma, membrane desolvation nebulization and standard interface; (G) hot plasma, membrane desolvation nebulization and ‘jet’interface; (H) cold plasma, membrane desolvation nebulization and ‘jet’ cones.

JAAS Paper

Publ

ishe

d on

25

Janu

ary

2017

. Dow

nloa

ded

by G

hent

Uni

vers

ity L

ibra

ry o

n 08

/03/

2017

19:

37:3

7.

View Article Online

At the same time, cold plasma conditions cannot fullyremove the 40Ar16O1H+ ions, not even by reducing the RF powerto values lower than 600 W. As a result, the 57Fe/54Fe ratio canonly be measured in MR. Transition to the cold plasma condi-tions is associated with a 2-fold loss in sensitivity, aer re-optimization of the nebulizer and auxiliary gas ow rates andtorch position. At the same time, when using LR and coldplasma conditions, a small gain in Fe sensitivity is achieved ascompared to MR and hot plasma conditions.

Critically, aer transition to cold plasma conditions, also thevoltages of the focus lens array need to be re-tuned substantiallyto achieve maximum sensitivity and ensure signal stability.Although we didn't directly monitor the amount of all the Ar-based ions (Ar+ and Ar-containing polyatomic ions), extractedinto the mass spectrometer from the ICP, the necessity tochange the voltages of the focusing lens array seems to indicatea lower amount of extracted Ar species ions with cold than withhot plasma conditions.

Fig. 1(C and D) presents the peak proles at m/z ratios of 54,56 and 57 measured with the high-transmission ‘jet’ interface


cones and with hot and cold plasma conditions, respectively.Based on one year of experience with Fe isotopic analysis usingthe ‘jet’ and standard interface congurations in our group, anapproximately two-fold improvement in isotope ratio measure-ment precision is observed when switching to the high-transmission ‘jet’ interface (d56/54Fe ¼ 0.450 � 0.030& versusd56/54Fe ¼ 0.450 � 0.070&, 2SD, n ¼ 17). In comparison withstandard interface cones, the sensitivity achieved with the ‘jet’interface is 2-fold higher. At the same time, the higher iontransmission efficiency is associated with a signicantly higherlevel of Ar-based spectral interferences, as can be seen inFig. 1(C). With the ‘jet’ interface, however, the occurrence of Ar-based polyatomic ions is not suppressed as a result of operatingthe ICP under cold plasma conditions, such that Fe isotopemeasurements absolutely require a higher mass resolutionunder these conditions, see Fig. 1(D). At the same time,conversion to cold plasma conditions is associated with a 3-folddecrease in Fe sensitivity, which renders application of coldplasma conditions with high-transmission ‘jet’ interface conesfutile.



Fig. 2 Results from 56Fe/54Fe isotope ratio measurement of thegeological referencematerials PCC-1, BHVO-2G and IF-G as obtainedunder cold plasma conditions and at LR. The reference values corre-spond to the values reported in (ref. 11, 40–44). The error barsrepresent the internal measurement precision, expressed as thestandard deviation of 4 replicate measurements of sample solution.

Paper JAAS

Publ

ishe

d on

25

Janu

ary

2017

. Dow

nloa

ded

by G

hent

Uni

vers

ity L

ibra

ry o

n 08

/03/

2017

19:

37:3

7.

View Article Online

3.2. Dry plasma conditions

Dry plasma conditions were achieved using the Teledyne CetacAridus II membrane desolvating nebulizer. This device utilizesa PFA C-type nebulizer (100 ml min�1) and a heated spraychamber, combined with a PTFE membrane desolvator unit.This combination provides an increase in sensitivity due toa more efficient analyte transfer into the plasma and a reduc-tion in solvent-based interferences (e.g., oxide, hydride andhydroxide ions). The desolvation unit was operated with a 4.85 lmin�1 Ar sweep gas ow rate (at 160 �C). The N2 gas ow rateneeds to be tuned nely. It was found that with hot plasmaconditions, an N2 ow rate of about 10 ml min�1 provides thehighest Fe sensitivity, while with cold plasma conditions, thehighest sensitivity was achieved at a near-to-zero N2 ow rate.Aerosol desolvation provided a 7-fold improvement in thesensitivity for Fe with hot plasma conditions in comparison topneumatic nebulization. Fig. 1(E and F) presents peak prolesat the m/z ratios of 54, 56 and 57 with hot and cold plasmaconditions, respectively, when using standard interface conesand dry plasma conditions. The 40Ar16O+ and 40Ar16O1H+ signalswere successfully removed, rendering the signals of 56Fe+ and57Fe+ interference-free, but the signal from 54Fe+, typically usedas the reference isotope, suffers from signicant overlap withthe 40Ar14N+ signal. It makes application of LR impossibleunder such conditions (except for the 56Fe/57Fe isotope ratio).This 40Ar14N+ peak most probably originates from air inside thenebulizer and spray chamber of the Teledyne Cetac Aridus II orentrained within the ICP, because it is present even when the N2

gas ow is switched off.Conversion to cold plasma conditions is associated with

a 1.5-fold loss in sensitivity compared to hot plasma conditions.In contrast to wet plasma conditions, the 40Ar14N+ interference


peak is not fully removed by conversion to cold plasma condi-tions, although it is reduced signicantly. This ArN+ interfer-ence makes it impossible to measure at LR under cold plasma/dry plasma/standard interface conditions.

Among all the congurations studied, the combination ofmembrane desolvation (dry plasma) and the high-transmission‘jet’ interface with hot plasma conditions provides the highestsensitivity for Fe. Similarly to the situation with the standardinterface cones, the signals of the ArO+ and ArOH+ ions are fullyremoved by membrane desolvation. However, this congura-tion results in an even higher level of 40Ar14N+, which saturatesthe faraday cup detector with a 1011 U amplier, renderingmeasurements at LR impossible. Although this interfering peakis resolved in MR, xFe/54Fe isotope ratio measurements werestill not adequate (d56/54Fe ¼ 0.522 � 0.041&, SD, n ¼ 5,compared with the reference value of 0.450 � 0.030&). This islikely the result of an insufficient abundance sensitivity of theinstrument to avoid the inuence of the tail of the huge 40Ar14N+

peak on the plateau of the 54Fe+ signal. At the same time,57Fe/56Fe ratios measured in MR were within the range expectedfor the in-house standard (d57/56Fe¼ 0.220� 0.053&, SD, n¼ 5,compared with the reference value of d57/56Fe ¼ 0.228 �0.017&).

Conversion to cold plasma conditions when using the high-transmission ‘jet’ interface and dry plasma conditions leads toa 10-fold drop in the sensitivity compared to that with hotplasma conditions. Although the occurrence of 40Ar14N+ is notfully avoided by cold plasma conditions, it is suppressedenough to allow 54Fe+ to bemeasured inMRmode. When the N2

gas ow rate in the membrane desolvator is decreased in orderto provide the highest sensitivity, a small 40Ar16O+ interferencestarts affecting the 56Fe+ peak. This small interference can beremoved fully when the N2 ow rate is optimized, but in thelatter case, the overall Fe sensitivity is compromised.

3.3. Instrumental mass discrimination

Instrumental mass discrimination is an inherent drawback ofMC-ICP-MS and it refers to a mass-dependence of the trans-mission efficiency of the ions of the target element's isotopes22,30

from the ICP ion source to the detector. Typically, ions of theheavier of two isotopes are transmitted more efficiently than theions of the lighter one. While the extent of mass discriminationis in the 1% per amu range at mid-mass, there is a strongereffect for light elements (up to several % per amu). Coulombrepulsion (space-charge effects) and scattering effects withinthe ion beam were previously suggested to be the maincontributors to the instrumental mass discrimination,31

although other effects were also described.32,33 Various strate-gies to correct for instrumental mass discrimination are effi-ciently used,28,34,35 but reducing the extent of discriminationwould possibly allow achieving even higher precisions forisotope ratio measurements and possibly avoid the need fortarget element isolation in many cases. Ar is the majorcomponent of the ICP discharge and its ions and excited atomstransfer the RF energy to the analyte via charge transfer andpenning ionization. Aer charge separation, the Ar+ ions



Fig. 3 Plots of isotope ratios measured using a Neptune MC-ICP-MS unit operated under 8 different instrument settings: hot/cold and dry/wetplasma conditions and standard/‘jet’ interface. Measured raw isotope ratios are presented in ln-scale for (A) Fe and Ni isotope ratios measured inIRMM-14 and NIST SRM 986, respectively, (B) Pb and Tl isotope ratios measured in NIST SRM 981 and NIST SRM 997, respectively (C) B isotoperatios measured in NIST SRM 951, (D) Li isotope ratios measured in IRMM SRM 16 and (E) Sr isotope ratios measured in NIST SRM 987. All the ratiospresented are raw data – i.e. not corrected for instrumental mass discrimination. The arrows indicate the direction of the change in instrumentalmass discrimination when the plasma power is decreased.

JAAS Paper

Publ

ishe

d on

25

Janu

ary

2017

. Dow

nloa

ded

by G

hent

Uni

vers

ity L

ibra

ry o

n 08

/03/

2017

19:

37:3

7.

View Article Online

dominate the ion beam in the ion optics aer the interface. Thepresence of large amounts of Ar+ ions in the plasma and theextracted ion beam not only results in spectral interferences asa result of the occurrence of Ar-based polyatomic ions, but alsoaffects the instrumental mass discrimination due to the afore-mentioned Coulomb repulsion. Fig. 3 presents the effect ofchanging from hot to cold plasma conditions on the extent ofinstrumental mass discrimination for several isotopic certiedreference materials measured with different instrumentcongurations, described in Table 1. The Sr isotope ratios werecorrected iteratively for Rb and Kr isobaric interferences relyingon natural isotopic abundances and the Russell's equation. Fordetails of the instrument settings and data acquisition param-eters for Pb, B, Li and Sr isotope ratio measurements, ref. 21 and36–38 are referred to.


It was found that when operating the ICP under cold plasmaconditions (600 W), when the presence of Ar-based species inthe beam is expected to be lower, the extent of instrumentalmass discrimination observed for IRMM-014 solution of Fe(2.8% for d56/54Fe or 1.4% per amu) is lower than in the case ofhot plasma conditions (6.4% for d56/54Fe or 3.2% per amu), asindicated in Fig. 3(A). This observation is in agreement with theobserved effect for U on Pb isotope ratios reported earlier in theliterature.39 However, with the high-transmission ‘jet’ interfaceand wet plasma conditions, the magnitudes of instrumentalmass discrimination with cold and hot plasma conditions arenearly similar. This unexpected behavior with the ‘jet’ interfacemight be a result of the extraction of a higher number of Ar-based ions from the plasma by the high-transmission inter-face, thus rendering the effect of changing over to cold plasma



Paper JAAS

Publ

ishe

d on

25

Janu

ary

2017

. Dow

nloa

ded

by G

hent

Uni

vers

ity L

ibra

ry o

n 08

/03/

2017

19:

37:3

7.

View Article Online

less signicant. Aer 1 year of experience with the ‘jet’ interface,we have also observed that its regular application leads to an atleast 2-fold faster degradation of the resolution slits comparedto the standard interface, even though target analyte intensitiesare kept the same. We associate this observation with anincreased amount of Ar-based ions extracted from plasma bythe ‘jet’ interface. An indirect conrmation of our assumptionconcerning the effect that Ar-based species exert on the extent ofinstrumental mass discrimination observed for Fe is that whenthe strongest effect of switching over to cold plasma conditionson instrumental mass discrimination is observed (with stan-dard interface cones and under wet plasma conditions), thevoltages of the focusing lens array had to be re-tuned moresignicantly in order to achieve the highest intensity for Fe+.

Differently to Fe and Ni, both heavier elements Sr, Pb and Tl(Fig. 3(B and E)) and lighter ones Li and B (Fig. 3(C and D)) werefound to suffer a larger extent of instrumental mass discrimi-nation with cold plasma conditions than with hot plasma ones.This observation suggests that the prevailing mechanisms ofmass discrimination are potentially different for Fe and theother elements evaluated. This indicates that the mechanismsbehind instrumental mass discrimination have not been fullyunraveled yet and systematic experimental and modelingstudies should be carried out under the different conditionsstudied to obtain a more profound insight into the underlyingmechanisms.31

4. Conclusions

In the standard conguration (standard interface cones andsample introduction via pneumatic nebulization), cold plasmaconditions can efficiently suppress/remove 40Ar16O+ and40Ar14N+ ions, but not 40Ar16O1H+. Once the interferences from40Ar16O+ and 40Ar14N+ are fully removed by cold plasma condi-tions, accurate 56Fe/54Fe measurement can be accomplished atLR, providing a gain in intensity, at the cost of the possibility tomonitor the 57Fe isotope. As additionally demonstrated by56Fe/54Fe isotope ratio measurement of several geologicalreference materials, application of cold plasma conditions at LRcan be an alternative to MR under hot plasma conditions.

Application of membrane desolvation nebulization (dryplasma) efficiently removes oxide-based interferences, but leadsto a substantial increase in the 40Ar14N+ signal intensity. Thispeak of 40Ar14N+ can be suppressed with cold plasma conditionsto some extent, but not fully removed.

Application of the high-efficiency ‘jet’ interface providesa higher sensitivity owing to a higher ion transmission effi-ciency, but it also signicantly enhances the level of spectralinterferences. None of the relevant interferences can be fullyremoved by application of cold plasma conditions with the ‘jet’conguration of the plasma interface.

Overall, conversion from hot to cold plasma conditions isassociated with a lower extent of instrumental mass discrimi-nation for Fe, except with the ‘jet’ interface, with which themass discrimination is nearly equal for hot and cold plasmaconditions. Unexpectedly, this is not the case for the otherisotopic systems studied, which again opens the questions


concerning the processes giving rise to instrumental massdiscrimination in MC-ICP-MS. For Fe, it is tempting to suggestthat this effect might result from the introduction of fewer Ar-based ions into the mass spectrometer, which would alleviatespace-charge effects to some extent. The fact that the voltages ofthe focusing lens array have to be re-optimized substantiallyseems to be an indirect indication of a signicant change in theion beam composition aer switching from hot to cold plasmaconditions.

Acknowledgements

The Flemish Research Foundation FWO-Vlaanderen is acknowl-edged for nancial support through research project“G023014N”.Marta Costas-Rodrıguez thanks FWO-Vlaanderen for her post-doctoral grant. Frank Vanhaecke and Stepan Chernonozhkinacknowledge the research project funding by the Planet Topers,Interuniversity Attraction Poles Program initiated by the BelgianScience Policy Office (BELSPO).

References

1 D. Malinovsky, A. Stenberg, I. Rodushkin, H. Andren, J. Ingri,B. Ohlander and D. C. Baxter, Performance of highresolution MC-ICP-MS for Fe isotope ratio measurementsin sedimentary geological materials, J. Anal. At. Spectrom.,2003, 18, 687–695, DOI: 10.1039/b302312e.

2 B. L. Beard and C. M. Johnson, Fe Isotope Variations in theModern and Ancient Earth and Other Planetary Bodies,Rev. Mineral. Geochem., 2004, 55, 319–357, DOI: 10.2138/gsrmg.55.1.319.

3 E. D. Young, C. E. Manning, E. A. Schauble, A. Shahar,C. A. Macris, C. Lazar and M. Jordan, High-temperatureequilibrium isotope fractionation of non-traditional stableisotopes: experiments, theory, and applications, Chem. Geol.,2015, 395, 176–195, DOI: 10.1016/j.chemgeo.2014.12.013.

4 M. Costas-Rodrıguez, J. Delanghe and F. Vanhaecke, High-precision isotopic analysis of essential mineral elements inbiomedicine: natural isotope ratio variations as potentialdiagnostic and/or prognostic markers, TrAC, Trends Anal.Chem., 2016, 76, 182–193, DOI: 10.1016/j.trac.2015.10.008.

5 B. L. Beard and C. M. Johnson, High precision iron isotopemeasurements of terrestrial and lunar materials, Geochim.Cosmochim. Acta, 1999, 63, 1653–1660, DOI: 10.1016/S0016-7037(99)00089-7.

6 T. Walczyk, Iron isotope ratio measurements by negativethermal ionisation mass spectrometry using FeF4�

molecular ions, Int. J. Mass Spectrom. Ion Processes, 1997,161, 217–227, DOI: 10.1016/S0168-1176(96)04532-6.

7 B. Hattendorf, B. Gusmini, L. Dorta, R. S. Houk andD. Gunther,Abundance and Impact of Doubly Charged Polyatomic ArgonInterferences in ICPMS Spectra, Anal. Chem., 2016, 88, 7281–7288, DOI: 10.1021/acs.analchem.6b01614.

8 C. H. Ebert, T. M.Witte and R. S. Houk, Investigation into thebehavior of metal–argon polyatomic ions (MAr+) in theextraction region of inductively coupled plasma-mass



JAAS Paper

Publ

ishe

d on

25

Janu

ary

2017

. Dow

nloa

ded

by G

hent

Uni

vers

ity L

ibra

ry o

n 08

/03/

2017

19:

37:3

7.

View Article Online

spectrometry, Spectrochim. Acta, Part B, 2012, 76, 119–125,DOI: 10.1016/j.sab.2012.06.046.

9 J. W. Ferguson, T. J. Dudley, K. C. Sears, S. M. McIntyre,M. S. Gordon and R. S. Houk, Polyatomic ions in inductivelycoupled plasma-mass spectrometry part II: origins of N2H

+

and HxCO+ ions using experimental measurements

combined with calculated energies and structures,Spectrochim. Acta, Part B, 2009, 64, 690–696, DOI: 10.1016/j.sab.2009.06.008.

10 J. Vogl, P. Klingbeil, W. Pritzkow and G. Riebe, High accuracymeasurements of Fe isotopes using hexapole collision cellMC-ICP-MS and isotope dilution for certication ofreference materials, J. Anal. At. Spectrom., 2003, 18, 1125–1132, DOI: 10.1039/B301812A.

11 N. Dauphas, P. E. Janney, R. A. Mendybaev, M. Wadhwa,F. M. Richter, A. M. Davis, M. van Zuilen, R. Hines andC. N. Foley, Chromatographic separation andmulticollection-ICPMS analysis of iron. Investigating mass-dependent and independent isotope effects, Anal. Chem.,2004, 76, 5855–5863, DOI: 10.1021/ac0497095.

12 S. D. Tanner, Characterization of ionization and matrixsuppression in inductively coupled “cold” plasma massspectrometry, J. Anal. At. Spectrom., 1995, 10, 905–921, DOI:10.1039/ja9951000905.

13 F. Vanhaecke, L. Balcaen, G. De Wannemacker and L. Moens,Capabilities of inductively coupled plasmamass spectrometryfor themeasurement of Fe isotope ratios, J. Anal. At. Spectrom.,2002, 17, 933–943, DOI: 10.1039/b202409h.

14 K. Kehm, E. Hauri, C. M. O. Alexander and R. Carlson, Highprecision iron isotope measurements of meteoritic materialby cold plasma ICP-MS, Geochim. Cosmochim. Acta, 2003, 67,2879–2891, DOI: 10.1016/S0016-7037(03)00080-2.

15 C. J. Bryant, M. T. McCulloch and V. C. Bennett, Impact ofmatrix effects on the accurate measurement of Li isotoperatios by inductively coupled plasma mass spectrometry(MC-ICP-MS) under “cold” plasma conditions, J. Anal. At.Spectrom., 2003, 18, 734–737, DOI: 10.1039/b212083f.

16 M. S. Choi, J.-S. Ryu, S.-W. Lee, H. S. Shin and K.-S. Lee, Arevisited method for Mg purication and isotope analysisusing cool-plasma MC-ICP-MS, J. Anal. At. Spectrom., 2012,27, 1955–1959, DOI: 10.1039/c2ja30191a.

17 M. Schiller, E. Van Kooten, J. C. Holst, M. B. Olsen andM. Bizzarro, Precise measurement of chromium isotopesby MC-ICPMS, J. Anal. At. Spectrom., 2014, 29, 1406–1416,DOI: 10.1039/C4JA00018H.

18 N. S. Lloyd, A. Trinquier and C. Bouman, in Improvements inIsotope Ratio Accuracy and Precision Using “Jet” Sample Conefor ICP-MS, Am. Geophys. Union, Fall Meet, 2013: p. #V53B-2776. http://adsabs.harvard.edu/abs/2013AGUFM.V53B2776L,December 23 2015.

19 K. Newman, Effects of the sampling interface in MC-ICP-MS:relative elemental sensitivities and non-linear massdependent fractionation of Nd isotopes, J. Anal. At.Spectrom., 2012, 27, 63–70, DOI: 10.1039/C1JA10222B.

20 M. G. Minnich and R. S. Houk, Comparison of cryogenic andmembrane desolvation for attenuation of oxide, hydride andhydroxide ions and ions containing chlorine in inductively


coupled plasma mass spectrometry, J. Anal. At. Spectrom.,1998, 13, 167–174, DOI: 10.1039/a704274d.

21 A. Bazzano, K. Latruwe, M. Grotti and F. Vanhaecke, Leadisotopic analysis of Antarctic snow using multi-collectorICP-mass spectrometry, J. Anal. At. Spectrom., 2015, 30,1322–1328, DOI: 10.1039/C4JA00484A.

22 F. Albarede, E. Albalat and P. Telouk, Instrumental isotopefractionation in multiple-collector ICP-MS, J. Anal. At.Spectrom., 2015, 30, 1736–1742, DOI: 10.1039/C5JA00188A.

23 K. Newman, P. A. Freedman, J. Williams, N. S. Belshaw andA. N. Halliday, High sensitivity skimmers and non-linearmass dependent fractionation in ICP-MS, J. Anal. At.Spectrom., 2009, 24, 742–751, DOI: 10.1039/b819065h.

24 S. M. Chernonozhkin, S. Goderis, L. Lobo, P. Claeys andF. Vanhaecke, Development of an isolation procedure andMC-ICP-MS measurement protocol for the study of stableisotope ratio variations of nickel, J. Anal. At. Spectrom.,2015, 30, 1518–1530, DOI: 10.1039/C5JA00080G.

25 C. N. Marechal, P. Telouk and F. Albarede, Precise analysis ofcopper and zinc isotopic compositions by plasma-sourcemass spectrometry, Chem. Geol., 1999, 156, 251–273, DOI:10.1016/S0009-2541(98)00191-0.

26 L. Van Heghe, E. Engstrom, I. Rodushkin, C. Cloquet andF. Vanhaecke, Isotopic analysis of the metabolicallyrelevant transition metals Cu, Fe and Zn in human bloodfrom vegetarians and omnivores using multi-collector ICP-mass spectrometry, J. Anal. At. Spectrom., 2012, 27, 1327–1334, DOI: 10.1039/c2ja30070b.

27 F. Vanhaecke, R. Dams and C. Vandecasteele, “Zone model”as an explanation for signal behaviour and non-spectralinterferences in inductively coupled plasma massspectrometry, J. Anal. At. Spectrom., 1993, 8, 433–438, DOI:10.1039/ja9930800433.

28 D. C. Baxter, I. Rodushkin, E. Engstrom and D. Malinovsky,Revised exponential model for mass bias correction using aninternal standard for isotope abundance ratio measurementsby multi-collector inductively coupled plasma massspectrometry, J. Anal. At. Spectrom., 2006, 21, 427–430, DOI:10.1039/b517457k.

29 F. Vanhaecke and L. Moens, Overcoming spectral overlap inisotopic analysis via single- and multi-collector ICP-massspectrometry, Anal. Bioanal. Chem., 2004, 378, 232–240,DOI: 10.1007/s00216-003-2175-8.

30 M. Rehkamper, M. Schonbachler and C. H. Stirling, MultipleCollector ICP-MS: Introduction to Instrumentation,Measurement Techniques and Analytical Capabilities,Geostand. Geoanal. Res., 2001, 25, 23–40, DOI: 10.1111/j.1751-908X.2001.tb00785.x.

31 N. Kivel, I. Gunther-Leopold, F. Vanhaecke and D. Gunther,Isotope fractionation during ion beam formation in multi-collector inductively coupled plasma mass spectrometry,Spectrochim. Acta, Part B, 2012, 76, 126–132, DOI: 10.1016/j.sab.2012.06.045.

32 H. Andren, I. Rodushkin, A. Stenberg, D. Malinovsky andD. C. Baxter, Sources of mass bias and isotope ratio variationin multi-collector ICP-MS: optimization of instrumental



Paper JAAS

Publ

ishe

d on

25

Janu

ary

2017

. Dow

nloa

ded

by G

hent

Uni

vers

ity L

ibra

ry o

n 08

/03/

2017

19:

37:3

7.

View Article Online

parameters based on experimental observations, J. Anal. At.Spectrom., 2004, 19, 1217–1224, DOI: 10.1039/b403938f.

33 N. Taylor and P. B. Farnsworth, Experimentalcharacterization of the effect of skimmer cone design onshock formation and ion transmission efficiency in thevacuum interface of an inductively coupled plasma massspectrometer, Spectrochim. Acta, Part B, 2012, 69, 2–8, DOI:10.1016/j.sab.2012.02.003.

34 L. Yang, Accurate and precise determination of isotopicratios by MC-ICP-MS: a review, Mass Spectrom. Rev., 2009,28, 990–1011, DOI: 10.1002/mas.

35 A. Al-Ammar and R. Barnes, Improving isotope ratioprecision in inductively coupled plasma quadrupole massspectrometry by common analyte internal standardization,J. Anal. At. Spectrom., 2001, 16, 327–332, DOI: 10.1039/b100973g.

36 S. Rich, S. W. Manning, P. Degryse, F. Vanhaecke and K. VanLerberghe, Strontium isotopic and tree-ring signatures ofCedrus brevifolia in Cyprus, J. Anal. At. Spectrom., 2012, 27,796–806, DOI: 10.1039/c2ja10345a.

37 V. Devulder, P. Degryse and F. Vanhaecke, Development ofa Novel Method for Unraveling the Origin of Natron FluxUsed in Roman Glass Production Based on B IsotopicAnalysis via Multicollector Inductively Coupled PlasmaMass Spectrometry, Anal. Chem., 2013, 85, 12077–12084,DOI: 10.1021/ac403176c.

38 K. Van Hoecke, J. Belza, T. Croymans, S. Misra, P. Claeys andF. Vanhaecke, Single-step chromatographic isolation oflithium from whole-rock carbonate and clay for isotopicanalysis with multi-collector ICP-mass spectrometry,


J. Anal. At. Spectrom., 2015, 30, 2533–2540, DOI: 10.1039/c5ja00165j.

39 R. Santos, M. J. Canto Machado, I. Ruiz, K. Sato andM. T. S. D. Vasconcelos, Space charge and massdiscrimination effects on lead isotope ratio measurements byICP-QMS in environmental samples with high uraniumcontent, J. Anal. At. Spectrom., 2007, 22, 783–790, DOI:10.1039/b618783h.

40 P. R. Craddock and N. Dauphas, Iron Isotopic Compositionsof Geological Reference Materials and Chondrites, Geostand.Geoanal. Res., 2011, 35, 101–123, DOI: 10.1111/j.1751-908X.2010.00085.x.

41 M. Oeser, S. Weyer, I. Horn and S. Schuth, High-Precision Feand Mg Isotope Ratios of Silicate Reference GlassesDetermined In Situ by Femtosecond LA-MC-ICP-MS and bySolution Nebulisation MC-ICP-MS, Geostand. Geoanal. Res.,2014, 38, 311–328, DOI: 10.1111/j.1751-908X.2014.00288.x.

42 M. A. Fehr, P. S. Andersson, U. Halenius and C.-M. Morth,Iron Isotope Variations in Holocene Sediments of theGotland Deep, Baltic Sea, Geochim. Cosmochim. Acta, 2008,72, 807–826, DOI: 10.1016/j.gca.2007.11.033.

43 J. Marin-Carbonne, C. Rollion-Bard and B. Luais, In situmeasurements of iron isotopes by SIMS: MC-ICP-MSintercalibration and application to a magnetite crystalfrom the Gunint chert, Chem. Geol., 2011, 285, 50–61,DOI: 10.1016/j.chemgeo.2011.02.019.

44 N. Dauphas, P. R. Craddock, P. D. Asimow, V. C. Bennett,A. P. Nutman and D. Ohnenstetter, Iron isotopes mayreveal the redox conditions of mantle melting fromArchean to present, Earth Planet. Sci. Lett., 2009, 288, 255–267, DOI: 10.1016/j.epsl.2009.09.029.



volume 32 number 3 march 2017 pages 423–694 jaas

Documents