Content uploaded by Lucie Sauzéat
Author content
All content in this area was uploaded by Lucie Sauzéat on Apr 20, 2021
Content may be subject to copyright.
Accepted Article
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1111/ggr.12161
This article is protected by copyright. All rights reserved.
Received Date : 21-Sep-2016
Revised Date : 25-Nov-2016
Accepted Date : 02-Dec-2016
Article type : Original Article
Nitrile, Latex, Neoprene and Vinyl Gloves: A Primary Source of
Contamination for Trace Element and Zn Isotopic Analyses in
Geological and Biological Samples
Marion Garçon (1, 2, 4)*, Lucie Sauzéat (3), Richard W. Carlson (2), Steven B. Shirey (2),
Mélanie Simon (3), Vincent Balter (3) and Maud Boyet (1)
(1) Laboratoire Magmas et Volcans, Campus Universitaire des Cézeaux, 6 avenue Blaise Pascal,
TSA 60026 - CS 60026, 63178 Aubière Cedex, France
(2) Carnegie Institution for Science, Department of Terrestrial Magnetism, Washington, DC 20015,
USA
(3) UMR 5276, Laboratoire de Géologie de Lyon, Ecole Normale Supérieure de Lyon, BP 7000 Lyon,
France
(4) Present address: Department of Earth Sciences, Institute of Geochemistry and Petrology, ETH
Zürich, Clausiustrasse 25, 8092 Zürich, Switzerland
* Corresponding author. e-mail: marion.garcon@erdw.ethz.ch
Accepted Article
This article is protected by copyright. All rights reserved.
Exogenic contamination is of primary concern for geochemical and biological clean laboratories
working with sample sizes at the nanogram or even sub-picogram level. Here we determined sixty
trace elements in fifteen different types of gloves from major suppliers worldwide to evaluate whether
gloves could be potential sources of contamination for routine trace element and isotope
measurements. We found that all gloves contain some trace elements that can be easily mobilised in
significant amounts. In weak acid at room temperature, the tested gloves released up to 17 mg of Zn,
more than 1 μg of Mg, Ti, Mn, Fe, Rb, Sr, Zr, Sn, Hf and Pb and between 100 and 1000 ng of Li, Sc,
V, Cr, Ni, Cu, Ga, As, Se, Y, Ag, Ba, La, Ce, Nd, Tl and Th. Vinyl gloves released lower quantities of
biologically and geologically important elements, with the exception of In and Sn. Isotopic analyses
indicate that all gloves share roughly the same Zn isotopic composition (average δ66Zn = +0.10 ±
0.32‰ (2s)). A single contact between glove and labware releases an average of ~ 6 ng of Zn, and
hence can significantly shift δ66Zn above the precision level when the amount of Zn determined is
below 500 ng.
Keywords: gloves, contamination, zinc, isotopic determinations, trace elements.
Received 21 Sep 16 – Accepted 02 Dec 16
Trace element contents and Zn isotopic compositions are used in various research fields such as
Earth and planetary sciences, biology and, more recently, medicine. The level of precision and
sensitivity of such measurements has significantly increased through time mainly due to technological
improvements in mass spectrometry. To date, routine analyses of Zn isotopic compositions and trace
element contents reach an average precision of ± 0.1‰ (2s) and ± 5% (2s), respectively, with sample
sizes as low as the nanogram to the low picogram range. To obtain accurate results, following the
best analytical procedures is necessary, in particular control of exogenic contamination during sample
preparation and analysis. In this context, most geochemical and biological clean laboratories carefully
acid-wash all pipette tips, tubes, and other labware, and use ultra-pure water and acids that they distil
themselves. While several precautions, such as measuring laboratory blanks in reagents, are taken to
Accepted Article
This article is protected by copyright. All rights reserved.
evaluate contamination levels, less attention has been given to less direct sources of surface
contamination as well as their potential impact on measurement quality. Three decades ago,
Hoffmann (1988) mentioned that the sources of contamination were often where one least expects to
find them, for example in labware such as gloves, pipette tips, lab wipes, polyethylene bottles and
spoons; an assumption that was confirmed twelve years later by Friel et al. (1996). Friel et al. (1996)
found that gloves were highly enriched in several elements such as Zn, Fe and Se and recommended
wearing vinyl gloves that were acid-washed before use for routine trace element work. In their study,
Friel et al. (1996) report only approximately twenty-five trace elements in brands of latex and vinyl
gloves that are mostly not used anymore in clean laboratories. The glove market has indeed
significantly evolved and diversified through the past two decades. Several new types of gloves, made
for example of nitrile or neoprene, are now commonly used in biological and geochemical laboratories
worldwide. In addition, the scientific community has been interested in studying an increasing diversity
of chemical and isotopic compositions highlighting the need to review and re-assess the risk of
contamination associated with the use of recent laboratory gloves on a wide range of trace and minor
elements.
In this paper, we quantitatively evaluate the risk of contamination caused by gloves on sixty trace
elements commonly measured in Earth and planetary sciences, biology and medicine. We tested
fifteen different types of gloves made of vinyl, neoprene, nitrile or latex from different suppliers such
as Kimberley-Clark®, Ansell® or Medline®. Since Zn is by far the most enriched trace element in
gloves (Friel et al. 1996), we also determined their Zn isotopic compositions (henceforth referred as
δ66Zn) to see whether glove contamination could significantly shift Zn isotopic compositions of
geological and biological samples. The aims of the study are thus multiple: (1) to determine whether
the gloves currently used in clean laboratories are still a major source of contamination for trace
element and isotopic work; (2) which gloves offer lower levels of contamination; (3) to quantify the
amount of each element that can be released from gloves into samples during preparation and
analytical processes and pinpoint which elements are most susceptible to contamination; (4) to
quantify the effect of glove contamination on Zn isotopic compositions of geological and biological
samples.
Accepted Article
This article is protected by copyright. All rights reserved.
Tested gloves and analytical methods
Tested gloves
The gloves tested in this study are among the most often used in geochemical and biological clean
laboratories. These include nitrile, vinyl, neoprene and latex powder-free gloves from ten main
suppliers worldwide: Kimberley Clark®, Ansell®, Polysem Medical®, LCH medical®, Microflex®,
Showa®, Clean-Dex®, Jet® and Medline®. Further details on the models, colours, packaging and
reference numbers of the tested gloves are provided in Table 1. To avoid sample bias and allow a
direct comparison between all analyses, we tested M size (7.5’’–8.5’’) gloves only. All gloves were
taken from unopened boxes to avoid external contamination. Complete duplicates were analysed for
three gloves (a vinyl, a nitrile and a neoprene glove) to assess the repeatability of the results and the
measurement precision.
Materials and analytical conditions
All tests were performed in a clean laboratory below laminar flow clean hoods using ultra-pure water
at 18.2 MΩ cm resistivity and acids that were distilled to ensure low trace element contents.
Polypropylene tubes and pipette tips used all through the analytical procedure were not acid-washed
before use to avoid any contamination by gloves worn during cleaning procedure. However, blanks
were systematically run to quantify detection limits and integrate potential contamination from the
material and/or the acids used to perform the experiments. Blank measurements thus eliminated the
possibility that pipette tips, polypropylene tubes and reagents used in the analyses contributed to the
results on for the gloves.
Experiments on gloves
We used two different techniques to test the potential contamination by gloves during analytical
preparation and analyses of geological or biological samples. They are both schematically illustrated
in Figure 1. The first test (Test A) consisted of soaking the gloves at room temperature for 40 hours in
20 ml of 0.4 mol l-1 HNO3 + 0.05 mol l-1 HF and then analysing the content of the soaking solutions to
Accepted Article
This article is protected by copyright. All rights reserved.
see which chemical elements were preferentially released in solutions. The test was done in weak
acid to mimic the effects of leaching by acid fumes that may occur when handling samples and
labware in the laboratory. The leaching acid also corresponds to that in which samples are commonly
diluted and then introduced into an ICP-MS for trace element determination (e.g., see Chauvel et al.
2011). The small amount of HF ensures the stability of the solution by preventing the precipitation of
high field strength elements such as Zr. The duration of the test was chosen arbitrarily and likely does
not correspond to the complete leaching of the gloves. The detailed procedure consisted of putting
each of the tested gloves in a 50 ml polypropylene vial being careful not to cross-contaminate the
different vials in between each tested glove. We then added 20 ml of 0.4 mol l-1 HNO3 + 0.05 mol l-1
HF in each vial and made sure that the whole gloves were soaked in the acid for 40 hours. The
soaking solutions were pipetted out the vials and analysed on the mass spectrometer following the
method described below.
The second test (Test B) aimed to evaluate how easily gloves could release trace and minor elements
just through contact with disposable labware (cf. Figure 1). For this test, we put 3 ml of 0.4 mol l-1
HNO3 in a 15 ml polypropylene tube, touched the extremity of a 1 ml pipette tip with a dry glove for a
second, washed the pipette tip with the solution into the tube and analysed the latter solution for trace
element concentrations. In parallel, we also rubbed each glove along the interior of another 15 ml
polypropylene tube for a second, added 3 ml of 0.4 mol l-1 HNO3, swished the acid around and then
determined the trace element content of the solution. The acid used in Test B was different from that
used in Test A for a technical reason. To reach higher sensitivity and improve detection limits, we
measured the solutions from Test B with a different instrument (see below) that was not conditioned
to analyse solutions with HF even in small amounts. Therefore, Test B was performed using 0.4 mol l-
1 HNO3 only.
Analyses for trace element contents
Trace element mass fractions were measured in solutions from Test A using the quadrupole ICP-MS
Agilent 7500 at the Laboratoire Magmas et Volcans (LMV) in Clermont-Ferrand, France. Solutions
from Test B were measured using the quadrupole ICP-MS Thermo iCap-Q at the Ecole Normale
Accepted Article
This article is protected by copyright. All rights reserved.
Supérieure (ENS) of Lyon, an instrument with lower detection limits. Most elements were measured in
“no gas mode” except Mg, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Ge, As and Se that were
determined in “He mode” using a collision cell to minimise spectral interferences caused by
polyatomic ions. Mass fractions were calibrated using repeated measurements of solutions containing
sixty trace elements mixed to concentrations of 0.02, 0.1, 0.5, 1, 2 and 60 ng ml-1. These solutions
were also used to monitor and correct the instrumental drift over one analytical sequence. Solutions
from Test A had to be diluted by factors of 50 or even 500 to precisely measure Zn mass fractions
and avoid detector saturation. For each test, detection limits (DL) were calculated following IUPAC
guidelines for each element i as DLi = xbi + k sbi where k = 3 (95% confidence level), and xbi and sbi
are, respectively, the mean and standard deviation of the number of counts measured in blanks (i.e.,
acids stored and handled with the same material used to test the gloves). The measurement precision
of Tests A and B was estimated using complete duplicates (see online supporting information
Appendices S1 and 2). Duplicated analyses are graphically represented in Appendix S3 where one
can see that Test A produces repeatable results that clearly differentiate the tested gloves. The
repeatability is poorer when the concentrations are low, hence closer to detection limits (cf. G1 and
G1 Dup in Appendices S1 and 3). The repeatability of Test B is not as good as for Test A probably
because the duration and surface of contact between glove and labware was variable from one tested
glove to another.
Analyses for Zn isotopic compositions
Zinc isotopic compositions were measured in the soaking solutions from Test A at the Ecole Normale
Supérieure (ENS) of Lyon, France following the procedure described by Maréchal et al. (1999). The
solutions were generally not purified by ion-exchange chromatography before measurement because
Zn was a hundred to a million times more concentrated than any other potential interfering element in
the solutions. In a recent paper, Chen et al. (2016) indicate that small amounts of Ni and Ti residues
could produce isobaric interferences and shift the measured δ66Zn values by more than 0.07‰ if the
Ni/Zn and Ti/Zn ratios were higher than 0.001 and 0.01, respectively. The Ni/Zn and Ti/Zn ratios of the
soaking solutions are either close to these values or significantly lower (see Appendices S1 and 4).
However, to ensure that the matrix of the solutions did not induce interferences and/or that Cu from
Accepted Article
This article is protected by copyright. All rights reserved.
the gloves did not influence the mass bias correction based on Cu-doping, three vinyl gloves (plus two
duplicates) with relatively high Cu/Zn, Ni/Zn and Ti/Zn ratios and one nitrile glove were processed
through column chemistry. For these samples, Zn was purified on quartz columns filled with 1.8 ml of
Bio-Rad AGMP-1 (100–200 mesh) anion-exchange resin by eluting 10 ml of HNO3 0.5 mol l-1 (see
Maréchal et al. 1999 for more details). The comparison of Zn isotopic compositions measured with
and without purification by ion-exchange chromatography is discussed in the results section.
On the day of the analyses, an aliquot of each solution from Test A was evaporated to dryness and
dissolved in a Cu-doped solution (Cu NIST SRM 976, National Institute of Standards and Technology,
Gaithersburg, MD, USA) to reach a Zn concentration of about 300 ng ml-1, which is similar within 10%
to the concentration of the standard solution that was run between each sample. Zinc isotopic
compositions, expressed as:
(1)
were measured on a Nu Plasma (Nu 500) MC-ICP-MS in wet plasma conditions to avoid potential
isotopic bias induced by the desolvation system on Zn isotopic ratios (Jaouen et al. 2013b).
Instrumental mass fractionation was corrected with an exponential law using the isotopic composition
of Cu that was introduced as dopant in the analysed samples as recommended by Maréchal et al.
(1999). δ66Zn were calculated by calibrator bracketing using Zn JMC 3-0749L (also called JMC-Lyon;
Johnson Matthey, Royston, UK) as reference calibrator. The latter was repeatedly measured in
between each sample to correct for the instrumental drift throughout the measurement sequence.
The accuracy of Zn isotopic compositions was assessed based on the analyses of an in-house
standard solution (sheep plasma; SHP) at the beginning of each measurement sequence. The
average δ66Zn measured for the in-house standard solution was +0.75 ± 0.12 (2s, n = 2), which is in
Accepted Article
This article is protected by copyright. All rights reserved.
good agreement with our reference in-house value of +0.73 ± 0.08 (2s, n = 8). Based on repeated
measurements of calibrator Zn JMC 3-0749L, re-run samples and complete duplicate analyses, we
estimate the precision of our measurements at ± 0.12‰ (2s). The long-term precision (duration = )
based on the repeated measurements of Zn JMC 3-0749L alone is better than ± 0.06‰ (2s, n = 140).
Results
The amount of each element released by gloves in the soaking solutions of Test A is reported in
Appendix S1 and shown normalised to geological and biological reference materials in Figure 2. The
first important result is that all soaking solutions contain amounts of almost all elements determined
that were well above detection limits. The exceptions are elements such as Ge, Te or Re (Appendix
S1, Figure 2). Gloves are thus highly enriched in leachable trace elements, especially in Zn for which
the amounts released in solutions are very large, between 11 μg and 17 mg (Figure 3). Large
amounts (> 1 μg) were also measured for Mg, Ti, Mn, Fe, Rb, Sr, Zr, Sn, Hf and Pb. Nitrile, neoprene
and latex gloves released the highest amounts of elements into solution. Latex gloves released the
most rare earth elements (REE). Vinyl gloves generally released much lower amounts of trace
elements, except G1 for In and Sn. Neoprene gloves released intermediate amounts of leachable
elements. No systematic variation was noticed as a function of the packaging or the colour of the
gloves.
Results from Test B are summarised in Appendix S2 and shown in Figure 4. The amounts of
elements released by gloves when touching a pipette tip were generally higher than when touching
the interior of a polypropylene tube (Figure 4). We think these features are directly related to the way
the tests were performed, mostly the fact that the surface of contact between glove and pipette tip
was probably larger than for the tube. In any case, results of Test B show that surface contaminants
on the gloves are easily transferable from glove to labware and then to solutions by simple and quick
contact. The most abundant transferable elements include Zn (up to 40 ng), Mg (up to ~ 20 ng), Fe
(up to ~ 17 ng), Ba (up to ~ 4 ng) and all REE (for example, up to ~ 10 ng for Ce and 17 ng for Nd).
Accepted Article
This article is protected by copyright. All rights reserved.
Amounts of Ti, Cr, Mn, Fe, Ni, Cu, Ga, Ge, As, Se, Sn, Sr and Pb > 0.1 ng were also detected for
some tests. Vinyl gloves generally released lower amounts of potentially contaminant elements than
all other gloves, except Sn. Note that most measurements are below detection limits (Appendix S2).
This does not mean that the risk of contamination is negligible since detection limits were often > 0.1
ng or even > 1 ng for some elements and highly variable from one element to another.
The Zn isotopic compositions of the five soaking solutions analysed with and without separation by
ion-exchange chromatography are compared in Appendix S4. No large difference was observed with
and without purification although the two solutions with the highest Cu/Zn ratios (i.e., Cu/Zn ~ 0.002
for G8 and 0.006 for G14; cf. Appendix S4) have δ66Zn values that do not totally overlap within
measurement precision. Since these variations remain small and within the range of values measured
for the other tested gloves, we conclude that not performing ion exchange purification of the soaking
solutions makes only a very minor difference in Zn isotopic composition and its omission probably has
no effect at all when Cu/Zn is < 0.002. Note that high Ti/Zn and Ni/Zn ratios (cf. G1 and G1 Dup,
Appendix S4) did not bias δ66Zn beyond measurement precision.
The Zn isotopic compositions of all the soaking solutions of Test A are reported in Table 2 and shown
in Figure 5. For the four samples (plus two duplicates) processed through column chemistry, we
reported the compositions measured after purification by ion-exchange chromatography in Table 2
(see Table caption for more details). The measured isotopic compositions span a relatively small
range of values with -0.20‰ < δ66Zn < +0.34‰. The average isotopic composition of the tested
gloves yielded δ66Zn = +0.10 ± 0.32‰ (2s). The most negative values were measured in nitrile and
latex gloves while the most positive were measured in vinyl gloves. Within the measurement
precision, there was, however, no significant difference of Zn isotopic composition between vinyl,
nitrile, neoprene and latex gloves or as a function of the glove colour.
Accepted Article
This article is protected by copyright. All rights reserved.
Discussion
Gloves as potential sources of contamination
Trace element contamination by gloves is of significant concern for all laboratories dealing with
analyses at the nanogram to picogram levels. The potential for contamination is a function of the
abundance of the element in the sample to be analysed relative to the amount of that element
contributed by the glove. To take this effect into account when discussing the results, we normalised
the amount of each element released by gloves into the soaking solutions (results from Test A) to the
amount of that element present in 1 mg of BHVO-2 (basalt), 1 mg of CI-chondrite (meteorite), 1 g of
SLRS-4 (natural river water) or 30 mg of NIST SRM 1598 (bovine serum). These amounts correspond
to the approximate amounts of geological and biological test portions usually run on ICP-MS for
routine trace element determinations. The normalised amounts are shown in Figure 2a–d. In Figure 3,
we report the amounts of selected elements leached from the gloves in binary plots as a function of
the Zn released from the same glove. The results of Test A (Figures 2 and 3) demonstrate that gloves
currently used in clean laboratories are important sources of contamination for both geological and
biological samples. For example, the amount of many elements leached from the gloves is
comparable and often higher than the amount of those elements present in 1 mg of geological
reference material BHVO-2 (basalt) (Figure 2a). In particular, the glove leaches contain up to 100,000
times more Zn; 1,000 times more Se, Ru, Ag, Cd, Sn, Ir, Au, Tl, Pb; 100 times more As, Rb, Zr, Pd,
In, Sb, Te, Cs, Hf, Re, Pt, Bi, Th, U; and ten times more Li, Be, Sc, Sr, Y, REE and W than in a
typically sized analytical aliquot of the geological reference material. Compared with materials with
lower trace element mass fractions such as chondrite or river water (Figures 2b and 2c), enrichment
factors reach higher values, for example more than 1000 for Be, Sc, Ti, V, Cr, Co, Ni, Zn, Se, Rb, Sr,
Y, Zr, Cd, Sn, Cs, REE, Hf, Tl, Pb, Bi, Th and U. The situation is even worse for biological samples
because their trace element contents are usually much lower than in geological samples. Figure 2d
shows that the gloves can release up to 100,000 times the amounts of V, Cr, Zn and Pb present in 30
mg of biological reference material NIST SRM 1598 (i.e., bovine serum); 10,000 times the amounts of
Mn, Ni, As, Cd, Cs and Tl; 1,000 times the amount of Co; 100 times the amounts of Mg, Fe, Se and
Accepted Article
This article is protected by copyright. All rights reserved.
Rb; and up to ten times the amounts of Cu and Mo. Whatever the chosen reference, the results of
Test A indicate that almost all elements are present at higher levels in the glove leaches than in the
samples themselves, which constitutes a tremendous potential for contamination. This conclusion is
supported by the results of our second test (Test B) aiming to quantify the amount of trace elements
released by gloves by a single contact with either a dry pipette tip that is then used to pipette the
sample or the interior of a polypropylene tube in which the sample is stored (Appendix S2 and Figure
4). This test demonstrates that high amounts (between 1 and ~ 40 ng) of Mg, Fe, Zn, Ba, La, Ce, Nd,
Sm, Gd, Dy, Ho, Er, Yb and non-negligible amounts (between 0.1 and 1 ng) of Ti, Cr, Mn, Cu, Ga,
Ge, As, Se, Sn, Eu, Tb, Tm, Lu and Pb can be transferred from the glove to the sample by a single
contact with labware. The contamination can, however, reach much higher values if repeated contacts
between labware and glove occur during sample preparation and analysis. Surprisingly, the neoprene
glove G5 and the nitrile gloves G10 and G15 released comparable amounts of rare earth elements in
the solutions of Test A and Test B although the two experiments were very different in design and
duration. During Test A, we noted that the soaking solutions of G5 and G10 were cloudy, possibly
indicating the presence of small particles in suspension. The surface of gloves G5, G10 and G15 is
also rougher and more granular than the other gloves, which may be related to the easier release of
particles from their surface by single dry contact. Whatever the case, Test B represents only one
possible scenario of contamination in the laboratory. The behaviour of wet gloves or gloves exposed
to acid fumes was not tested and could be very different. Careful handling of the samples is thus
crucial to reduce the risk of contamination. Contamination of labware by gloves when acid-washing
pipette tips, beakers and polypropylene tubes is, in particular, a serious concern. The operation
requires several handlings with gloves that may contaminate the labware rather than clean it. We thus
suggest the use of plastic tongs at every step of the acid-washing procedure to avoid direct contact
between gloves and labware, and subsequent contamination of the samples.
The two tests show that all gloves contain high amounts of trace and minor elements, which is a
potentially serious source of contamination for trace element and isotopic determinations. However,
the amount of elements released depends on the material from which the glove is made. Except for In
and Sn, nitrile, latex and neoprene gloves generally contain higher amounts of leachable elements
than vinyl gloves, on average between 10 and 100 times more Mg, Ti, V, Cr, Mn, Co, Ni, Zn, Se, Rb,
Accepted Article
This article is protected by copyright. All rights reserved.
Sr, Y, Zr, Nb, Rh, Cd, Cs, Ba, LREE, Hf, Pt, Tl, Pb and U (Figures 2 and 3). Similarly, results from
Test B (Figure 4) show that vinyl gloves generally release less Zn and other elements (below
detection limits), with the exception of Sn and Fe, by single contact with labware. The adequacy of
each type of glove for handling a particular type of chemical needs to be established by rigorous
testing and should not be presumed without such testing. However, the wearer needs to keep in mind
that the risk of sample contamination is higher for most elements when using gloves made of nitrile,
latex or neoprene to acid-wash labware and to prepare and analyse geological and biological
samples.
Why are the gloves so enriched in trace elements?
The high Zn content in rubber gloves (i.e., neoprene, latex and nitrile) is very likely explained by
manufacturing processes and the use of activators and accelerators such as Zn oxide and Zn organic
carbamates or thiazoles (Nieuwenhuizen et al. 2001). These catalysts are used to “cure” the rubber
(i.e., sulfur vulcanisation), an industrial process that allows the cross-linking of rubber
macromolecules to confer strength, elasticity and good ageing properties to the gloves. Even if most
suppliers ensure low levels of accelerators in the finished products, the certificates of analysis that we
were able to consult for some of the tested gloves (nitrile and latex) indicate Zn mass fractions
between 2 and 60 μg g-1, which is still extremely high if the gloves are used for trace element or
isotopic work. Since other trace elements are generally positively correlated with Zn (Figure 3), we
suspect that they derive from the Zn catalysts used to cure the rubber. Note that the data are more
scattered for Zr-Hf and rare earth elements (Figures 3d and 3g), which could be related to the type of
Zn catalyst. Trace elements are probably present in different proportions in Zn oxides and Zn
complexes (carbamates and thiazoles). The type of Zn catalysts added to rubber mixtures depends
on industrial processes and is highly variable from one glove to another. According to the information
available on the websites of some of the glove manufacturers, two nitrile gloves from the same brand
can for example be made from rubber treated with Zn oxides or Zn complexes or a combination of
both. Unfortunately, the exact recipe used by glove manufacturers is not available for each tested
glove making it difficult to establish a direct relationship between the type of Zn catalyst used and the
trace element budget of the finished product.
Accepted Article
This article is protected by copyright. All rights reserved.
For vinyl gloves, the lower but still significant mass fractions of Zn cannot be accounted for the use of
Zn activators and accelerators for sulfur vulcanisation since this process is used for rubber only, not
for plastic such as vinyl (i.e., PVC). The presence of Ca-Zn stabilisers is however clearly mentioned in
the composition of all tested vinyl gloves. Such stabilisers are used to prevent the degradation of PVC
when heated to soften during the extrusion or moulding processes (Balköse et al. 2001, Fang et al.
2009). Ca-Zn stabilisers are very likely the sources of most trace and minor elements released by
vinyl gloves.
For both plastic and rubber gloves, zinc and associated minor and trace elements are additives
entering directly in the composition of the gloves. They are not carried in a coating or disseminated at
the surface of the gloves, except possibly REEs for the neoprene glove G5 and the nitrile gloves G10
and G15. Pre-washing the gloves under water or acid-washing them before use as recommended by
Friel et al. (1996) might slightly reduce the risk of contamination, but is probably not enough to totally
rule out the risk of contamination. In our opinion, the most efficient way to minimise the contamination
is to avoid as much as possible the contact between gloves and labware during washing, preparation
and analyses.
Effect of glove contamination on Zn isotopic compositions
Zinc is clearly the most abundant leachable element in all types of glove, which means that glove
contamination is a serious concern for the measurement of the Zn mass fraction in biological and
geological samples. But, what of zinc isotopic compositions? Zinc isotopic compositions have been
used both to understand Earth and planetary processes including the formation of the Solar System
(Bridgestock et al. 2014), the origin of the Moon (Paniello et al. 2012), magmatic differentiation
processes occurring in the bulk silicate Earth (Chen et al. 2013), as well as biological processes
(Aucour et al. 2015, Cloquet et al. 2008, Moynier et al. 2009, Pichat et al. 2003), metabolic and
pathological reactions in humans and mammals (Balter et al. 2010, Büchl et al. 2008, Moynier et al.
2013, Ohno et al. 2005, Stenberg et al. 2005), dietary habits (Costas-Rodríguez et al. 2014, Jaouen
et al. 2013b, Van Heghe et al. 2012), and as markers of gender and age (Jaouen et al. 2013a). More
recently, the use of this analytical tool is also increasing in the disease-related research field where
stable isotope studies are used as new biomarkers of diverse pathologies such as breast cancer
Accepted Article
This article is protected by copyright. All rights reserved.
(Larner et al. 2015). In these studies, the relevance of the results and the associated interpretations
are based on highly precise isotopic determinations by MC-ICP-MS because the range of isotopic
variation in geological samples and, in particular, in biological samples is relatively small, a few per
mil at most. To date, the measurement precision on δ66Zn measured by MC-ICP-MS can be lower
than ± 0.1‰ (2s) when performed under optimal analytical conditions (i.e., high sensitivity, stable
plasma, elevated Zn mass fractions in test portions).
To quantify the impact of glove contamination on Zn isotopic compositions, we calculated the Zn
isotopic composition of a binary mixture of few nanograms of Zn from gloves and the rest from
geological or biological samples that have different Zn isotopic compositions (Figure 6). We re-wrote
the classical mixing equation to express the Zn isotopic composition of a binary mixture as a function
of the amount of Zn present in the two end-members:
(2)
where and are the amount of Zn (generally in ng) from the glove and the sample
respectively. For the glove end-member, we used the average Zn isotopic composition of all gloves
(δ66Zn = +0.10 ± 0.32‰ (2s); Table 2) and the average amount of Zn released in Test B for both
pipette tips and tubes (= 6.0 ± 19.3 ng (2s); Appendix S2). Note that 6.0 ng is a minimum since the
average amount of Zn released by one contact with glove was calculated taking analyses below
detection limit equal to zero ng in Appendix S2. Then, we considered three different scenarios to
account for different degrees of contamination: (1) glove in contact with labware once during acid-
washing, sample preparation or analysis such as = 6.0 ng; (2) glove in contact with labware
twice such as = 2*6.0 ng; (3) glove in contact with labware three times such as =
3*6.0 ng. For each scenario, we calculated the effect of contamination on Zn isotopic compositions as
a function of the amount of Zn processed through chemistry and analysed on MC-ICP-MS (i.e., 30 ng
< < 500 ng, cf. Figure 6).
Accepted Article
This article is protected by copyright. All rights reserved.
Geological samples: In terrestrial and extra-terrestrial samples, the range of variability of
Zn isotopic compositions can be quite large. Extreme δ66Zn values of -7.4‰ and +6.4‰ have been
reported by Moynier et al. (2011) and Herzog et al. (2009) in meteorites (EL6 chondrites) and lunar
soils, respectively. Gloves have an average δ66Zn of ~ 0‰. As expected, Figure 6 shows that the
larger the difference is between δ66Zn of glove and sample, the greater is the effect of contamination.
For extreme sample compositions of -5‰ and +5‰, one contact between glove and labware during
washing, preparation or analysis is enough to shift the measured isotopic compositions beyond their
uncertainties (i.e., ± 0.1‰, 2s) given typical amounts of Zn processed through chemistry and
subsequently analysed on MC-ICP-MS (Figure 6a and h). For low amounts of Zn determined (< 50
ng), shifts in δ66Zn due to contact with gloves can reach ± 0.5‰ (one contact) and almost ± 2‰ in the
worst scenario (i.e., three contacts).
In geological materials, most δ66Zn values however cluster between -1‰ and +1‰ (see for example,
the values reported by Chen et al. (2016) for seventeen whole-rock reference materials). For these
more typical Zn isotopic compositions (cf. Figure 6c–f), the effect of glove contamination is less
important though shifts of more than ± 0.1‰ can occur if less than 200 ng is processed through
chemistry and three contacts between glove and labware occur during the analytical procedure.
Taking into account the whole range of possible isotopic compositions for terrestrial and extra-
terrestrial samples, we thus recommend that a minimum of 500 ng of Zn be processed through
chemistry and determined. Extreme precaution when acid-washing labware, preparing and analysing
samples is however required since any contact with glove would shift extreme Zn isotopic
compositions beyond the precision level taken here at ± 0.1‰. If duplicate analyses allow for a better
measurement precision on δ66Zn values, for example down to 0.03‰ (2s) as reported by Chen et al.
(2016), then higher amounts of Zn should be determined to make sure the effect of glove
contamination remains negligible.
Biological samples: The range of δ66Zn measured in biological samples varies between -
1‰ and +2‰ (e.g., Moynier et al. 2009, Balter et al. 2010, Jaouen et al. 2013a, b, Larner et al. 2015).
To our knowledge, no extreme values outside this range have been reported so far. Biological
Accepted Article
This article is protected by copyright. All rights reserved.
samples such as body fluids are often not very concentrated in Zn, for example no more than a few
μg g-1 in urine (Balter et al. 2010) and few hundred ng g-1 in cerebro-spinal fluids. Moreover, the
amount of sample available for research purposes is restricted to extremely low quantities; hence the
total amount of Zn available for isotopic determinations in a biological sample sometimes does not
exceed 50 ng. Figure 6 shows that for such low amounts of Zn, one contact between glove and
labware could significantly shift the measured Zn isotopic composition beyond the measurement
uncertainties (i.e., > ± 0.1‰, cf. Figure 6c to 6f). For instance, for a typical δ66Zn of +1‰ (Figure 6c),
one contact with a glove can bias δ66Zn by ~ 0.1‰ and three contacts by more than 0.3‰ if the
amount of Zn determined is below 50 ng. The effect of contamination becomes negligible, that is
within the measurement precision of ± 0.1‰, when the amount of Zn processed through chemistry
and subsequently determined is higher than 300 ng. We thus recommend that, whenever possible,
measurements of Zn isotopic compositions in biological samples be done on amounts > 300 ng. In
the case where the measurement precision is estimated to be better than ± 0.1‰, the amount of Zn
determined should be increased accordingly.
Conclusions
The new generation of mass spectrometers and the curiosity of scientists to determine an increasing
number of elements with high precision and sensitivity requires increasingly low procedural blanks
and control on potential exogenic contamination during sample preparation and analysis. The results
of this study clearly show that gloves are a primary and important source of contamination in clean
laboratories for both trace element and isotope studies. Of the sixty elements determined in nitrile,
latex, neoprene and vinyl gloves, very few appear to be safe from potential contamination. Soaking of
gloves in weak acid (0.4 mol l-1 HNO3 + 0.05 mol l-1 HF) at room temperature for several hours
released significant amounts of minor and trace elements, up to 17 mg of Zn, 200 μg of Mg, and 10 to
60 μg of Ti, Fe, Sr and Zr. Compared with vinyl gloves, nitrile, latex and neoprene gloves are much
more enriched in leachable elements, with the exception of In and Sn.
Accepted Article
This article is protected by copyright. All rights reserved.
The tests also indicate that most elements are easily transferred from the gloves to the sample by
simple contact of a pipette tip or the interior of a polypropylene tube. Acid-washing labware is in
particular a potential major source of contamination since it requires several handlings with gloves.
Zinc has by far the highest risk of being contaminated by gloves. A single dry contact between
labware and gloves releases an average of 6.0 ± 19.3 (2s) ng of Zn into the samples. Such
contamination has variable effects on the Zn isotopic composition of the sample depending on the
amount of Zn processed through chemistry and the difference between the Zn isotopic composition of
the studied sample and that of the glove. All tested gloves shared roughly the same δ66Zn (average
δ66Zn = +0.10 ± 0.32‰ (2s)). The lower the amount of Zn processed through chemistry, and the
higher the difference between δ66Zn of the sample and the glove, the greater is the effect of
contamination. Whatever the amount of Zn determined (from 30 to 500 ng), a single contact between
glove and labware is enough to bias the composition of a sample with extreme δ66Zn of -5 or +5‰
beyond the measurement precision taken at ± 0.1‰ here. A bias as large as ± 1.5‰ can be reached
for samples with extreme δ66Zn of -5 or +5‰ if the amount of Zn determined is low (< 50 ng) and
labware is touched repeatedly during washing, preparation or analysis. For typical δ66Zn values
between -1‰ and +2‰, a minimum of 300 ng of Zn should be determined to limit the potential bias
due to glove contamination below 0.1‰.
We recommend the following best practices to keep glove-derived contamination in control. (1)
Recognise that gloves can be a significant source of laboratory contamination. (2) Carry out every
step of washing procedures with clean tongs to minimise contact between gloves and labware (3)
Take extreme precautions when handling samples for trace element and Zn isotopic determinations
with any type of glove. (4) Monitor and minimise glove contamination using Zn contents in blanks
even if Zn is not in the analysis protocol. (5) For Zn isotopic determinations, measure whole
procedural blanks routinely (i.e., in every batch of samples) and calculate the effect of contamination
on each sample using Equation (2). If the isotopic bias exceeds the measurement precision, reject the
data.
Accepted Article
This article is protected by copyright. All rights reserved.
Acknowledgements
We would like to thank C. Chauvel and M. Horan for providing some of the tested gloves, C. Bosq for
ensuring optimal working conditions in the clean laboratory, T. Mock, P. Télouk and J-L. Piro for their
assistance with the mass spectrometers. We acknowledge two anonymous reviewers for their
constructive comments on the first version of the manuscript. This study was financed by the
Carnegie Institution for Science, the French Government Laboratory of Excellence initiative n°ANR-
10-LABX-0006, the Région Auvergne and the European Regional Development Fund. This is
Laboratory of Excellence ClerVolc contribution number 225.
References
Albarède F. and Beard B. (2004)
Analytical methods for non-traditional isotopes. Mineralogy and Geochemistry, 55, 113–152.
Aucour A.-M., Bedell J.-P., Queyron M., Magnin V., Testemale D. and Sarret G. (2015)
Dynamics of Zn in an urban wetland soil–plant system: Coupling isotopic and EXAFS approaches.
Geochimica et Cosmochimica Acta, 160, 55–69.
Balköse D., Gökçel H.İ. and Göktepe S.E. (2001)
Synergism of Ca/Zn soaps in poly (vinyl chloride) thermal stability. European Polymer Journal, 37,
1191–1197.
Balter V., Zazzo A., Moloney A.P., Moynier F., Schmidt O., Monahan F.J. and Albarède F. (2010)
Accepted Article
This article is protected by copyright. All rights reserved.
Bodily variability of zinc natural isotope abundances in sheep. Rapid Communications in Mass
Spectrometry, 24, 605–612.
Bridgestock L.J., Williams H., Rehkämper M., Larner F., Giscard M.D., Hammond S., Coles B.,
Andreasen R., Wood B.J., Theis K.J., Smith C.L., Benedix G.K. and Schönbächler M. (2014)
Unlocking the zinc isotope systematics of iron meteorites. Earth and Planetary Science Letters,
400, 153–164.
Büchl A., Hawkesworth C.J., Ragnarsdottir K.V. and Brown D.R. (2008)
Re-partitioning of Cu and Zn isotopes by modified protein expression. Geochemical Transactions, 9,
11, doi:10.1186/1467-4866-9-11.
Chauvel C., Bureau S. and Poggi C. (2011)
Comprehensive chemical and isotopic analyses of basalt and sediment reference materials.
Geostandards and Geoanalytical Research, 35, 125–143.
Chen S., Liu Y., Hu J., Zhang Z., Hou Z., Huang F. and Yu H. (2016)
Zinc isotopic compositions of NIST SRM 683 and whole-rock reference materials. Geostandards and
Geoanalytical Research, 40, 417–432.
Chen H., Savage P.S., Teng F.-Z., Helz R.T. and Moynier F. (2013)
Zinc isotope fractionation during magmatic differentiation and the isotopic composition of the bulk
Earth. Earth and Planetary Science Letters, 369–370, 34–42.
Accepted Article
This article is protected by copyright. All rights reserved.
Cloquet C., Carignan J., Lehmann M.F. and Vanhaecke F. (2008)
Variation in the isotopic composition of zinc in the natural environment and the use of zinc isotopes in
biogeosciences: A review. Analytical and Bioanalytical Chemistry, 390, 451–463.
Costas-Rodríguez M., Van Heghe L. and Vanhaecke F. (2014)
Evidence for a possible dietary effect on the isotopic composition of Zn in blood via isotopic analysis
of food products by multi-collector ICP-mass spectrometry. Metallomics, 6, 139–146.
Fang L., Song Y., Zhu X. and Zheng Q. (2009)
Influence of lanthanum stearate as a co-stabilizer on stabilization efficiency of calcium/zinc stabilizers
to polyvinyl chloride. Polymer Degradation and Stability, 94, 845–850.
Friel J.K., Mercer C., Andrews W.L. and Simmons B.R. (1996)
Laboratory gloves as a source of trace element contamination. Biological Trace Element Research,
54, 135–142.
Herzog G.F., Moynier F., Albarède F. and Berezhnoy A.A. (2009)
Isotopic and elemental abundances of copper and zinc in lunar samples, Zagami, Pele’s hairs, and a
terrestrial basalt. Geochimica et Cosmochimica Acta, 73, 5884–5904.
Hoffmann J. (1988)
Experience with the sources of contamination when preparing samples for the analysis of trace
Accepted Article
This article is protected by copyright. All rights reserved.
elements. Fresenius Zeitschrift für analytische Chemie, 331, 220–222.
Jaouen K., Gibert M., Lamboux A., Télouk P., Fourel F., Albarède F., Alekseev A.N., Crubézy E.
and Balter V. (2013a)
Is aging recorded in blood Cu and Zn isotope compositions? Metallomics, 5, 1016–1024.
Jaouen K., Pons M.-L. and Balter V. (2013b)
Iron, copper and zinc isotopic fractionation up mammal trophic chains. Earth and Planetary Science
Letters, 374, 164–172.
Larner F., Woodley L.N., Shousha S., Moyes A., Humphreys-Williams E., Strekopytov S.,
Halliday A.N., Rehkämper M. and Coombes R.C. (2015)
Zinc isotopic compositions of breast cancer tissue. Metallomics, 7, 112–117.
Maréchal C.N., Télouk P. and Albarède F. (1999)
Precise analysis of copper and zinc isotopic compositions by plasma-source mass spectrometry.
Chemical Geology, 156, 251–273.
Moynier F., Fujii T., Shaw A.S. and Le Borgne M. (2013)
Heterogeneous distribution of natural zinc isotopes in mice. Metallomics, 5, 693.
Moynier F., Paniello R.C., Gounelle M., Albarède F., Beck P., Podosek F. and Zanda B. (2011)
Accepted Article
This article is protected by copyright. All rights reserved.
Nature of volatile depletion and genetic relationships in enstatite chondrites and aubrites inferred from
Zn isotopes. Geochimica et Cosmochimica Acta, 75, 297–307.
Moynier F., Pichat S., Pons M.-L., Fike D., Balter V. and Albarède F. (2009)
Isotopic fractionation and transport mechanisms of Zn in plants. Chemical Geology, 267, 125–130.
Nieuwenhuizen P.J. (2001)
Zinc accelerator complexes: Versatile homogeneous catalysts in sulfur vulcanization. Applied
Catalysis A, 207, 55–68.
Ohno T., Shinohara A., Chiba M. and Hirata T. (2005)
Precise Zn isotopic ratio measurements of human red blood cell and hair samples by multiple
collector ICP-mass spectrometry. Analytical Sciences, 21, 425–428.
Palme H. and O'Neill H. StC. (2014)
Cosmochemical estimates of mantle composition. In: Holland H.D. and Turekian K.K. (eds),
Treatise on Geochemistry (2nd edition), vol. 3. Elsevier-Pergamon, 1–39.
Paniello R.C., Day J.M.D. and Moynier F. (2012)
Zinc isotopic evidence for the origin of the Moon. Nature, 490, 376–379.
Pichat S., Douchet C. and Albarède F. (2003)
Accepted Article
This article is protected by copyright. All rights reserved.
Zinc isotope variations in deep-sea carbonates from the eastern equatorial Pacific over the last 175
ka. Earth and Planetary Science Letters, 210, 167–178.
Stenberg A., Malinovsky D., Öhlander B., Andrén H., Forsling W., Engström L.-M., Wahlin A.,
Engström E., Rodushkin I. and Baxter D.C. (2005)
Measurement of iron and zinc isotopes in human whole blood: Preliminary application to the study of
HFE genotypes. Journal of Trace Elements in Medicine and Biology, 19, 55–60.
Van Heghe L., Engström E., Rodushkin I., Cloquet C. and Vanhaecke F. (2012)
Isotopic analysis of the metabolically relevant transition metals Cu, Fe and Zn in human blood from
vegetarians and omnivores using multi-collector ICP-mass spectrometry. Journal of Analytical
Atomic Spectrometry, 27, 1327.
Supporting information
The following supporting information is available online:
Appendix S1. Results of Test A (gloves)
Appendix S2. Results of Test B (tube or pipette)
Appendix S3. Comparison of duplicate analyses (Test A)
Appendix S4. Comparison of zinc isotopic compositions.
This material is available as part of the online article from:
http://onlinelibrary.wiley.com/doi/10.1111/ggr.00000/abstract
(This link will take you to the article abstract).
Accepted Article
This article is protected by copyright. All rights reserved.
Figure captions
Figure 1. Illustration showing how Tests A and B were performed.
Figure 2. Amount of trace and minor elements released by gloves when soaked in 20 ml of 0.4 mol l-1
HNO3 + 0.05 mol l-1 HF for 40 hours (results of Test A) normalised to the amount of elements present
in (a) 1 mg of basalt BHVO-2, (b) 1 mg of chondrite, (c) 1 g of natural river water SLRS-4 and (d) 30
mg of bovine serum NIST SRM 1598. Enrichment factors were calculated relative to the preferred
values compiled by GeoReM for reference materials BHVO-2, SLRS-4 and NIST SRM 1598.
Reference values were taken from Palme and O’Neill (2014) for CI chondrites. For panel (c) and (d),
only a few elements are shown because the mass fractions of other elements in SLRS-4 and NIST
SRM 1598 are not available in the literature. The grey field represents values below detection limits
defined following IUPAC guidelines as the average number of counts measured in blanks plus three
times the standard deviation (see section Analyses for trace element contents for more details). When
the amount of element detected in a solution was below the detection limit (< DL in Appendix S1), it is
represented as equal to the detection limit.
Figure 3. Binary plots showing the amount of chemical elements released by gloves when soaked in
20 ml of 0.4 mol l-1 HNO3 + 0.05 mol l-1 HF for 40 hours (results of Test A). The grey field represents
values below detection limits defined following IUPAC guidelines as the average number of counts
measured in blanks plus three times the standard deviation (see section Analyses for trace element
content for more details). When the amount of element detected in a solution was below the detection
limit (< DL in Appendix S1), it is represented as equal to the detection limit: this is the case of some
vinyl glove in panels (e) and (g).
Figure 4. Amount of trace and minor elements (ng) released by gloves by short contact with (a) the
interior of a polypropylene tube and (b) a 1 ml pipette tip (results of Test B). The grey field represents
values below detection limits defined following IUPAC guidelines as the average number of counts
measured in blanks plus three times the standard deviation (see section Analyses for trace element
Accepted Article
This article is protected by copyright. All rights reserved.
content for more details). When the amount of element detected in a solution was below the detection
limit (< DL in Appendix S2), it is represented as equal to the detection limit in the two panels.
Figure 5. Zinc isotopic compositions (δ66Zn) of the tested gloves. Error bars are fixed at ± 0.12‰
based on complete duplicate analyses.
Figure 6. Effect of contamination by gloves on Zn isotopic compositions as a function on the amount
of Zn processed through chemistry and determined by mass spectrometry. For this calculation, we
used the average δ66Zn of all tested gloves (δ66Zn = +0.10‰) and assumed that the amount of Zn
released by gloves for one contact was equal to 6.0 ng (average amount of Zn released in Test B),
2*6.0 ng for two contacts and 3*6.0 ng for three contacts. The green field shows the isotopic
composition of the sample and its measurement precision that is fixed at ± 0.1‰ (2s) for this figure.
More information about the calculation can be found in the main text.
Accepted Article
This article is protected by copyright. All rights reserved.
Accepted Article
This article is protected by copyright. All rights reserved.
Accepted Article
This article is protected by copyright. All rights reserved.
Accepted Article
This article is protected by copyright. All rights reserved.
Accepted Article
This article is protected by copyright. All rights reserved.
Accepted Article
This article is protected by copyright. All rights reserved.