Study of the cytotoxicity of CCVD carbon nanotubes
ABSTRACT The cytotoxicity of different samples of carbon nanotubes synthesised by catalytic chemical vapour deposition was investigated towards human umbilical vein endothelial cells, using two cytotoxicity standard tests (neutral red assay for the cell viability and MTT assay-tetrazolinium salt-for the cell metabolic activity). No toxicity was found for any sample, although a slight dilution effect may exist for two of them. © 2006 Springer Science + Business Media, Inc.
Article: Solid C60: a new form of carbon[show abstract] [hide abstract]
ABSTRACT: A new form of pure, solid carbon has been synthesized consisting of a somewhat disordered hexagonal close packing of soccer-ball-shaped C60 molecules. Infrared spectra and X-ray diffraction studies of the molecular packing confirm that the molecules have the anticipated 'fullerene' structure. Mass spectroscopy shows that the C70 molecule is present at levels of a few per cent. The solid-state and molecular properties of C60 and its possible role in interstellar space can now be studied in detail.09/1990; 347(6291):354-358.
Investigation of the cytotoxicity of CCVD carbon nanotubes
towards human umbilical vein endothelial cells
E. Flahaut a , M.C. Durrieu b, M. Remy-Zolghadri a, R. Bareille b and Ch.
a CIRIMAT UMR-CNRS 5085/LCMIE, Centre Interuniversitaire de Recherche et d’Ingénierie
des Matériaux, Université Paul-Sabatier, 31062 Toulouse Cedex 4, France
b INSERM U 577, Université Victor Segalen, 146, Rue Léo Saignat, 33076 Bordeaux Cedex,
The cytotoxicity of different samples of carbon nanotubes synthesised by catalytic chemical
vapour deposition was investigated towards human umbilical vein endothelial cells, using two
cytotoxicity standard assays (neutral red assay for the cell viability and MTT assay—
tetrazolinium salt—for the cell metabolic activity). No cytotoxicity was found for any sample.
Keywords: Carbon nanotubes; Chemical vapor deposition; Biocompatibility
2.1. Synthesis of CNTs
2.2. Evaluation of the CNTs’ cytotoxicity
3. Results and discussion
In the heart of the dispute about the possible negative effects of nanotechnologies and the
potential toxicity of nanoscale objects, the question of the cytotoxicity of carbon nanotubes
(CNTs) is very relevant. Before CNTs can be used in medical applications such as drug-delivery
agents for example, their innocuity towards the human body has to be assessed. Compared to the
abundant literature dealing with the synthesis, characterisation and potential applications of
CNTs, it is striking to note the lack of data on their possible bio-toxicity. CNTs are unique
tubular 1D-objetcs which can be described as a rolled hexagonal carbon network, ended by half-
fullerene caps. They can be single-walled (SWNTs) or multi-walled (MWNTs) in the case of
concentric tubes arranged in a Russian dolls-like structure. It is important to keep in mind that
CNTs can be produced by different methods, the main three ones being the electric arc-discharge
, Laser ablation  and catalytic chemical vapour deposition (CCVD) ,  and . Each
method gives CNTs of different kind and all require the use of metals as catalysts. As a general
trend, arc-produced CNTs tend to be about 1 μm long and to gather into large bundles up to more
than 100 CNTs. These CNTs are generally very straight and, said to be free of structural defects.
They are produced together with carbon soot containing amongst others fullerenes and graphitic
particles. Laser ablation produces purer samples (containing fewer by-products) and the CNTs
form very long and large bundles, up to hundreds of μm long. CCVD can be used to produce a
wide variety of different CNTs ranging from SWNTs to MWNTs (possibly containing structural
defects). When the number of walls is low (less than 5), CCVD CNTs are generally forming thin
bundles, with a length generally ranging between a few micrometers to hundreds of micrometers.
The diameter distribution is ranging between less than 1 nm and 4–5 nm, as opposed to the arc-
discharge and laser ablation methods where the diameter distribution is narrower and mainly
centred around 1.3 nm. Each synthesis method produces a different quality of CNTs, the term of
quality involving the chemical purity (residual catalyst—mainly transition metals—and presence
or not of other forms of carbon species such as amorphous carbon and graphitic-like particles)
and the structural quality of the CNTs. The control of the number of walls of the CNTs and the
diameter distribution can also be seen as a sort of purity (in terms of selectivity).
The first study (published in 2001) by Huczko et al.  was focusing on the comparison between
arc-discharge produced CNTs and asbestos fibres. CNTs were introduced in guinea pigs by
intratracheal instillation (suspension in sterile saline media with addition of surfactant). The
results of the bronchoalveolar lavage examinations 4 weeks after the instillation revealed no
change in the pulmonary function and did not induce any measurable inflammation in
bronchoalveolar space, concluding to the innocuity of the tested samples.
Shvedova et al.  have investigated the assessment of raw-CNTs cytotoxicity using human
keratinocyte cells, showing that the dermal exposure to raw CNTs may lead to dermal toxicity
due to accelerated oxidative stress in the skin, mainly because of important amounts of iron in the
studied samples (HiPCO CNTs, CCVD process). Another study by the same group  has
compared the aerosols formed during the handling of different CNTs samples (HiPCO, laser
ablation) and has shown that the airborne concentrations were generally lower than 53 μg m−3. It
is only recently in 2004 that two reports on the toxicological effects of CNTs have been issued,
both based on a comparative toxicological approach and intratracheal instillation route of
exposure to examine the pulmonary toxicity of SWNTs. Lam et al.  focused on the
histopathological alterations in mice 7 and 90 days after the exposure to CNTs, produced by two
different CCVD methods and containing various amounts of residual catalytic metals (Fe, Co, Ni,
Mo, etc.). They reported that CNTs could be more dangerous than quartz particles and that all the
CNTs tested induced lung lesions characterised by the presence of granulomae. They also
suggested that the metal contaminants could not be the only cause. Warheit et al.  studied the
ability of CNTs to induce pulmonary inflammation as well as to alter lung cellular proliferation.
They confirmed the formation of granulomae after instillation of SWNTs (laser-ablation
produced soot containing about 60% SWNTs) and revealed that the unique mechanical and
chemical properties of CNTs make them more persistent in biological systems. Pantarotto et al.
 reported the translocation of water soluble SWNT derivatives across cell membranes and
have shown that cell death can be induced by functionalised CNTs (bioactive peptides),
depending on their concentration in the media. Very recent results of in vitro tests performed on
human embryo kidney cells  suggest that SWNTs can induce cell apoptosis but also that these
cells are able to isolate CNTs-attached cells from the unaffected ones by secretion of “isolation”
proteins. it has also been shown that CNTs can generate an inflammatory response of the human
immune system, both via the classical and alternative pathways .
We report here the study of the cytotoxicity of different samples of carbon nanotubes synthesised
by CCVD. We took advantages of two existing standard assays, Neutral Red assay for the cell
viability and MTT assay (tetrazolinium salt) for the cell metabolic activity. These assays are
optimized for indirect cytotoxicity and we adapted them to test the effect of the CNTs on cells.
Human umbilical vein endothelial cells (HUVEC) are especially good candidates for these tests,
because CNTs are often thought to be used for drug-delivery  and would be injected in the
human body to be carried to targeted places by the blood.
2.1. Synthesis of CNTs
The CNTs were prepared by catalytic chemical vapour deposition using bimetallic MgO-based
catalysts  and . Mg1−x(CoyM1−y)xO catalysts were prepared by combustion synthesis .
The different compositions were as follows: x = 0.01 with y = 0.75 and M = Mo (sample A) or W
(sample B); x = 0.05 with y = 0.9 and M = Mo (sample C). Magnesium and cobalt nitrates were
dissolved in deionised water and the required amount of (NH4)6Mo7O24 · 4H2O or
Na2WO4 · 2H2O was added to the solution. Citric acid was used as the fuel, using the
stoichiometric proportion . For each experiment, 1.5 g of the starting oxide was reduced in a
H2–CH4 mixture (18 mol.% CH4, heating and cooling rates 5 °C/min, maximum temperature
1000 °C, no dwell). This resulted in a mat of composite powder (which density depended on the
composition of the catalyst), which was then treated with a concentrated aqueous HCl solution
(12 M) to separate the CNTs by dissolving all the remaining oxide material, as well as
unprotected metal particles  and . The suspensions were filtered and washed on cellulose
nitrate membranes (Whatman, 0.45 μm) with deionised water until neutrality. The samples were
dried overnight in air in an oven at 80 °C, yielding different amounts of CNTs according to the
composition of the corresponding catalyst. The samples were characterised by electron
microscopy (scanning electron microscopy and transmission electron microscopy using a JEOL
2010 operated at 120 kV); the specific surface area (noted Se, m2 g−1) was measured by the BET
method using N2 adsorption at liquid N2 temperature (Micromeritics FlowSorb II 2300) and the
carbon content (noted Ce, wt. %) was obtained by elemental analysis (flash combustion).
2.2. Evaluation of the CNTs’ cytotoxicity
Primary cultures of human umbilical vein endothelial cells (HUVEC) were prepared and cultured
as previously described . HUVEC were submitted to the samples by the way of fragments of
sterile (autoclaved: 121 °C–20 min.) material. HUVEC were seeded at a density of 6000 cells per
cm2 in 96-well microtiter plates (Nunc, Denmark) and the culture was maintained at 37 °C after
cell plating until confluency was reached. At confluency the medium was replaced by the
material samples. To obtain such samples, fragments of sterile material were immersed in culture
medium (IMDM) in order to obtain 5 cm2 of the sample surface (estimated from the measure of
the specific surface area of each CNTs sample) to 1 mL of IMDM (Iscove Modified Dulbecco’s
Medium) according to the standards of indirect cytotoxicity . Samples were treated in
borosilicate glass tubes at 37 °C for 120 h and 5% CO2 without stirring according to the standard
procedures. Borosilicate tubes containing IMDM with either no material or a solution of phenol
at a concentration of 64 g L−1 (known to be cytotoxic) were processed under the same conditions
to provide negative and positive controls, respectively. After this step, all samples were
centrifuged at 1200 rpm (180g) and the supernatant was recovered, according to the standard
procedures. In the case of CNTs, centrifugation at such a low acceleration for very dilute
suspensions is not able to withdraw the CNTs from the suspensions. Suspensions were
supplemented to 10% FCS (Fetal Calf serum) and final dilutions were achieved using IMDM
containing 10% FCS to obtain the 50% (v/v), 10% (v/v) and 1% (v/v) samples. When HUVEC
confluency was reached, the culture medium was removed and replaced by the samples (100μL)
at various dilutions [100% (v/v), 50% (v/v), 10% (v/v), 1% (v/v)] in the culture medium
(0.33 cm2 wells containing approximately 30,000 endothelial cells at the time of the assay) and
supplemented with 10% FCS for 24 h at 37 °C and 5% CO2. At the end of the incubation period,
the supernatants (material or control samples) were discarded, and two different assays , ,
 and  were performed: cell viability (Neutral Red assay) and cell metabolic activity (MTT
assay). The intensity of the colours obtained (red and blue respectively) is directly proportional to
the viability and metabolic activity of cell populations and inversely proportional to the toxicity
of the material. For the statistical analyses, values are expressed as mean ± SD. The statistical
significance was tested by an unpaired Student t test and p < 0.05 was considered significant.
3. Results and discussion
Using the same CCVD conditions, three different kinds of CNTs were prepared from three
different catalysts because we wanted to investigate both the possible influence of the nature of
the catalytic metal (Co combined to either Mo (samples A and C) or W (sample B)) and of the
amount of catalytic metal (catalyst used to synthesize sample C contains a higher content of
catalytic metal than catalyst used to prepare sample A). Fig. 1 shows typical electron microscope
images (SEM, TEM) of the CNTs samples. SEM images were obtained on the raw CCVD
samples, i.e. before the elimination of the oxide catalyst which can be seen in the background of
SEM images: 1(a, c, d). These images reveal a high density of CNTs in all three samples. TEM
images (Fig. 1b and inset of Fig. 1d) show that the filaments observed by SEM are in fact
individual or small-diameter bundles of CNTs. Careful analysis of high-magnification TEM
images allowed the measure of the number of walls and diameter of the CNTs (Fig. 2). Sample A
(Fig. 2a, 96 individual CNTs imaged) contains CNTs having between 1 and 3 walls, with ca. 80%
double-walled CNTs (DWNTs) and outer diameters ranging from 1.1 to 3.2 nm (Fig. 2d). Sample
B (Fig. 2b, 164 individual CNTs imaged) contains CNTs having between 1 and 4 walls, with
37% DWNTs, 48% triple-walled CNTs (TWNTs) and 12% CNTs with four walls. The outer
diameters are ranging from 1.1 to 4.3 nm (Fig. 2e). Sample C (Fig. 2c, 183 individual CNTs
imaged) contains CNTs having between 1 and 6 walls, with 9% single-walled CNTs (SWNTs),
72% DWNTs and 15% TWNTs. The outer diameters are ranging from 0.7 to 6.3 nm (Fig. 2f).
Fig. 1. Typical images of the CNTs samples. SEM images were obtained on the raw CCVD
samples, before the elimination of the oxide catalyst (which can be seen in the background).
Sample A observed by SEM (a) and TEM (b); note the small diameter of the bundles of CNTs.
Typical SEM images of samples B (c) and C (d—the inset is a TEM image showing a cross-
section of some of the few triple-walled CNTs found in this sample).
Fig. 2. Distribution of the numbers of walls (a–c) and the inner (di) and outer (do) diameter (d–f)
for each sample (whole population of CNTs). Sample A (a) and (d), respectively, sample B (b)
and (e) respectively and sample C (c) and (f), respectively.
The main characteristics of the CNTs samples are summarized in Table 1: they have a specific
surface area ranging from 550 m2 g−1 (sample C) to more than 1100 m2 g−1 (sample A). The
carbon content is higher than 90 wt.% (corresponding to more than 97.8 at.%) for samples A and
B and close to 80 wt.% (equivalent to more than 95 at.%) for sample C. The diameter distribution
widens from sample A (1.2–3.2 nm) to sample C (0.7–6.3 nm).
Characteristics of the CNTs samples
Reference Catalyst composition (nm)
Se (m2 g−1)
1.2 – 3.2 2.01
1.0 – 4.3 3.11
The catalyst composition is the elemental one, although these catalysts are not solid solutions
0.7 – 6.3 2.37
Data obtained from the biocompatibility studies following incubation of cells with CNTs
suspensions were expressed for Neutral Red and MTT assays as a percentage of the values
obtained from cells incubated with negative control extracts. Phenol (64 g L−1) was used as a
toxic control. Fig. 3 compares the results (given in percent of the response to the assay compared
to an untreated control blank sample) of the cell viability (Neutral Red assay, Fig. 3a) and cell
metabolic activity (MTT assay, Fig. 3b). Any response lower than the dotted line at 75% of the
response of the blank sample may be considered as cytotoxic. The results of the cells viability
assay (Fig. 3a) indicated that none of the samples was found to be cytotoxic, although for
samples A and B the error bars sometimes extend below the cytotoxicity threshold (75%). There
may be a slight dilution effect for samples A and B, with the HUVEC viability slowly decreasing
with the dilution, as confirmed by a significant statistical difference between diluted samples
(results marked with a star in Fig. 3a) and the maximum concentration (100%). The cell
metabolic activity assay (Fig. 3b) confirmed that none of the sample showed a cytotoxic effect
but revealed no obvious dilution effect in this case. HUVEC were chosen in this study because of
their possible future contact with CNTs when intravenous route is envisaged and because such
cells are very sensitive to modifications of their environment. Indeed, they are known to respond
quickly to various stimuli (biochemical, mechanical) and are thus good indicative cells  of a
possible cytotoxic effect.
Fig. 3. HUVEC viability (a) and metabolic activity (b) assays for samples A, B and C. Results
expressed in percentage of the answer of the untreated control sample and compared to phenol
(cytotoxic control reference); a star (*) indicates that the value is statistically different from
The biocompatibility tests performed in this study required the use of very small amounts of
CNTs (5 cm2 of CNTs per mL for the non-dilute suspension). Because the specific surface area of
each CNTs sample was different (and very high), the concentration in weight per volume was
different for each of them. The concentrations for the more concentrated suspensions were thus
close to 0.5 μg mL−1 for sample A, 0.64 μg mL−1 for sample B and 0.9 μg mL−1 for sample C.
Several successive dilutions were required to reach such low concentrations and systematic errors
thus accumulate. Although the samples were prepared very carefully, the validity of these widely
used standard tests may be questioned in the case of very high specific surface area nanomaterials
such as CNTs.
The question of the chemical purity of the samples has already been raised: is the toxicity coming
from the CNTs themselves or from residual catalysts still present in the purified samples? Lam et
al.  suggested that the metal contaminants could not be the only cause of the toxicity, in
agreement with the results of this study because the only sample showing no cytotoxicity at any
concentration is the one which contains the more residual metal nanoparticles. These metal
nanoparticles are in fact encapsulated by graphene-like layers and thus efficiently protected
against their environment ; the metals should thus never be in contact with the biological
material. If the toxicity does not come from carbon-encapsulated residual catalyst, it may rather
come from the structural characteristics of the CNTs, such as their gathering into bundles of
different diameter and their outer diameter distribution. In terms of bundling, one can imagine
that the lower the concentration of the dispersion, the higher the dispersion of the CNTs and thus
the smaller the diameter of the bundles. A lower concentration would favour a higher dispersion
of the CNTs which may improve their interactions with the cells and thus possibly lead to some
toxicity. The question of the outer diameter distribution of the CNTs is important as well. The
response of the three samples to the tests does not show any clear correlation with the mean
diameter of the CNTs populations. In fact, this is probably because the diameter of the CNTs has
no relationship with the size of the HUVEC, which are adherent cells of cobblestone morphology
in such culture conditions, with a mean size usually ranging from 30 to 50 μm on the culture
support. Such small variations of the CNTs outer diameter distribution should thus have no
consequence. About the effect of concentration, Pantarotto et al.  and Maynard et al.  also
worked with different concentrations of CNTs and both claimed a lower toxicity when the
concentration was decreased (in the case of Maynard, it is mainly the contamination with Fe
which is responsible for the toxicity); it is also possible (although statistically relevant for
samples A and B) that in the present study the potential dilution effect could be only fortuitous
and may fall within the experimental error. The results of the MTT assay would agree with this
second explanation. Experiments to confirm this observation are currently in progress.
The inflammatory response of these cells has not been investigated in this study but would be
important to know. However, results obtained on sample A indicate that the DWNTs can activate
the human complement via both the classical and alternative pathways. The C1q protein, which is
involved in the classical pathway, adsorbs readily onto CNTs . We do not have any data
about the possible endocytosis of CNTs by the cells but Pantarotto et al.  have reported that
functionalised CNTs are able to cross the cell membrane and to accumulate in the cytoplasm or
even reach the nucleus without being toxic for the cell at low enough concentration. The exact
concentration (in weight per volume) of the CNTs suspensions was not given but it was reported
that the toxicity appeared with increasing the concentration, which is different from the results
presented here. However, there are two main differences between these studies because (i)
Pantarotto et al.  used very short CNTs (0.3 to 1μm long), and (ii) these CNTs were
functionalised by bioactive peptides. The length of the CNTs used in this work is more likely to
be ranging between 5 μm and up to 100 μm, which makes an important difference. A microscopy
study of the HUVEC after the cytotoxicity test would give important information about the
possible endocytosis of these non-functionalised CNTs.
In summary, the cytocompatibility study of three CCVD-synthesised CNTs samples was
investigated by two standard assays (Neutral Red and MTT assays). It revealed no cytotoxic
effect according to the standards of the method. However, we have observed that for the CNTs
having the highest specific surface area, a cytotoxicity effect (decrease of HUVEC viability)
seems to appear slowly with increasing the dilution of the suspension. This dilution effect may be
related to the amount of CNTs present in the different suspensions (the higher the specific surface
area, the lower the amount of CNTs) and/or to bundling effects. Although these results do not
indicate any cytotoxicity towards HUVEC, and because each CNTs sample is different (presence
of residual metals or additional carbon contamination such as graphitic-like particles and/or
amorphous carbon deposits), the handling of CNTs should always be done in safe conditions by
wearing gloves and appropriate respiratory protection. The results of the study of the cytotoxicity
of a given sample could probably not be extrapolated without risk to another sample prepared
using a different technique or even from a different catalyst for the same synthesis process.
Potential physical damages to the cells and the likely chemical stability of CNTs in the human
body incite to carefulness but the results of this study however suggest that the vectorisation of
substances by CNTs via injection in the blood circulation seems conceivable.
We acknowledge Prof. A. Peigney for SEM imaging and Lucien Datas from the Service Commun
de Microscopie à Transmission of the University Paul Sabatier for his help with HRTEM
 W. Krätschmer, L.D. Lamb, K. Fostiropoulos and D.R. Huffman, Solid C60: a new form of
carbon, Nature 347 (1990), pp. 354–358.
 T. Guo, P. Nikolaev, A. Thess, D.T. Colbert and R.E. Smalley, Catalytic growth of single-
walled nanotubes by laser vaporization, Chem Phys Lett 243 (1995), pp. 49–54.
 M.J. Bronikowski, P.A. Willis, D.T. Colbert, K.A. Smith and R.E. Smalley, Gas-phase
production of carbon single-walled nanotubes from carbon monoxide via the HiPco process: a
parametric study, J Vac Sci Technol A 19 (2001), pp. 1800–1805.
 H. Dai, A.G. Rinzler, P. Nikolaev, A. Thess, D.T. Colbert and R.E. Smalley, Single-wall
nanotubes produced by metal-catalyzed disproportionation of carbon monoxide, Chem Phys Lett
260 (1996), pp. 471–475
 E. Flahaut, A. Peigney, Ch. Laurent and A. Rousset, Synthesis of single-walled carbon
nanotube-Co–MgO composite powders and extraction of the nanotubes, J Mater Chem 10
(2000), pp. 249–252.
 A. Huczko, H. Lange, E. Calko, H. Grubek-Jaworska and P. Droszcz, Physiological testing of
carbon nanotubes: are they asbestos-like?, Fullerene Sci Technol 9 (2001), pp. 251–254
 A.A. Shvedova, V. Castranova, E.R. Kisin, D. Schwegler-Berry, A.R. Murray and V.Z.
Gandelsman et al., Exposure to carbon nanotube material: assessment of nanotube cytotoxicity
using human keratinocyte cells, J Toxicol Env Health A 66 (2003), pp. 1909–1926
 A.M. Maynard, P. Baron, M. Foley, A.A. Shvedova, E.R. Kisin and V. Castranova, Exposure
to carbon nanotube material: aerosol release during the handling of unrefined single-walled
carbon nanotube material, J Toxicol Env Health A 67 (2004), pp. 87–107
 C.W. Lam, J.T. James, R. McCluskey and R.L. Hunter, Pulmonary toxicity of single-wall
carbon nanotubes in mice 7 and 90 days after intratracheal instillation, Toxicol Sci 77 (2004), pp.
 D.B. Warheit, B.R. Laurence, K.L. Reed, D.H. Roach, G.A. Reynolds and T.R. Webb,
Comparative pulmonary toxicity assessment of single-wall carbon nanotubes in rats, Toxicol Sci
77 (2004), pp. 117–125.
 D. Pantarotto, J.P. Briand, M. Prato and A. Bianco, Translocation of bioactive peptides
across cell membranes by carbon nanotubes, Chem Commun 16–17 (2004).
 D. Cui, T. Furong, C.S. Ozkan, M. Wang and H. Gao, Effect of single wall carbon nanotubes
on human HEK293 cells, Toxicol Lett 155 (2005), pp. 73–85.
 C. Salvador-Morales, E. Flahaut, E. Sim, J. Sloan, M.L.H. Green and R.B. Sim,
Complement activation and protein adsorption by carbon nanotubes, Mol Immun 43 (2005), pp.
 E. Bekyarova, Y. Ni, E.B. Malarkey, V. Montana, J.L. McWilliams and R.C. Haddon et al.,
Applications of carbon nanotubes in biotechnology and biomedicine, J Biomed Nanotech 1
(2005), pp. 3–17.
 E. Flahaut, R. Bacsa, A. Peigney and Ch. Laurent, Gram-scale CCVD synthesis of double-
walled carbon nanotubes, Chem Commun 1442–1443 (2003).
 K.C. Patil, Advanced ceramics: combustion synthesis and properties, Bull Mater Sci 16
(1993), pp. 533–541.
 L. Bordenave, C. Baquey, R. Bareille, F. Lefebvre, Ch. Lauroua and V.J. Guerin,
Endothelial cell compatibility testing of three different Pellethanes, Biomed Mater Res 27 (1993),
 ISO document 10993. Biological compatibility of medical devices. 5. Test for cytotoxicity:
in vitro methods. December 1992, revised 1999.
 C.R. Parish and A. Mullbacher, Automated colorimetric assay for T-cell cytotoxicity, J
Immunol Meth 58 (1983), pp. 225–237.
 T.J. Mosmann, Rapid colorimetric assay for cell growth and survival: application to
proliferation and cytotoxicity assays, Immunol Meth 65 (1983), pp. 55–63
 F. Denizot and R.J. Lang, Rapid colorimetric assay for cell growth and survival:
modification to the tetrazolium dye procedure giving improved sensitivity and reliability,
Immunol Meth 89 (1986), pp. 271–277.
 H. Wan, R. Williams, P. Doherty and D.F. Williams, A study of the reproducibility of the
MTT test, J Mater Sci Mater Med 5 (1994), pp. 154–159.
 T. Bachetti and L. Morbidelli, Endothelial cells in culture: a model for studying vascular
functions, Pharmacol Res 42 (2000), pp. 9–19.
 E. Flahaut, F. Agnoli, J. Sloan, C. O’Connor and M.L.H. Green, CCVD synthesis and
characterization of cobalt-encapsulated nanoparticles, Chem Mater 14 (2002), pp. 2553–2558.
Corresponding author : E. Flahaut
Fax: +33 5 61 55 61 63.
e-mail : email@example.com
Volume 44, Issue 6, May 2006, Pages 1093-1099
Toxicology of Carbon Nanomaterials