AM. omq> Hyr., Vol 41, No. 3, pp 269-279, 1997
Bnlah Occupational Hygiene Society
O 1997 AEA Technology pfc
PuWnhed by Ebeyier Science Ltd
Printed in Great Britain
0003-^878/97 S17.00 + 0.00
PII: S0003-^878(97)00002-l
LEACHING STUDIES OF NEUTRON-IRRADIATED
CHYRSOTILE ASBESTOS,
A. M o r g a n and R. J. Talbot
Biomedical Research.JAEA Technology! 551 Harwell,(BidcotloX 11 ORA.ju.lT}
(Received 2$ March 1996)
Abstract—Samples of chrysotile from Quebec (UICC B and Jeffrey 4T-30) and from the Coalinga
region of California (Calidna RG-144) were irradiated with thermal neutrons in a reactor. The
main activation products induced were 46 Sc, 51 Cr, 59 Fe and " t o . Accurately weighed samples of
the irradiated materials were dispersed in N HC1 by hand shaking for 10 s. After leaching for
predetermined periods at 25°C, the samples were filtered and the concentrations of Mg determined
in the filtrates by inductively-coupled plasma atomic emission spectrometry (IPC-AES) and the
activities of the four radionuclides by high resolution y-ray spectrometry. Similar measurements
were made on solutions obtained by refluxing samples of irradiated chrysotile with 2N H O for 2 h.
The specific activities of each of the four activation products were calculated in arbitrary units and,
as the concentrations of Sc, Cr, Fe and Co in the UICC B sample had already been determined, it
was possible to estimate the concentrations of these elements in the other two samples. Similarities
in the leaching patterns of magnesium and of the activation products showed that with all samples,
high proportions of the parent trace elements were present in the form of isomorphous substitutes
for magnesium in structural brucite. Agreement was closest with the Calidna chrysotile in which all
the radionuclides had a similar pattern. With the Jeffrey and UICC B samples, the presence of a
high proportion of the iron as relatively insoluble magnetite accounted for the observed
discrepancy in behaviour between w F e and Mg. More detailed calculation of leaching rates over
specific time intervals showed that, initially, 3l Cr and '"Co dissolved more rapidly than Mg but
that this was followed by a period in which the opposite was the case. It was concluded that the
Calidria RG-144 sample is an ideal candidate for studies of magnesium dissolution in v/vol © 1997
AEA Technology pic. Published by Elsevier Science Ltd
"~~^
INTRODUCTION
In a previous study (Morgan, 1997), the acid-leaching characteristics of a number of
chrysotile samples from various Canadian mines and from the Coalinga region of
California were compared. This showed that it was difficult to compare the intrinsic
leaching rates of magnesium from samples of chrysotile. With milled samples from
Quebec, breakage and surface damage to fibrils caused by milling resulted in a rapid
initial leaching rate of the structural magnesium compared with hand-picked
unopened samples from the same mine. The initial Mg leaching rates of Coalinga
chrysotiles were much lower, due to the fact the fibres are opened more readily and,
in consequence, have suffered less surface damage. With milled samples from
Quebec, about 20% of the total Mg dissolved during the rapid leaching phase
compared with only about 5% at the most for the Coalinga materials. Following
this rapid dissolution phase, leaching rates were rather similar for all opened
chrysotile samples, falling within the range of 1-2% h~'.
It has been shown (Morgan et al., 1973) that acid-leaching studies of neutronirradiated chrysotile give useful information on the chemical form of some of the
associated trace elements. The y-ray spectra of neutron-irradiated chrysotile samples
269
270
A. Morgan and R. J. Talbot
Table 1. Principal y-emitting activation products in neutron-irradiated chrysotile with
half-lives > 10 days
Activation
product
4
*Sc
31
Production process
45
Half-life
Sc(n,y)46Sc
84 days
Cr
Mn
58
Co
59
Fe
"Mn(n,2n)MMn
5
'Ni(n,p)5'Co
"Cifr.tf'Cr
27.8 days
312 days
71 days
45 days
""Co
"CoCn.^Co
5.27 years
M
E,(MeV)
(> 10% abundance)
0.89 (100%)
1.12(100%)
0.32 (8%)
0.84 (100%)
0.81 (99%)
1.10(56%)
1.29(44%)
1.17 (100%)
1.33(100%)
exhibit the same range of radionuclides, irrespective of the source. The principal,
relatively long-lived, activation products are ^Sc, 5lCr, 59Fe and ^Co, which are
induced in (n,y) reactions on the corresponding stable elements. Scandium,
chromium, iron and cobalt are invariably present as impurities either in, or
associated with, chrysotile asbestos. Smaller activities of MMn and 58Co are also
generally present. Information on the physical properties of these radionuclides is
given in Table 1.
With some samples of chrysotile, the leaching characteristics of induced
radionuclides follow closely that of magnesium, indicating that the parent elements
are present as isomorphous substitutes. With other samples, some radionuclides
dissolve either more or less rapidly than the magnesium, indicating that at least a
proportion of the parent element is present in a different form. For example, in
samples that contain magnetite, the fraction of radioactive 59Fe that is insoluble
corresponds to the fraction of iron that can be separated magnetically as magnetite.
This does not readily dissolve in dilute mineral acids (Morgan et ai, 1973).
With samples of chrysotile in which trace elements are present uniquely as
isomorphous substitutes for magnesium in structural brucite, it is possible to
determine the dissolution rate of the structural magnesium in vivo, following the
administration of neutron-irradiated material to experimental animals (Morgan et
al., 1971). This is done by measuring the cumulative excretion of either 5lCr or "^Go
as a fraction of the corresponding administered activity. Once these elements
dissolve from the fibre in vivo, they are excreted rapidly, mainly in urine, and
therefore can be used as surrogates for magnesium to estimate its dissolution rate in
vivo.
The purpose of the present study was to compare the acid leaching
characteristics of neutron-irradiated chrysotile samples from Quebec and from the
Coalinga region of California to determine whether they were appropriate for use in
studies of magnesium dissolution in vivo.
MATERIALS
Two of the samples used were Canadian in origin. One was the Standard
Reference Sample B, prepared under the auspices of the Union Internationale
Contre le Cancer (UICQ by bulking samples from eight mines—seven in Quebec
Neutron-irradiated chyrsotile asbestos
271
Table 2. Concentrations of trace elements in chrysotile samples determined by
neutron activation analysis (Holmes el at., 1971)
Chrysotile
Fe (%)
Cr (ppm)
Co (ppm)
Sc (ppm)
UICC B
Jeffrey (Mine H)
2.6
3.2
490
380
50
53
5
6
and one in British Columbia, roughly in proportion to their annual output.
Approximately half the material originated from the Jeffrey mine operated by
Johns-Manville. The UICC sample has been shown to be homogeneous at the 10-mg
level (Morgan and Timbrell, 1971). A second sample from Quebec, designated 4T30, was also from the Jeffrey mine and, despite having been milled, the fibres are
relatively long compared with the UICC material. The third sample, designated RG144, was from the Calidria mine in the Coalinga region of California and is a highpurity product in which the fibres are short and well-dispersed.
The concentrations of iron, chromium, cobalt and scandium in the UICC B
sample, and also in a sample of chrysotile from the Jeffrey mine, which formed a
component (Mine H) of the standard reference sample, have been determined by
neutron activation analysis (Holmes et al., 1971). The values obtained are given in
Table 2.
METHODS
Neutron irradiation of chrysotile
Two samples (about 200 mg) of each type of chrysotile were weighed into small
polytainers. All six samples were packed into a standard polythene irradiation tube
and irradiated in the University of London reactor at Silwood Park. The samples
were irradiated for a total of 120 h in a thermal neutron flux of about 2xlO12
neutrons c m " 2 s " ' .
Magnesium and radionuclide leaching studies
The magnesium leaching rates for the three samples were determined using
similar methods to those described previously (Morgan, 1997), except that
somewhat smaller weights (10-15 mg) of chrysotile were used for each
measurement. Accurately weighed samples were suspended in 25 ml of N HC1 for
predetermined times at 25°C and then filtered through Whatman No. 1filterpapers.
The magnesium contents of the filtrates were determined by inductively-coupled
plasma atomic emission spectrometry (ICP-AES). Additional 5-ml aliquots of the
filtrates were measured into glass liquid scintillation vials for radiometric
determination of the dissolved activation products by high-resolution y-ray
spectrometry.
The total magnesium and radionuclide activities of the samples were determined
by refluxing accurately weighed amounts of irradiated material in 25 ml of 2N HC1
for 2 h. The resulting suspension was diluted to 50 ml, and 5-ml aliquots of both the
filtered and unfiltered suspensions were analysed for Mg by ICP-AES and y-counted
272
A. Morgan and R. J. Talbot
to ascertain whether any of either the Mg, or radioactivity, was associated with the
siliceous or other insoluble material.
Measurement of radioactivity
Measurements of radioactivity were made using a high-purity germanium crystal
(46 mm diameter x 59 mm thick with a 31-mm diameter x46-mm well) housed in a
lead shield 100 mm thick. A multi-channel analyser was used with a channel width
of 0.65 kev. All samples were counted for 1000 s. The counting rates in appropriate
channels were corrected for background and radioactive decay. The resulting
radioactivities were corrected to the total volume of the samples and divided by the
sample weight to give a relative specific activity for each radionuclide in that sample.
The specific activities were then expressed as a percentage of the corresponding
specific activity in the unfiltered refluxed sample.
RESULTS
Magnesium leaching results
The cumulative magnesium leaching patterns are shown in Figs 1-3 for the
UICC B, Jeffrey 4T-30 and RG-144 samples, respectively. Because of the limited
quantities of irradiated material available, most of the points in these figures
represent single measurements only. In the few cases in which duplicate
determinations were made, the agreement was very good. The amounts of
magnesium leached are expressed as a percentage of the mean total magnesium
UICCB
100
—0
0
80 -
A
—u
HI
60 -
o
S3
0
D
o
40 -
'•6
a
20
O «Sc
51
D
Cr
¥
r
Mg
D
59
A
"CO
Fe
i
i
i
i
i
i
I
20
40
60
80
100
120
140
160
Leach time (hours)
Fig. 1. Cumulative leaching patterns of radionuclides and of stable magnesium from the UICC B
Standard Reference chrysotile.
Neutron-irradiated chyrsotile asbestos
273
JEFFREY 4T-30
100
a
z.
A
80 -
-a
91
>
c
o
n
20
o Mg
)
40 -i
o
•
V
f
4S
Sc
D
A
«°Co
1
1
i
i
i
i
I
20
40
60
80
100
120
140
160
Leach time (hours)
Fig. 2. Cumulative leaching patterns of radionuclides and of stable magnesium from Jeffrey 4T-30
chrysotile.
CAUDRIA RG-144
100
80 -
<U
>
60 -
o Mg
O <«Sc
O
40 -
t;
20
IF
f
20
40
60
80
100
•
51
n
59
A
"CO
Cr
Fe
i
i
120
140
160
Leach time (hours)
Fig. 3. Cumulative leaching patterns of radionuclides and of stable magnesium from Calidria RG-144
chrysotile.
274
A. Morgan and R. J. Talbot
Table 3. Relative activities of filtered and unfiltered reflux samples
Sample
UICCB
Jeffrey 4T-30
RG-144
5l
Cr
0.825
0.926
0.986
M
Co
1.029
0.903
0.904
46
Sc
0.874
0.892
0.925
w
Fe
0.970
1.025
1.001
contents reported in the previous study (Morgan, 1997). The total magnesium
contents measured in the present study, obtained by refluxing the samples for 2 h in
2N HCl are slightly lower (95-98%) than the mean values obtained in the previous
study. However, the leaching curves are very similar to those reported previously for
the same samples, indicating that magnesium solubility had not been significantly
affected by neutron irradiation.
Radionuclide leaching results
The ratios of the filtered to unfiltered relative counting rates for the four main
radionuclides after refluxing are given in Table 3. With the exception of 59 Fe, most
of the values are less than unity, indicating that small fractions of the other
radionuclides remained undissolved after refluxing. For this reason, the fraction of
radionuclides leached with N HCl at 25°C is expressed as a percentage of the
unfiltered, rather than the filtered, specific activity as this is likely to be a more
accurate estimate of the total radioactivity.
The cumulative leaching results for the four main activation products are
compared with the corresponding values for magnesium in Figs 1-3 for the UICC B,
Jeffrey 4T-30 and RG-144 samples, respectively. From these, it can be seen that,
initially, most radionuclides dissolve somewhat more rapidly than the magnesium,
but at later times, the opposite is the case.
DISCUSSION
The concentrations of iron, chromium, cobalt and scandium in the UICC B
standard reference sample have been measured by Holmes et al. (1971) using
neutron activation analysis (NAA). Using these values, it is possible to estimate the
concentrations of the same elements in the other two samples used from their
relative specific activities. The concentrations derived in this way are given in
Table 4. The trace element concentrations in the Jeffrey 4T-30 sample are similar to
those measured previously by neutron activation in a sample from the same mine
(see Table 2). The iron content of the Calidria RG-144 chrysotile is in exact
Table 4. Estimated concentrations of trace elements in Jeffrey 4T-3O and RG144 chrysotile
Chrysotile
Fe (%)
Cr (ppm)
Co (ppm)
Sc (ppm)
Jeffrey 4T-3O
RG-144
2.8
1.6
455
1690
45
65
6
6
Neutron-irradiated chyrsotile asbestos
275
agreement with that measured previously by ICP-AES (Morgan, 1997) and less than
is generally found in samples from Quebec. The chromium content, however, is
about three times greater. The trace metal profile of the Calidria chrysotile is very
similar to that of chrysotile from the Cassiar mine in British Columbia reported by
Morgan et al. (1973), which also has a relatively high chromium content.
The similarity in the magnesium and radionuclide leaching patterns shown in
Figs 1-3 indicates that, in most cases, high proportions of the parent trace elements
are present as isomorphous substitutes for structural magnesium. An exception is
59Fe in the UICC B and Jeffrey samples in which only about 50% had been leached
when virtually all the magnesium had dissolved. Similar behaviour had been noted
in an previous investigation (Morgan et al., 1973), in which it was pointed out that
the undissolved fraction of 59Fe is almost identical to that which could be separated
magnetically as magnetite. With the Calidria RG-144 chrysotile, the leaching
pattern of 59Fe is much closer to that of magnesium, which can be accounted for by
the virtual absence of magnetite. For chromium, cobalt and scandium, more of the
corresponding radionuclide than magnesium is dissolved during the first few hours
of leaching but, after about 4 h, the situation is generally reversed. This
phenomenon was also reported in the earlier study (Morgan et al., 1973) with
samples from Quebec.
With the Canadian samples, the leaching pattern of ^Sc appears to follow most
closely that of magnesium. Unfortunately, much of the scandium that dissolves in
vivo is retained in various organs, including Gl-tract, liver, spleen and kidneys,
which limits its usefulness for studies of magnesium dissolution. It is also
noteworthy that the chromium content of the Jeffrey sample appears to be
considerably more soluble than that of the UICC B material, of which Jeffrey
chrysotile is only one component. With the Calidria chrysotile, the radionuclide that
mimics the behaviour of magnesium most closely is 51Cr.
In Figs 4-6, the dissolution rates for magnesium and for 51Cr and ^Co in the
different leaching periods are plotted against the mid-time of the leaching period for
the RG-144, UICC B and Jeffrey 4T-30 samples, respectively. The initial leaching
rates for magnesium are virtually identical to those reported earlier (Morgan, 1997)
and, after 4 h, decline in an approximately exponential manner. For all the samples,
the leaching rates of 51Cr and ^Co are greater than that of magnesium during the
first 2 min of contact with acid. Between 2 min and 1 h, the leaching rates of the
magnesium and radionuclides are similar, but between 1 h and 20 h, the magnesium
dissolves more rapidly than the activation products. During this period, the leaching
rate of 5l Cr is generally slightly higher than that of ^Co and, at least with the single
mine samples, the leaching rates of the 5lCr and ^Co in the 4- to 16-h leach period
are greater than in the 1- to 4-h period, resulting in a pronounced 'kink'. At later
times, within the limits of experimental error, the leaching rates are similar for all
components. In most cases, the points in the Figs 4-6 represent individual
measurements, so that no indication of errors can be given.
It should be noted that the magnesium in chrysotile asbestos is not all present as
structural brucite. In some samples, the space between fibrils is filled with an
amorphous hydrated magnesium silicate matrix material, possibly alpha sepiolite, as
proposed by Bates and Comer (1959). Chrysotile samples from Quebec also
generally contain free brucite. In most samples, the concentration of free brucite is
A. Morgan and R. J. Talbot
276
UICCB
1000
100 -
o
_2
o
K
b
0.01
20
40
60
80
100
120
140
Mid-time of leach period (hours)
Fig. 4 Dissolution rates of radioactive chromium and cobalt from the UICC B Standard Reference
chrysotile compared to that of stable magnesium.
less than 5%, but it can be as high as 20% (Wagner et al., 1970). Samples of
chrysotile from the Calidria mine are reputed to contain little non-structural brucite
in either free or interstitial form. Differences in the initial dissolution rates of
magnesium and the activation products could be accounted for if, for example, the
concentrations of the parent trace elements are greater in free and interstitial than in
structural brucite. Indeed, levels of impurities in the interstitial matrix material are
probably greater than in the fibres themselves, since analysis of fibre veins and
surrounding rock samples usually shows that the former have lower impurity
concentrations (Reimschussel, 1975). The reason for this is that thefibrescrystallize
out first in a relatively pure form, leaving a higher concentration of impurities in the
277
Neutron-irradiated chyrsotile asbestos
JEFFREY 4T-30
1000
0.01
40
60
80
100
120
140
Mid-time of leach period (hours)
Fig. 5. Dissolution rates of radioactive chromium and cobalt from Jeffrey 4T-3O chrysotile compared to
that of stable magnesium.
matrix, which crystallizes last. As magnesium in free brucite and matrix materials
will dissolve more rapidly than magnesium in structural brucite, such differences
could account for the observed discrepancies in relative solubility between
magnesium, on the one hand, and chromium and cobalt on the other.
Finally, it has to be considered whether the use of induced radioactivity would
enable the dissolution rate of structural magnesium to be determined in vivo for any
of the samples examined. In an earlier study with rats (Morgan et al., 1971), a
radioactive tracer technique was used to study magnesium dissolution from
chrysotile in vivo following intrapleural injection of samples from Zimbabwe
(UICC A), British Columbia (Cassiar) and Quebec. A high proportion of dissolved
A. Morgan and R. J. Talbot
278
CALIDRIA RG-144
1000
0.01
40
60
80
100
120
140
Mid-time of leach period (hours)
Fig. 6. Dissolution rates of radioactive chromium and cobalt from Calidna RG-144 chrysotile compared
with that of stable magnesium.
^ C o was excreted, principally in urine, following its removal from the fibre. Some of
the dissolved 51 Cr was retained in the liver so that a slightly smaller fraction was
excreted. When this systemic 5l Cr was taken into account, the dissolution patterns
of these two activation products in vivo were very similar to those measured in vitro
for all three samples, although, of course, the rates of dissolution were much less.
Data from the Cassiar sample indicated that about 30% of the magnesium had
dissolved by 2 months after administration but that the subsequent dissolution rate
was much slower.
The present study shows that Calidria RG-144 chrysotile would be an ideal
candidate for studies of magnesium dissolution in vivo because of the low levels of
Neutron-irradiated chyrsotile asbestos
279
non-fibrous impurities and the fact that the cumulative dissolution patterns of 51Cr
and ^Co are similar in overall shape to that of the magnesium. Although there are
small differences in the initial dissolution rates of these activation products and
magnesium, over longer periods, these differences become less important. It is
possible that the radioactive tracer techniques also could be applied to measure
magnesium dissolution rates from less pure materials, such as those from Quebec,
but further studies would be required to confirm this. Clearly, as a prerequisite to
any in vivo studies, it is important to obtain as accurate in vitro leaching information
as possible for both the Mg and the relevant activation products.
Finally, trial intratracheal instillation studies with the irradiated chrysotile used
in the present study indicated that higher specific activities would be necessary for
investigations of magnesium dissolution in vivo. Higher specific activities would
enable small amounts of chrysotile to be administered by intratracheal instillation
within the limits recommended by Sykes et al. (1983) and would also enable
investigations to be pursued over longer periods. Reactors are available with
neutron fluxes approximately 100 times greater than in the University of London
reactor, which could be used for the production of chrysotile with a higher specific
activity than that used in the present study.
Acknowledgement—This investigation was supported by Union Carbide.
REFERENCES
Bates, T. F. and Comer, J. J. (1959) Further observations on the morphology of chrysotile halloysite. In
Proc. 6th National Conference on Clays and Clay Minerals, NAS-NRC, pp. 237-248.
Holmes, A., Morgan, A. and Sandalls, F. J. (1971) Determination of iron, chromium, cobalt, nickel and
scandium in asbestos by neutron activation analysis. Amer. ind. Hyg. Assoc. J. 32, 281-286.
Morgan, A. (1997) Acid leaching studies of chrysotile asbestos from mines in the Coahnga region of
California and from Quebec and British Columbia. Ann. occup. Hyg. 41, 249-268.
Morgan, A., Holmes, A. and Gold, C. (1971) Studies of the solubility of constituents of chrysotile
asbestos in vivo using radioactive tracer techniques. Environ. Res. 4, 558—570.
Morgan, A., Lally, A. E. and Holmes, A. (1973) Some observations on the distribution of trace metals in
chrysotile asbestos; Ann. occup. Hyg. 16, 231-240.
Morgan, A. and Timbrell, V. (1971) The use of neutron activation analysis to determine the composition
of blended samples of asbestos. Int. J. appl. Radial. Isotopes 22, 745-751.
Reimschussel, G. (1975) The association of trace metals with chrysotile asbestos. In Physics and Chemistry
of Asbestos Minerals. Third International Conference, Quebec.
Sykes, S. E., Morgan, A., Moores, S. R., Davison, W., Beck, J. and Holmes, A. (1983) The advantages
and limitations of an in vivo test system for investigating the cytotoxicity and fibrogenicity of fibrous
dust*. Environ. Hlth Perspect. 51, 267-273.
Wagner, J. C , Berry, G. and Timbrell, V. (1970) Mesothelioma in rats following the intrapleural
inoculation of asbestes. In Pneumoconiosis. Proceedings of an International Conference, Johannesburg
1969, ed. H. A. Shapiro, pp. 216-219. Oxford University Press, Cape Town.