PAPER www.rsc.org/jem | Journal of Environmental Monitoring

Publ

ishe

d on

09

Sept

embe

r 20

08. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

28/1

0/20

14 0

9:17

:34.

View Article Online / Journal Homepage / Table of Contents for this issue

A study of the bio-accessibility of welding fumes†

Balazs Berlinger,*ab Dag G. Ellingsen,b Miklos Naray,a Gyula Zaraycd and Yngvar Thomassenb

Received 18th April 2008, Accepted 20th August 2008

First published as an Advance Article on the web 9th September 2008

DOI: 10.1039/b806631k

The respiratory bio-accessibility of a substance is the fraction that is soluble in the respiratory

environment and is available for absorption. In the case of respiratory exposure the amount of

absorbed substance plays a main role in the biological effects. Extensive bio-accessibility studies have

always been an essential requirement for a better understanding of the biological effects of different

workplace aerosols, such as welding fumes. Fumes generated using three different welding techniques,

manual metal arc (MMA) welding, metal inert gas (MIG) welding, and tungsten inert gas (TIG)

welding were investigated in the present study. Each technique was used for stainless steel welding.

Welding fumes were collected on PVC membrane filters in batches of 114 using a multiport air sampler.

Three different fluids were applied for the solubility study: deionised water and two kinds of lung fluid

simulants: lung epithelial lining fluid simulant (Gamble’s solution) and artificial lung lining fluid

simulant (Hatch’s solution). In order to obtain sufficient data to study the tendencies in solubility

change with time, seven different leaching periods were used (0.5, 1, 2, 4, 8, 16, 24 h), each of them with

three replicates. The effect of dissolution temperature was also studied. The total amounts of selected

metals in the three different welding fumes were determined after microwave-assisted digestion with the

mixture of aqua regia and hydrofluoric acid. The most obvious observation yielded by the results is that

the solubility of individual metals varies greatly depending on the welding technique, the composition

of the leaching fluid and leaching time. This study shows that the most reasonable choice as a media for

the bio-assessment of solubility might be Hatch’s solution by a dissolution time of 24 h.

Introduction

Exposure to welding fumes may have a significant impact on

human health.1 Over the past decades numerous studies have

addressed the effects of welding fumes on the pulmonary func-

tion of workers, and several different pulmonary diseases have

been reported.2–11 Increased mortality due to ischemic heart

disease among welders has also been observed recently.12

Welding fume constituents such as aluminium (Al) and

manganese (Mn) have been suspected of causing psychiatric and

neurological symptoms in exposed workers in specific occupa-

tions. Although adverse neuropsychological and neurological

health effects are well known among workers with high Mn

exposures in mining, ore-processing and ferroalloy production,13

the risks among welders are still poorly understood. Therefore

several investigations have been made in recent years to study the

aChemical Laboratory, Hungarian Institute of Occupational Health, P.O.Box 22, H-1450 Budapest, Hungary. E-mail: berlingerb@omfi.hu;Fax: +36 1476 1374; Tel: +36 1476 1185bNational Institute of Occupational Health, P.O. Box 8149 Dep, N-0033Oslo, Norway. E-mail: yngvar.thomassen@stami.no; Fax: +4723195206; Tel: +47 23195320cDepartment of Analytical Chemistry, Eotvos Lorand University,Budapest, Hungary. E-mail: zaray@ludens.elte.hu; Fax: +36 1372 2608;Tel: +36 13722607dCooperative Research Centre of Environmental Sciences, Budapest,Hungary. E-mail: zaray@ludens.elte.hu; Fax: +36 1372 2608; Tel: +3613722607

† Presented at AIRMON 2008, January 28–31, 2008, Dr Holms Hotel,Geilo, Norway.

1448 | J. Environ. Monit., 2008, 10, 1448–1453

possible neurological health effects among welders exposed to

Mn-containing welding fumes.14–17 Although manganism has

repeatedly been observed in highly exposed welders, some

authors do not find it justifiable to associate welding with clinical

neurotoxicity based on the scant exposure–response data avail-

able for welders.16 However, a more recent study indicates dose–

response relationship between neurobehavioral functions and

welders’ exposure to welding fumes containing high concentra-

tion of Mn.17

For a better understanding of the uptake and biological effects

of welding fumes it is important to find the relationship between

exposure to the different welding fume components, especially

metals, and the diseases occurring among welders. In these

attempts the estimation of bioavailability can be useful. Metal

bioavailability can be defined as the extent of systemic absorption

of metal species. The amount of metal ion binding the target would

represent its true bioavailability and thus the toxic dose.18

Assessing metal bioavailability in humans requires bio-monitoring

strategies. To evaluate metal bioavailability, studies must be

conducted using in vivo models.19 Alternatively, the potential

bioavailability of metal compounds or species can be investigated

by measuring their solubility in artificial human tissue fluids or in

samples of natural tissue fluids such as human serum. In this

context, the availability of metal ions from a particular metal

compound for absorption can be measured when dissolved in body

fluids or their in vitro surrogates. The amount of metals dissolved

in this procedure is defined as the bio-accessible fraction.20

Bio-accessibility experiments can not only be applied when

there is no opportunity to perform bio-monitoring or tests with

animals, but they can also provide important preliminary

This journal is ª The Royal Society of Chemistry 2008

Table 1 Composition and pH of extraction fluids

Extraction fluid pH Composition (per litre DI water)

Gamble’s solution 7.4 Magnesium chloride hexahydrate (0.2033g), sodium chloride (6.0193 g), potassiumchloride (0.2982 g), dibasic sodiumphosphate (0.1420 g), sodium sulfate(0.0710), calcium chloride dihydrate(0.3676), sodium acetate trihydrate(0.9526), sodium bicarbonate (2.6043 g),sodium citrate dehydrate (0.0970 g)

Hatch’s solution 7.4 Calcium chloride (0.2251 g), magnesiumchloride hexahydrate (0.21 g), magnesiumsulfate (0.0342 g), potassium chloride (0.37g), potassium dihydrogen phosphate (0.03g), sodium bicarbonate (2.27 g), sodiumchloride (7.0 g), dibasic sodium phosphate(0.1196 g), D-glucose (1.0 g), phosphatidylcholine (10 g), a-tocopherol (0.001 g), uricacid (0.025 g), serum albumin (10 g),lysosyme (2.5 g), apo-transferrin (0.2 g),ascorbate (0.05 g), glutathione (0.05 g)

Publ

ishe

d on

09

Sept

embe

r 20

08. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

28/1

0/20

14 0

9:17

:34.

View Article Online

information to a bioavailability study. Ellingsen et al.21

investigated if there was any association between Mn concen-

trations in workroom air and Mn concentrations in whole

blood in welders. Pearson’s correlation coefficient of 0.31

(p < 0.01) was calculated between blood-Mn and Mn in the

workroom air that was collected the day before blood

sampling. However, correlation might improve if the deposited

dose of bio-accessible Mn could also be taken into account.

Several recent publications support the theory that the solu-

bility of metals has a major role in inducing systemic biological

effects. In another study Ellingsen et al.22 reported an associ-

ation between urinary Mn levels of workers in the Mn alloy

production industry and soluble respirable Mn air concentra-

tions, suggesting that the solubility of Mn is important for the

uptake. This report is compatible with the study of Roels

et al.,23 who showed that rats administered soluble MnCl2intratracheally had substantially higher Mn concentrations in

striatum (in the brain), the target tissue, than rats administered

less soluble MnO2.

Animal studies have also indicated the importance of solubility

in the local effects of welding fumes on test animals. In 2003

Taylor et al.24 found that different welding fumes caused varied

responses in the lungs of rats, correlated to their metal compo-

sition and ability to produce free radicals. However, both soluble

and insoluble fractions of the welding fume generated during

manual metal arc (MMA) welding were required to produce

most effects, which indicates that the responses were not exclu-

sively dependent on the soluble metal fractions. A later study by

McNeilly et al.25 suggested that the soluble amounts of transition

metals present in welding fumes were responsible for inflamma-

tion in the lungs of rats, because removal of the soluble fraction

by chelation or washing the welding fume particles abolished

their ability to cause inflammation.

Based on these studies it is likely that the solvent in which

solubility is determined influences the results. Therefore the

main aim of the present work is to provide information about

the bio-accessibility of different metals in welding fumes using

different media for dissolution. Three different fluids were

applied for the solubility study: deionised water and two kinds

of lung fluid simulant, lung epithelial lining fluid simulant

(Gamble’s solution)26 and artificial lung lining fluid simulant

(Hatch’s solution).27 Gamble’s solution has been used for more

than 30 years for determining the solubility of nucleotides and

man-made fibres. It represents the interstitial fluid deep within

the lung. Citrate is used in the place of proteins and acetate to

represent organic acids. Hatch’s solution contains some proteins

and enzymes in addition. The main reason for the selection of

Gamble’s and Hatch’s solutions was that these are the most

frequently used lung fluid simulants. Since a solubility study

using pure water is more attractive from an analytical point of

view, in this work we have also compared the solubility of

welding fume components in water to their solubility in Gam-

ble’s and Hatch’s solutions. Three different welding techniques:

MMA welding, metal inert gas (MIG) welding and tungsten

inert gas (TIG) welding were investigated during the study.

These are the most commonly used welding techniques in

industrial processes. The selected elements for examination were

the main metal components of the welding materials and the

welded steels.

This journal is ª The Royal Society of Chemistry 2008

Experiment

Air sampling

Each type of welding electrode that was applied for welding

contained 19% of chromium (Cr), 12% of nickel (Ni), 3% of

molybdenum (Mo) and approximately 0.8% of silicon (Si) and

0.8% of Mn, but they were purchased from different suppliers

(Sandvik Materials Technology, Sandviken, Sweden; Avesta

Welding AB, Avesta, Sweden). Each technique was used for

stainless steel welding. For the bio-accessibility study welding

fumes were collected on membrane filters in batches of 114 using

a multiport air sampler, which has been developed at the

National Institute of Occupational Health (NIOH), Oslo, Nor-

way, for the collection of parallel samples of workplace aerosols.

It has been successfully used by a number of laboratories to

prepare certified reference materials of metals on filters, quality

assurance filter samples and test samples for validating analytical

methods for harmful substances in workplace air.28,29 The mul-

tiport air sampler was equipped with 25 mm Millipore filter

cassettes containing 5.0 mm pore-size Millipore PVC membrane

filters (Millipore Corp., Billerica, USA). The length of sampling

times varied between 30 and 180 min depending on the intensity

of fume generation.

Sample preparation and analytical procedure

Chemicals. The deionised (DI) water used throughout was

prepared by a Milli Q System (18.2 MU cm, Millipore Corp.,

Billerica, USA). Nitric, hydrofluoric and hydrochloric acids for

trace analysis were used (Fluka, Buchs, Switzerland). Each

inorganic salt used in the preparation of the biological fluid

simulants, D-glucose and ascorbic acid were of analytical grade

quality (Merck, Darmstadt, Germany). Phosphatidyl choline,

a-tocopherol, uric acid, serum albumin, lysosyme, apo-trans-

ferrin and glutathione were purchased from Sigma (St. Louis,

USA). The compositions of the lung fluid simulants are listed in

Table 1. The biological fluid simulants were prepared by first

dissolving inorganic salts in DI water and then adjusting the pHs.

J. Environ. Monit., 2008, 10, 1448–1453 | 1449

Table 2 Concentrations (in mg mg�1) of elements in the different weldingfumes

TIG/mg mg�1 MIG/mg mg�1 MMA/mg mg�1

Ag <0.3 1.1 0.7Al 2.5 6.8 5.5Ca <48 <9 34.4Cr 12.6 64.2 30.6Cu 2.3 4.9 0.5Fe 43.7 203 185K <15 11.1 203Li <0.2 0.2 5.2Mg <2.4 1.0 1.3Mn 59.5 77.0 70.1Mo 1.3 12.1 1.7Na 9.0 3.8 59.7Ni 5.0 35.1 5.6Pb <1.7 238 30.4Si 8.7 16.4 29.2Sn <1.8 1.7 0.3Ti 0.5 0.6 11.0V <0.8 <0.2 0.1Zn <10 5.0 1.3Zr <0.2 <0.03 0.03

Collected weldingfume in mg (n ¼ 25)

0.14a � 0.03b 0.75 � 0.02 1.47 � 0.04

a Arithmetic mean. b Standard deviation of the mean.

Publ

ishe

d on

09

Sept

embe

r 20

08. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

28/1

0/20

14 0

9:17

:34.

View Article Online

The organic compounds were dissolved in the final step to

produce the Hatch’s solution.

Leaching was done within five hours after sampling in 50 mL

volume VectaSpin 20TM polypropylene (PP) centrifuge tubes

with 25 mL filter cup inserts equipped with 0.2 mm nylon

membranes (Whatman International Ltd., Maidstone, England,

UK) by adding 10 mL of the leaching solutions to the PVC filter

sample. In order to get sufficient data to calculate dissolution

curves, seven different leaching periods were used (0.5, 1, 2, 4, 8,

16, 24 h), each of them with three replicates. The choice of 24 h as

the longest leaching time was based on the finding of Semmler-

Behnke et al.,30 who reported in a recent publication dealing with

the elimination of inhaled nanoparticles from the alveolar region

that 90% of the nanoparticles (17–20 nm) disappeared from the

epithelium due to sequential lung retention and clearance within

24 h. Therefore we may presume that the uptake of metals from

the epithelial region will take place dominantly in the first 24 h

period. Each tube containing the welding fume filter and the

leaching solution was placed in a laboratory oven set to

a temperature of 37 � 1 �C for the specified leaching time.

Leaching with DI water was also done at room temperature with

the same leaching periods as described above. The obtained

solutions were then filtered by centrifugation at 3000 rev min�1

with a 12 tube capacity centrifuge (Model 4K15, Sigma, Osterode

am Harz, Germany). As an internal spectrometric standard

known quantity of beryllium chloride was added to the portions

for the inductively coupled plasma optical emission spectro-

metric (ICP-OES) measurements. To ensure that the metal

species were kept in solution after leaching, 2 mL of aqua regia

was also added to each solution after centrifugation.

For the determination of the total amount of metals in the

welding fume samples, 10 – 12 filters were digested from each

kind of welding fume in Teflon autoclaves with a mixture of 2 mL

aqua regia and 0.2 mL hydrofluoric acid. A known quantity of

beryllium chloride solution was added as an internal spectro-

metric standard before acid digestion. Sixteen autoclaves were

heated in a microwave unit simultaneously (MLS 1200, Teflon

Container SV140, 10 bar, Milestone, Sorisole, Italy). The

dilution volume (with DI water) for the digests was 14 mL.

Metal contents of the extracts and the digests were measured

by ICP-OES (Perkin Elmer Optima 3000, Perkin Elmer Inc.,

Waltham, USA). Details of the ICP-OES method were as

described previously.31 Background corrections were made by

using field and laboratory blanks which were sufficiently below

the concentration levels studied.

Since welding fumes are formed with an individual particle size

of <1 mm, or even <0.1 mm, it was reasonable to test the filtration

efficiency of the centrifuge tube filters for ultrafine/nanoparticles.

Nanomaterials (titanium dioxide (TiO2) with different particles

sizes: 5 nm, 10 nm, 30–40 nm) supplied by NanoAmor (Houston,

USA) were used for the testing. Portions of 0.5–2.0 mg of the

TiO2 powders were weighed into the filter cup inserts and 10 mL

of DI water was added to each of them to suspend the particles.

The centrifugations were done at 3000 rev min�1. The aqueous

centrifugates were analysed by ICP-OES for their titanium (Ti)

contents. Filtration efficiencies for the different nanopowders

were calculated by comparing the amounts of Ti found after

filtration to the amounts added. Filtration efficiencies were

greater than 99.99% for all powders, even for 5 nm TiO2.

1450 | J. Environ. Monit., 2008, 10, 1448–1453

For quality control of the analytical procedure used for

assessing the elemental compositions of the welding fume filters,

commercially available 37 mm cellulose ester membrane filter

samples (Standard Reference Material A 2, NIOH, Oslo,

Norway) were analysed, which were prepared by spiking each

filter with an aqueous solution containing 24 elements with

concentrations gravimetrically traceable to ultrapure metals or

to stoichiometrically well defined oxides. The recoveries for all

the measured metals were between 98 and 102%.

Results and discussion

The collected masses and the elemental compositions of the

different welding fumes are summarised in Table 2. The masses

on the filters in case of TIG welding were quite low in spite of the

relatively long sampling time. This resulted in reduced absolute

detection limits of the different elements in TIG welding fumes,

especially when the solubility of metals in different media was

examined. The dominant elements in TIG and MIG welding

fumes were iron (Fe), Mn, Cr, Ni and Si, while the MMA fume

contained considerable amounts of potassium (K) and calcium

(Ca) originating from the coating of the MMA welding rod. In

case of MIG welding lead (Pb) appeared in a large amount. It

turned out that Pb was present in the welding fume as

a contaminant, because some soldering operations using

Pb-containing materials had taken place at the table where the

welding was later performed. In this regard solubility results for

Pb in different media may be useful, since such Pb contamination

can also happen in a real welding situation.

The most obvious observation from the results is that the

solubility of each metal varies greatly depending on the welding

technique, the composition of the leaching fluid and leaching

time. Results for the most important elements are summarised in

Tables 3–9. It has to be mentioned that the relative standard

This journal is ª The Royal Society of Chemistry 2008

Table 3 Dissolved amount of Fe in % compared to the total amount of Fe in the three different welding fumes

Extraction time/h

Water at 20 �C Water at 37 �C Gamble’s solution Hatch’s solution

TIGa MIGa MMAa TIGa MIGa MMAa TIGa MIGa MMAa TIGa MIGa MMAa

0.5 3.5 <0.11 3.0 0.85 <0.11 1.8 1.0 0.14 0.51 18 0.96 0.821 1.2 <0.11 2.9 2.0 <0.11 1.3 0.47 0.08 0.50 17 0.92 0.772 1.0 <0.11 2.7 0.52 <0.11 0.59 0.10 0.11 0.25 18 1.0 0.764 1.1 <0.11 3.4 1.4 <0.11 0.58 0.41 0.08 0.24 17 1.1 0.768 2.7 <0.11 1.5 0.58 <0.11 0.28 0.16 0.06 0.20 19 1.4 0.8116 1.1 <0.11 1.1 0.35 <0.11 0.13 0.04 0.07 0.05 20 2.0 0.8324 1.0 <0.11 0.81 0.93 <0.11 0.35 0.33 0.03 0.09 20 3.1 1.1

a Average of the three parallel extractions.

Fig. 1 Solubility of Mn in % in Hatch’s solution compared to the total

amount of Mn for the different welding techniques.

Publ

ishe

d on

09

Sept

embe

r 20

08. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

28/1

0/20

14 0

9:17

:34.

View Article Online

deviations (RSD) for the three replicates were typically between 5

and 20%.

The solubility of Fe was very low in all welding techniques and

leaching solutions, except when Hatch’s solution was used for

dissolving TIG welding fumes (Table 3). This may partly be

explained by the very low solubility of Fe(OH)2 and Fe(OH)3

likely to be formed at pHs of the applied solutions. In spite of the

low solubility values in percentages it is clear that the solubility of

Fe slightly decreased at higher temperature in each welding

technique, but considerably higher in Hatch’s solution compared

to other solutions in TIG and MIG welding, while in the case of

MMA welding no such difference was observed. The first

observation can be a consequence of the temperature dependence

of the solubility of the different components. The second

observation may be explained with possibly different oxidation

states of dissolved Fe in the different welding fumes. This may

suggest that the dissolved Fe can be present in the trivalent form

in solution in the case of MIG and TIG welding, since enzymes—

like apo-transferrin—in Hatch’s solution can bind Fe(III)

strongly, while in MMA welding it is more likely that Fe is in

the divalent form, so the solubility constant of Fe(OH)2 will

determine the dissolved amount of Fe(II).

There was a significant increase with time in the solubility of

Mn in most cases (Table 4). For example, in the case of TIG

welding the dissolved amount of Mn in Hatch’s solution changed

from 11% to 37% during the 24 h period (Fig. 1). At the same

time the solubility of Mn in MMA fumes decreased with time in

Gamble’s solution and it was quite constant in Hatch’s solution.

A larger amount of Mn was dissolved at a water temperature of

37 �C than at 20 �C. These very different behaviours of Mn

Table 4 Dissolved amount of Mn in % compared to the total amount of M

Extraction time/h

Water at 20 �C Water at 37 �C

TIGa MIGa MMAa TIGa MIGa M

0.5 3.9 0.93 9.0 4.6 0.931 6.3 1.0 9.4 7.0 1.02 10 1.2 9.7 8.5 1.4 14 9.3 1.4 11 14 1.6 18 11 2.0 11 13 2.0 116 11 1.9 11 21 2.3 124 17 2.0 12 23 3.0 1

a Average of the three parallel extractions.

This journal is ª The Royal Society of Chemistry 2008

between welding techniques and leaching methods make it

challenging to choose the appropriate solvent and adequate

dissolution time to be used in bioavailability studies for the

estimation of the bio-accessible doses of welding fumes.

Much larger amounts of soluble Cr were found in the case of

MMA welding than in TIG or MIG welding (Table 5). It is well

known that MMA welding fumes contain soluble Cr as chro-

mates, and the amount of hexavalent Cr can be up to 90–100% of

the total amount of Cr in certain welding fumes.32 In our

experiment Gamble’s and Hatch’s solutions dissolved larger

amounts of Cr in each welding fume than did water at any

temperature. If the two biological fluid simulants are compared,

Gamble’s solution is found to dissolve more Cr than Hatch’s

solution in the case of MMA welding, while TIG and MIG

welding present inverse situations.

n in the three different welding fumes

Gamble’s solution Hatch’s solution

MAa TIGa MIGa MMAa TIGa MIGa MMAa

9.2 1.7 0.64 5.1 11 2.7 6.09.8 1.8 0.65 5.0 8.0 3.4 5.90 1.3 1.0 4.2 15 4.7 6.11 2.1 1.8 3.9 11 6.2 6.13 1.4 2.7 3.2 17 7.8 6.45 1.0 1.9 1.2 36 12 6.26 1.4 1.1 1.0 37 14 6.8

J. Environ. Monit., 2008, 10, 1448–1453 | 1451

Table 5 Dissolved amount of Cr in % compared to the total amount of Cr in the three different welding fumes

Extraction time/h

Water at 20 �C Water at 37 �C Gamble’s solution Hatch’s solution

TIGa MIGa MMAa TIGa MIGa MMAa TIGa MIGa MMAa TIGa MIGa MMAa

0.5 <2.1 <0.1 43 <2.1 <0.1 45 4.3 1.4 61 12 1.7 531 <2.1 <0.1 43 <2.1 <0.1 45 4.1 1.2 59 9.4 1.7 522 <2.1 <0.1 42 <2.1 <0.1 44 2.9 1.3 61 12 1.8 524 <2.1 <0.1 41 <2.1 <0.1 43 2.9 1.5 55 9.5 1.7 528 <2.1 <0.1 41 <2.1 <0.1 44 <2.3 1.5 62 11 2.1 5216 <2.1 <0.1 40 <2.1 <0.1 46 <2.3 1.6 62 14 2.1 5024 <2.1 <0.1 39 <2.1 <0.1 47 2.3 1.5 60 11 2.5 52

a Average of the three parallel extractions.

Publ

ishe

d on

09

Sept

embe

r 20

08. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

28/1

0/20

14 0

9:17

:34.

View Article Online

No evaluation for Ni could be done in the TIG welding fume

because all the measured Ni values were lower than the detection

limit for Ni. If the dissolved Ni amounts are compared in MIG

and MMA welding fumes, significantly larger amounts of

dissolved Ni are found in MMA than in MIG welding in each

solvent (Table 6). If we look at the tendency of changes in the

dissolved Ni amounts, an increase with leaching time can be

observed in the case of MIG welding, while solubility of Ni in the

MMA welding fume seems to be more constant. Since there are

no large differences between the values in the different solutions

for a certain welding fume, theoretically each solvent could be

appropriate for the determination of bio-accessible Ni in the

examined welding fumes from an analytical point of view.

There is a very large difference in the solubility of Mo in water

as compared to biological fluid simulants (Table 7). Five to ten

times higher amounts of soluble Mo were found in biological

Table 6 Dissolved amount of Ni in % compared to the total amount of Ni

Extraction time/h

Water at 20 �C Water at 37 �C

MIGa MMAa MIGa M

0.5 2.1 15 2.2 11 2.3 15 2.5 12 2.6 15 3.3 14 3.0 16 3.6 18 3.6 15 3.9 116 3.6 15 3.5 124 3.6 15 4.3 1

a Average of the three parallel extractions.

Table 7 Dissolved amount of Mo in % compared to the total amount of M

Extraction time/h

Water at 20 �C Water at 37 �C

MIGa MMAa MIGa M

0.5 0.2 7.5 0.3 11 0.1 11 0.32 0.1 15 0.14 0.3 9.6 0.38 0.5 4.1 0.116 0.2 4.0 0.324 0.2 2.9 0.3

a Average of the three parallel extractions.

1452 | J. Environ. Monit., 2008, 10, 1448–1453

fluid simulants than in water in the case of MMA welding, while

these differences were almost 100-fold in MIG welding fume. Ti

solubility in the case of MMA welding proved to be completely

different since water could dissolve 5–10 times more Ti than

biological fluid simulants (Table 8). The solubility decreased with

time in each solvent. It may be suggested that the soluble forms

for Mo and Ti could be molybdate (MoO42�) and titanate

(TiO32�), respectively, but this suggestion could only be

confirmed by an another analytical technique, e.g. ion chroma-

tography (IC).

The solubility of Pb was approximately 10 times higher in

water than in biological fluid simulants in the case of MIG

welding (Table 9). This observation may be explained by the

pH-dependence of the solubility of Pb(OH)2 and/or some salting

out of Pb in biofluids due to the high concentration of chloride,

bicarbonates and phosphates. The tendencies of solubility were

in the MIG and MMA welding fumes

Gamble’s solution Hatch’s solution

MAa MIGa MMAa MIGa MMAa

5 1.7 17 2.4 145 1.7 16 2.5 145 2.0 17 3.3 156 1.8 15 3.7 157 1.9 17 3.8 178 1.6 17 5.0 149 1.7 16 6.3 19

o in the MIG and MMA welding fumes

Gamble’s solution Hatch’s solution

MAa MIGa MMAa MIGa MMAa

1 24 66 30 588.4 20 63 33 552.3 24 69 39 534.6 25 59 39 531.7 24 69 40 540.5 28 67 45 561.1 25 67 57 58

This journal is ª The Royal Society of Chemistry 2008

Table 8 Dissolved amount of Ti in % compared to the total amount ofTi in the MMA welding fume

Extractiontime/h

Water at20 �Ca

Water at37 �Ca

Gamble’ssolutiona

Hatch’ssolutiona

0.5 26 21 2.5 4.71 27 18 1.8 4.12 26 13 0.8 4.14 24 8.7 0.7 3.98 19 4.8 0.4 3.916 17 2.6 0.2 3.024 11 2.2 0.2 3.1

a Average of the three parallel extractions.

Table 9 Dissolved amount of Pb in % compared to the total amount ofPb in MIG welding fume

Extractiontime/h

Water at20 �Ca

Water at37 �Ca

Gamble’ssolutiona

Hatch’ssolutiona

0.5 16 18 2.6 1.61 19 20 2.5 1.02 20 23 3.2 1.34 18 26 2.1 1.78 21 24 0.1 1.916 25 24 0.3 1.924 22 27 0.4 2.6

a Average of the three parallel extractions.

Publ

ishe

d on

09

Sept

embe

r 20

08. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

28/1

0/20

14 0

9:17

:34.

View Article Online

increasing in the case of water at different temperatures and

decreasing in the case of Gamble’s solution. No clear tendency

was observed in the case of Hatch’s solution.

Concluding remarks

This study showed that the solubility of each metal varies greatly

depending on the welding technique, the composition of the

leaching fluid and leaching time. However, if the bio-accessible

part of the airborne exposure is to be estimated among welders

for the study of the relationship between air exposure and whole

blood/urine concentrations, it is essential to choose the most

biologically relevant leaching solution and dissolution time.

According to results of the presented study, the most reasonable

choice might be Hatch’s solution as leaching solution and

a 24 hour dissolution time for the following reasons: (1) based on

its composition, Hatch’s solution is the most similar to the lung

lining fluid; (2) in most cases the highest concentrations of the

most important metals (Fe, Mn, Cr, Ni) in the investigated

welding fumes were found in Hatch’s solution after a 24 hour

period. This finding combined with the observation of other

researchers that 90% of the particles are cleared from the

epithelial region within the first 24 hour period30 supports the

suggestion that the most adequate dissolution time is 24 hours.

This journal is ª The Royal Society of Chemistry 2008

References

1 J. M. Antonini, Crit. Rev. Toxicol., 2003, 33, 61.2 W. S. Beckett, P. E. Pace, S. J. Sferlazza, G. D. Perlman, A. H. Chen

and X. P. Xu, J. Occup. Environ. Med., 1996, 38, 1229.3 A. Sobaszek, C. Boulenguez, P. Frimat, H. Robin, J. M. Haguenoer

and J.-L. Edme, J. Occup. Environ. Med., 2000, 42, 923.4 J. R. Beach, J. H. Avery, C. L. Bromly, R. J. Ward, E. H. Walters,

S. C. Stenton and D. J. Hendrick, Am. J. Respir. Crit. Care Med.,1996, 154, 1394.

5 C. Vogelmeier, G. Konig, K. Benze and G. Fruhmann, Chest, 1987,92, 946.

6 J. E. Cotes, E. L. Feinman, V. J. Male, E. S. Remie andC. A. G. Wickham, Br. J. Ind. Med., 1989, 46, 292.

7 A. T. Doig and A. G. McLaughlin, Lancet, 1936, 1, 771.8 J. A. Roesler and H. J. Woitowiltz, Eur. Resp. J., 1996, 9, 220S.9 D. Coggon, H. Inskip, P. Winter and B. Pannett, Lancet, 1994, 344,

41.10 N. Becker, J. Claude and R. Frentzel-Beyme, Scand. J. Work,

Environ. Health, 1984, 11, 75.11 B. Sjogren, A. Gustavsson and L. Hedstorm, Scand. J.Work, Environ.

Health, 1987, 13, 247.12 B. Sjogren, F. Gyntelberg and B. Hilt, Scand. J. Work, Environ.

Health, 2006, (Suppl. 2), 50.13 J. Donaldson, NeuroToxicology, 1987, 8, 451.14 R. M. Park, R. M. Bowler, D. E. Eggerth, E. Diamond, K. J. Spencer,

D. Smith and R. Gwiazda, NeuroToxicology, 2006, 27, 373.15 R. M. Bowler, H. A. Roels, S. Nakagawa, M. Drezgic, E. Diamond,

R. Park, W. Koller, R. P. Bowler, D. Mergler, M. Bouchard,D. Smith, R. Gwiazda and R. L. Doty, J. Occup. Environ. Med.,2007, 64, 167.

16 A. B. Santamaria, C. A. Cushing, J. M. Antonini, B. L. Finley andF. S. Mowat, J. Toxicol. Environ. Health, Part B, 2007, 10, 417.

17 D. G. Ellingsen, R. Konstatinov, R. Bast-Pettersen, L. Merkurjeva,M. Chashchin, Y. Thomassen and V. Chashchin, NeuroToxicology,2008, 29, 48.

18 Casarett and Doull’s Toxicology: The Basic Science of Poisons, ed.C. D. Klaassen, McGraw Hill, New York, 5th edn, 1996, pp. 190,691–693.

19 T. Brock and W. Stopford, J. Environ. Monit., 2005, 5, 71N.20 K. M. Ellickson, R. J. Meeker, M. A. Gallo, B. T. Buckley and

P. J. Lioy, Arch. Environ. Contam. Toxicol., 2001, 40, 128.21 D. G. Ellingsen, L. Dubeikovskaya, K. Dahl, M. Chashchin,

V. Chashchin, E. Zibarev and Y. Thomassen, J. Environ. Monit.,2006, 8, 1078.

22 D. G. Ellingsen, S. M. Hetland and Y. Thomassen, J. Environ.Monit.,2003, 5, 84.

23 H. Roels, G. Meiers, M. Delos, I. Ortega, R. Lauwerys, J. P. Buchetand D. Lison, Arch. Toxicol., 1997, 71, 223.

24 M. D. Taylor, J. R. Roberts, S. S. Leonard, X. Shi andJ. M. Antonini, Toxicol. Sci., 2003, 75, 181.

25 J. D. McNeilly, L. A Jimenez, M. F. Clay, W. MacNee, A. Howe,M. R. Heal, I. J. Beverland and K. Donaldson, Toxicol. Lett., 2005,158, 152.

26 O. Moss, Health Phys., 1979, 36, 1209.27 G. E. Hatch, in Comparative biology of the normal lung, ed. R. A.

Parent, CRC Press, Boca Raton, FL, 1992, pp. 617–632.28 O. T. Butler and A. M. Howe, J. Environ. Monit., 1999, 1, 23.29 R. D. Foster, X-Ray Spectrom., 2000, 29, 467.30 M. Semmler-Behnke, S. Takenaka, S. Fertsch, A. Wenk, P. Mayer,

G. Oberdorster and W. G. Kreyling, Environ. Health Perspect.,2007, 115, 728.

31 Y. Thomassen, E. Nieboer, D. Ellingsen, S. Hetland, T. Norseth,J. O. Odland, N. Romanova, S. Chernova and V. P. Tchachtine,J. Environ. Monit., 1999, 1, 15.

32 D. Naranjit, Y. Thomassen and J. C. Van Loon, Anal. Chim. Acta,1979, 110, 307.

J. Environ. Monit., 2008, 10, 1448–1453 | 1453