Certification of a boron carbide powder reference material for chemical analysis

Article abstract of DOI:10.1111/j.1551-2916.2011.04657.x

The certification of a boron carbide reference material for chemical composition is described. The mass fractions of 16 elements and two boron species are certified in an international interlaboratory comparison with 35 participating laboratories from six different countries. Beside chemical characterization the certification process includes homogeneity and stability testing of the candidate material boron carbide (type F360, 305M422). Details of the analytical methods used for chemical characterization and of the calculation of the uncertainties of the certified mass fractions are given. The new reference material ERM-ED102 with certified mass fractions of 18 parameters (elements and species) and of the amount fraction of the isotope ¹⁰B is a valuable tool for laboratories working in the field of advanced ceramic materials analysis to improve their analytical results.

Full text of DOI:10.1111/j.1551-2916.2011.04657.x

J. Am. Ceram. Soc., 94 [11] 4009–4016 (2011)
DOI: 10.1111/j.1551-2916.2011.04657.x
©
2011 The American Ceramic Society
ournal
J
Certification of a Boron Carbide Powder Reference
Material for Chemical Analysis
‡
‡
‡
Ralf Matschat, Angelika Dette, Silke Richter, Sebastian Recknagel, and Peter Barth
‡,†
§
‡
BAM Federal Institute for Materials Research and Testing, Richard-Willsta
¨
tter-Str. 11, 12489 Berlin, Germany
§
ESK Ceramics GmbH & Co. KG, Max-Schaidhauf-Str. 25, 87437 Kempten, Germany
The certification of a boron carbide reference material for chem-
ical composition is described. The mass fractions of 16 elements
and two boron species are certified in an international interlabo-
ratory comparison with 35 participating laboratories from six
different countries. Beside chemical characterization the certifi-
cation process includes homogeneity and stability testing of the
candidate material boron carbide (type F360, 305M422).
Details of the analytical methods used for chemical character-
ization and of the calculation of the uncertainties of the certified
be milled and then purified to reduce metallic impurities.
Besides the advantageous mechanical properties B C also
4
shows an outstanding chemical resistance. While this is very
useful for its industrial applications it makes it quite difficult
to analyze its chemical composition including the determina-
tion of impurities. However, chemical analysis must be per-
formed frequently to ensure the constant quality of the
material produced, i.e., its chemical, electrical, physical,
mechanical, and thermal properties. In different fields of
application lists of specification exist concerning the chemical
composition of the specific boron carbide material, because
main-, minor- and trace elements have a great impact to sev-
eral important properties aimed for. The elements boron,
carbon, nitrogen, and oxygen are of interest in monitoring
the manufacturing process. The main components boron and
carbon define the stoichiometric composition of the specific
boron carbide and an accurate determination of boron is
essential for nuclear applications. The amounts of oxygen
and soluble (free) carbon serve as indicators of the complete-
ness of transformation and a high concentration of nitrogen
is a hint of increased influx of air during the melting process.
Soluble boron compounds and oxygen serve also as indica-
tors for the oxidation of boron carbide during milling/
homogenization. Trace elements give information about con-
tamination, e.g., from starting materials or with materials
from other product lines.
®
mass fractions are given. The new reference material ERM -
ED102 with certified mass fractions of 18 parameters (elements
1
0
and species) and of the amount fraction of the isotope B is a
valuable tool for laboratories working in the field of advanced
ceramic materials analysis to improve their analytical results.
I. Introduction
ORON carbide (B
known, only surpassed in its hardness by diamond and
4
C) is one of the hardest materials
B
cubic boron nitride. This makes this material very attractive
for numerous industrial applications. It is widely used as an
abrasive for polishing and lapping as well as for cutting
applications, e.g., for water jet cutting. It can also be
employed for sharpening diamonds and tools, sapphire slic-
ing, and polishing and, due to the excellent wear and abra-
sion resistance, as material for nozzles for grit blasting and
in water jet cutters. Boron carbide materials can also be
found in many precision tools, and it is used as an antioxi-
dant for refractory compositions, e.g., carbon containing
The methods for a comprehensive determination of the
chemical composition of boron carbide are well described
and also Standard Test Methods (e.g. ASTM C-791 ) exist.
3
1
bricks and slide-gates.
The two most commonly used ways to analyze boron car-
bide are wet-chemical analysis after dissolution/decomposi-
tion and solid sampling analysis. Heinrich et al. describe the
The combination of hardness, high elastic modulus, and
low density makes boron carbide well suited as ballistic
armor material. Boron carbide sheets have an extraordinarily
high specific stopping power to defeat/protect against high
dissolution of B
4
C in a mixture of nitric acid, hydrofluoric
4
acid, and sulfuric acid for 12 h at 250°C under pressure.
2
5
6
velocity projectiles.
Blumenthal and Watanabe showed that alkaline fusion
alone or in combination with acidic decomposition is suitable
also to dissolve B C. However, alkaline fusion is mainly used
In addition to its remarkable mechanical properties, boron
carbide has the ability to absorb neutrons very effectively
without forming long-lived radio nuclides. This is due to the
4
if main components, e.g., boron, have to be analyzed. For
trace and ultra trace analysis chemicals used for fusion are
often not clean enough and high salt contents are often not
desirable for the analytical methods performed. Solid
sampling methods such as instrumental photon activation
1
0
high neutron absorption cross section of the B isotope. The
fact that B C is much easier to fabricate in defined shapes
4
than the extremely brittle pure boron makes it the material
of choice as absorbent for neutron radiation in nuclear
power plants. It is used there as shielding material as well as
in control and shut down rods.
7
analysis (IPAA), total reflection X-ray fluorescence spec-
trometry (TXRF), ICP optical emission spectrometry (ICP
8
9
OES) combined with slurry technique, DC arc optical emis-
Boron carbide is mainly produced by carbo-thermal reduc-
1
0
sion spectrometry (DC-arc-OES), electrothermal vaporation
tion of B
2
O
tions. For most commercial applications B
3
in an electric arc furnace or by gas-phase reac-
C must usually
1
1
4
ICP OES (ETV-ICP OES), solid sampling electrothermal
evaporation atomic absorption spectrometry (SS ET AAS) or
carrier gas hot extraction (CGHE, for the elements hydro-
gen, nitrogen, oxygen) can be used for analysis of boron car-
bide and other advanced ceramics. These methods avoid any
time-consuming sample dissolution steps before the analysis,
W. Fahrenholtz—contributing editor
1
2
thus bearing fewer risks of contamination. However, these
techniques normally require reference materials for
calibration and validation which are often not available. To
Manuscript No. 29243. Received January 24, 2011; approved April 29, 2011.
Author to whom correspondence should be addressed. e-mail: sebastian.recknagel@
†
bam.de
4009

4
010
Journal of the American Ceramic Society—Matschat et al.
Vol. 94, No. 11
overcome the lack of suitable certified reference materials
CRMs) in the field of advanced ceramic materials BAM
the same way as the other sub-samples of the homogeneity
study.
(
Federal Institute for Materials Research and Testing in coop-
eration with the Working Group “Special Materials” of the
Committee of Chemists of GDMB Society for Mining, Met-
allurgy, Resource and Environmental Technology has started
a program on reference material development. CRMs based
The determination of the metallic traces Al, Ca, Cr, Cu,
Fe, Mg, Mn, Ni, Ti, Zr was carried out by inductively cou-
pled plasma optical emission spectrometry (ICP OES) using
aliquots of digestion solutions. The digestion procedure of
the sub-samples is described in Ref. [4] An ICP OES spec-
trometer “IRIS-advantage Duo” (Thermo Fisher Scientific
Inc., Waltham, MA) was used for the measurements. To
minimize influences of drifts, drift corrections were made.
The solutions of the sub-samples were measured at two dif-
ferent days and the mean values of the results from both
days of measurements were calculated. To further improve
the precision of the measurements more than one analytical
spectral line was used for analysis (number of spectral lines
in parentheses) for the analytes Al(2), Cr(3), Fe(3), Mg(2),
Mn(3), Ni(2), Ti(3), Zr(3).
1
3
®
on silicon carbide (BAM-S003 ) and silicon nitride (ERM -
ED101 ) are already available. The certification of CRM
ERM -ED102 boron carbide described in this article is
1
4
®
1
5
related to the third material in this series. All reference
materials were certified following the relevant guidelines, i.e.
1
6–18
ISO-Guides 31, 34, and 35,
“Guidelines for the produc-
1
9
tion of BAM reference materials” and “Technical Guide-
lines for the Production and Acceptance of a European
Reference Material”.
20
Homogeneity measurements for Si and Na were carried
out by the direct solid sampling method of ETV-ICP OES
using the spectrometer ICP IRIS Intrepid XSP (Thermo
Fisher Scientific Inc.) in combination with the ETV system
II. Experimental Procedure
1) Candidate Material
About 32 kg of boron carbide material (type F360,
05M422) was taken from a single batch of the customary
(
3
¨
ETV4000 (Spectral Systems, Furstenfeldbruck, Germany)
production line of the producer (ESK Ceramics GmbH &
Co. KG, Kempten, Germany). For the intended purpose of
the reference material the stoichiometry of the boron carbide
powder was not relevant. It was therefore not important to
select a boron carbide powder with a specific B/C stoichio-
metry as candidate material but to select a customary boron
carbide powder with measurable concentrations of relevant
trace elements and where, with respect to grain size, chemical
stability can be assumed.
since the precision of ICP OES with liquid samples was not
sufficient for these elements. For the determination of both
elements Freon R12 (dichlorodifluoromethane) was added.
The temperature in the evaporation step was 2300°C which
is only slightly below the melting temperature of boron car-
bide. The sub-sample mass intake was 2.5 mg.
The homogeneity of the total C distribution was deter-
mined with the combustion instrument Leco WC 200 (Leco
Instruments, St. Joseph, MI) using an inductively heated
furnace and infrared detection. Tungsten and iron granules
were added to the sub-samples. The sub-sample mass intake
was 25 mg.
Oxygen and nitrogen were determined in 50 mg sub-sam-
ples in one step by carrier gas hot extraction (CGHE) using
a Leco TC 436 instrument (Leco Instruments) equipped with
a resistance furnace device with graphite crucible and infra-
red and thermal conductivity detection cells.
The material was bottled into 320 bottles each containing
1
00 g of the material. The particle size of the material was
investigated by laser light diffraction using a Mastersizer
000 (Malvern Instruments Ltd, Malvern, Worcestershire,
U.K.). The results are given in Table I.
2
(
2) Homogeneity Testing
Twenty out of the 320 bottles were representatively picked
by a combination of random access and systematic selection
to test for homogeneity of the whole material and to ensure
that each bottle is representative for the whole batch of 320
bottles. Four sub-samples from each of the 20 bottles were
filled into vials and distributed to each laboratory participa-
ting in the homogeneity investigation. Thus, most analytes
were measured 80 times for the homogeneity testing. For
analytes for which the determination was very time-consum-
ing (B, Na, Si, total C, free C, O, and N), sub-samples from
only 10 of the 20 selected bottles were used for homogeneity
determinations, i.e., 40 determinations in total were
performed.
The mass fraction of total boron was determined using a
titration device (Metrohm AG, Zofingen, Switzerland). The
determination was carried out after an alkaline digestion of
100 mg of material according to Blumenthal followed by
titration with 0.1 mol/L NaOH and addition of mannitol
(mannitoboric acid procedure).
The mass fraction of HNO soluble boron was determined
3
using the same titration device as for total boron. The mea-
surements were carried out after boiling 4 g of the sample in
5
3
1.6 mol/L HNO under reflux followed by titration of the
dissolved boron with 0.1 mol/L NaOH with addition of man-
nitol. The investigation of the water soluble boron was done
using the same analytical technique, but after extraction of
A thoroughly homogenized sample was produced to deter-
mine the repeatability of the methods used for the homoge-
neity investigations. This sample was prepared from about
2
the sample in H O at 60°C. The result was then expressed as
mass fraction of boron oxide (so called “adherent boron
oxide”). Both extraction methods are empirical but they have
been found useful in characterizing boron carbide. It is
assumed that water extracts only boron oxide and that nitric
acid extracts both boron oxide and elemental boron.
2
(
0 g of the material highly homogenized in the “Mixer/Mill”
Spex Industries Inc., Edison, NJ) for 10 min (5 9 2 min)
using polypropylene vessels and balls and distributed to the
laboratories involved in homogeneity testing. The laborato-
ries in most cases took 20 sub-samples (in some cases less)
from the highly homogenized material and analyzed them in
Free carbon was determined from 100 mg sub-samples by
oxidizing carbon to carbon dioxide (CO
sulfuric iodic acid at a temperature of 100°C. The liberated
CO was carried by an inert gas stream to the coulometric
CO2 detection system (Strohlein Bruker Elemental, Kalkar,
2
) using hot chromic
2
¨
2
1
Table I. Particle Size of Boron Carbide Obtained
with Laser Light Diffraction
Germany).
Particle size in lm
(3) Stability Testing
Due to its physical and chemical properties the B
can be assumed to be stable under normal storing conditions.
However, there is a slight chance of instability due to oxida-
tion of boron carbide to boron oxide according to the
formula:
4
C material
D
D
10
21.5
33.6
51.4
60.4
50
D90
D97

November 2011
Certification of Boron Carbide Powder
ð1Þ
4011
B4C þ 4O2 ! 2B2O3 þ CO2
such an interlaboratory comparison for certification is to
obtain results from as many different analytical techniques as
possible. The laboratories were free to use any method they
liked (with only one exception, see below). In this way possi-
ble systematic errors of a specific method should be detectable
and avoided. For the calibration of the methods pure chemi-
cals or solutions prepared from them had to be used to estab-
lish direct traceability to SI. The calibration with matrix
reference materials was therefore not allowed. Using different
analytical methods for element detection is a desired necessity
but not always a sufficient condition to avoid systematic
deviations. Therefore, the application of different dissolution
procedures and the use of liquid as well as solid sampling
techniques are desirable. Table III shows the variety of
analytical methods used. For most of the elements several dif-
ferent methods were used although for many elements ICP
OES prevailed.
In this case the oxygen content should be a sensitive
parameter to indicate an aging of the material. To assess the
long-term stability of the material the oxygen mass fraction
in a boron carbide material with properties very similar to
the candidate CRM was determined over a period of
1
37 months using carrier gas hot extraction. The long term
instability contribution from change of oxygen mass fraction
over a period of 10 years (120 months) was then approxi-
mated by a linear interpolation from the maximum difference
of the values measured at the beginning and the end of the
period of 137 months.
Other constituents of interest in the reference material
could be affected by oxidation as well, namely the mass frac-
2 3 3
tions of B O , in HNO soluble boron, total and free carbon.
To estimate the long-term stability of total carbon, nitrogen,
total boron, boron oxide and free carbon in the candidate
CRM over a period of 10 years (as for oxygen) measurements
over a longer period of time were carried out at a similar
material. The long term stability of the mass fraction of in
Depending on the analytical method different sample
pretreatment procedures were adopted. But in all cases the
material had to be dried at (135 ± 5)°C for 12 h before
weighing. For wet-chemical methods (ICP OES, AAS,
ICP-MS) acid dissolution with hydrofluoric acid, nitric acid,
and in some cases sulfuric acid in closed vessels under pres-
HNO soluble boron was extrapolated for the 10-year period
3
based on measurements at the actual candidate CRM over a
period of 3 years. The certified metallic impurities cannot
become volatile and are therefore only indirectly affected by a
change of the mass of the reference material insofar as their
mass fractions would change if the total mass of the material
increases due to oxidation. Therefore, the total mass of the
material was checked over a 3.5-year period.
sure was mainly used. Alternatively fusion with Na
2 3
CO and
Na ·10H O was carried out. For the elements Al, Ca,
2
4
B O
7
2
Co, Cr, Cu, Fe, Mg, Mn, Na, Ni, Si, Ti, and Zr also solid
sampling techniques which do not require sample dissolution
steps were performed. ETV-ICP OES using Freon R12 to
mobilize the analytes of interest from the matrix was used as
well as DC-arc-OES. For the determination of the elements
Ca, Co, Fe, Mn, Ti, Zr, and N instrumental photon activa-
tion analysis (IPAA) was also employed. Especially for
(
4) Chemical Characterization
nitrogen the latter method represents
alternative to carrier gas hot extraction used by all other
a very valuable
The analytical determinations of the constituents selected for
certification were carried out by all 33 laboratories participating
in the interlaboratory comparison (see Table II). The aim of
7
laboratories.
Table II. Laboratories Participating in the Certification Interlaboratory Comparison
Asahi Glass Ceramics Co., LTD, Development Center, Japan
Bundesanstalt fur Materialforschung und -prufung, Germany
¨
¨
Ceram Testing & Environmental Services, U.K.
Dunhua Zhengxing Abrasive Co., Ltd., China
Eagle Picher Technologies Boron Dept., USA
ESK Ceramics GmbH & Co. KG, Germany
ESK-SiC GmbH, Abteilung MQ, Germany
¨
EUROPAISCHE KOMMISSION Gemeinsame Forschungsstelle, Institut fu
Ferro GmbH, Germany
¨
Forschungszentrum Julich GmbH, Zentralabt. fur chemische Analysen, Germany
¨
r Transurane, Germany
¨
Framatome ANP GmbH, Abt. NGTR, Germany
H. C. Starck GmbH & Co. KG, Germany
HORIBA, Ltd., Application Center, Japan
Japan Analyst Corporation, Japan
JFE Refractories R & D Laboratory, Japan
Johannes Gutenberg Universita
Krosaki Harima Co., LTD., Technical Examination Center, Japan
Leibnitz-Institut fur Festkorper- und Werkstoffforschung, Germany
Max-Planck-Institut fur Metallforschung, Germany
¨
¨
t Mainz, Institut fur Kernchemie, Germany
¨
¨
¨
NSL Analytical Services, Inc., USA
Osram GmbH, Germany
Plansee AG, Werkanalytik, Austria
PTB, Physikalisch Technische Bundesanstalt, Germany
¨
Revierlabor Chemische Laboratorien fur Industrie und Umwelt GmbH, Germany
Rigaku Industrial Corp., Japan
SGL Carbon GmbH, Laboratory Services, Germany
Shanghai Institute of Ceramics, Chinese Academy of Sciences, China
Shinagawa Refractories Co., Ltd., Testing & Evaluation Center, Japan
Taiko Refractories Co., Ltd, Research & Development Laboratory, Japan
Treibacher Industrie AG, Austria
TYK Corporation, Research & Development Center, Japan
Verein fur Kernverfahrenstechnik und Analytik Rossendorf e.V., Germany
¨
Zhuzhou Cemented Carbide Group Corp., Ltd., China

4
012
Journal of the American Ceramic Society—Matschat et al.
Vol. 94, No. 11
Table III. Analytical Methods Used for the Determination of Certified and of Indicative Values
Element
Analytical method used
Number of accepted data sets
Range in mg/kg
Al
DC-arc-OES
ET AAS
1
1
141–177
ETV-ICP OES
ICP-MS
1
2
ICP OES (acid sample digestion)
ICP OES (decomposition by fusion)
14
3
Ca
DC-arc-OES
ETV-ICP OES
2
1
61–135
F AAS
ICP-MS
2
2
ICP OES (acid sample digestion)
ICP OES (decomposition by fusion)
13
2
IPAA
ETV-ICP OES
ET AAS
ICP-MS
ICP OES (acid sample digestion)
IPAA
ET AAS
ETV-ICP OES
ICP-MS
ICP OES (acid sample digestion)
ICP OES (decomposition by fusion)
DC-arc-OES
ET AAS
ETV-ICP OES
ICP-MS
ICP OES (acid sample digestion)
1
1
1
3
3
1
1
1
3
10
2
1
1
1
3
7
1
Co
Cr
Cu
Fe
0.28–0.53
4.0–9.3
1.4–3.2
DC-arc-OES
ETV-ICP OES
599–792
1
F AAS
ICP-MS
2
2
ICP OES (acid sample digestion)
ICP OES (decomposition by fusion)
12
3
IPAA
MAS
1
1
Mg
Mn
DC-arc-OES
ET AAS
ETV-ICP OES
ICP-MS
ICP OES (acid sample digestion)
ICP OES (decomposition by fusion)
DC-arc-OES
1
1
1
3
10
2
1
1.3–7.3
8.1–12.7
ET AAS
ETV-ICP OES
ICP-MS
1
1
2
ICP OES (acid sample digestion)
ICP OES (decomposition by fusion)
IPAA
14
1
1
Na
Ni
ET AAS
ETV-ICP OES
F AAS
ICP-MS
ICP OES (acid sample digestion)
DC-arc-OES
ET AAS
1
1
5
1
2
1
1
5.3–7.6
6.2–11.1
ETV-ICP OES
ICP-MS
1
2
ICP OES (acid sample digestion)
ICP OES (decomposition by fusion)
DC-arc-OES
9
1
1
Si
205–323
ET AAS
ETV-ICP OES
ICP-MS
1
1
1
ICP OES (acid sample digestion)
ICP OES (decomposition by fusion)
MAS
5
1
3

November 2011
Certification of Boron Carbide Powder
4013
Table III. (continued)
Element
Analytical method used
Number of accepted data sets
Range in mg/kg
Ti
DC-arc-OES
ET AAS
1
1
90–105
ICP-MS
ICP OES (acid sample digestion)
ICP OES (decomposition by fusion)
IPAA
2
14
3
1
W
Zr
ICP-MS
ICP OES
DC-arc-OES
ETV-ICP OES
ICP-MS
3
2
1
1
1.1–5.6
37.3–55.4
2
ICP OES (acid sample digestion)
ICP OES (decomposition by fusion)
IPAA
13
1
1
Range in %
Ctotal
Comb.-Coul.
Comb.-Grav.
Comb.-IR
Comb.-Vol.
Coul.
Wet Chem. Oxidation-Coul. (Method M4)
CGHE-Coul.
CGHE-IR
CGHE-TC
IPAA
ICP OES
TITR (Method M1)
ICP OES
TITR (Method M2)
ICP OES
2
1
18
1
1
4
20.5–21.6
Cfree
O
0.39–0.66
1
0.067–0.122
0.172–0.233
78.09–79.06
0.098–0.137
0.056–0.108
11
11
1
2
14
3
N
B
total
B
soluble
4
3
B O
2
3
TITR (Method M3)
ICP-MS
TIMS
6
4
3
1
0
10
11
B /(B + B )
DC-arc-OES, direct current arc optical emission spectrometry; ET AAS, electrothermal atomic absorption spectrometry; ETV-ICP OES, electrothermal vaporiza-
tion inductively coupled plasma optical emission spectrometry; ICP-MS, inductively coupled plasma mass spectrometry; ICP OES, inductively coupled plasma optical
emission spectrometry; FAAS, flame atomic absorption spectrometry; IPAA, instrumental photon activation analysis; MAS, molecular absorption spectrometry;
TIMS, thermal ionization mass spectrometry; TITR, titration; CGHE, carrier gas hot extraction (Coul, coulometry; IR, infrared detection; TC, thermal conductivity
detection); Comb, combustion (Coul, coulometry; IR, infrared detection; Grav, gravimetry; Vol, volumetry.
For the determination of total boron (Btotal), adherent
boron oxide (B O ), and HNO -soluble boron specific
Since free carbon (Cfree) is an exception insofar as it is the
only method dependent parameter, it had to be determined
by all laboratories using the following method described in
detail in the appendix of Ref. [15]: Determination of Cfree in
2
3
3
analytical methods were recommended to the laboratories
see appendix of Ref. [15]). These methods are summarized
in short as given below:
(
B C by wet-chemical oxidation: the C
4
of the sample is oxi-
free
dized to CO
perature of 100°C. The developed CO
gas stream to the coulometric detection system.
2
using hot chromic sulfuric iodic acid at a tem-
is carried by an inert
(1) Determination of Btotal in B
Powdered B C is decomposed by melt-fusion with
4
C by titrimetric method:
2
4
sodium-carbonate or
a
mixture of potassium-
Due to the fact that boron carbide is also used as neutron
absorbing material in nuclear industry, the amount fraction
of B, which is the only neutron absorbing isotope, was also
sodium-carbonate and subsequently dissolved in
hydrochloric acid. The boric acid in the sample solu-
tion is then determined in the presence of mannitol
as mannitoboric acid by potentiometric titration
with sodium hydroxide solution.
10
22
of interest. ICP-MS and thermal ionization mass spectro-
metry (TIMS) were used by the participants to determine this
parameter. Most laboratories used some type of acid diges-
tion for sample preparation. One laboratory performed an
alkaline oxidizing decomposition and another laboratory pre-
pared a mixed suspension of the sample before analysis.
(
2) Determination of HNO -soluble boron in B C by
3
4
titrimetric method: Solid-liquid extraction of HNO
soluble boron in powdered B C is carried out with
boiling 1.6 mol/L HNO . After precipitation of
3
-
4
3
metallic impurities dissolved during the extraction
procedure as hydroxides the boric acid in the sample
solution is determined in presence of mannitol as
mannitoboric acid by potentiometric titration with
sodium hydroxide solution.
(5) Calculation of Certified Mass Fractions and Their
Uncertainties
The statistical evaluation of all reported data was performed
using the program “SoftCRM”. The decision on whether
23
(
3) Determination of B
2
O
3
in B
4
C by titrimetric method:
in powdered B C is
the mass fraction of an element or parameter was suitable
for certification or not was based on the number of different
analytical techniques used and the conformity of the results.
Table III gives an overview of the different analytical meth-
ods used, the number of datasets for each parameter, and the
Solid–liquid extraction of B
2
O
3
4
performed with H O at 60°C. The boric acid in the
2
sample suspension is determined in the presence of
mannitol as mannitoboric acid by potentiometric
titration with sodium hydroxide solution.

4
014
Journal of the American Ceramic Society—Matschat et al.
Vol. 94, No. 11
Table IV. Indicative Values and Their Uncertainties in CRM
ERM -ED102 Boron Carbide
Table V. Certified Values and Their Uncertainties in CRM
ERM -ED102 Boron Carbide
®
®
Parameter
Indicative value ± Uncertainty
Parameter
Certified value ± Uncertainty
Mass fraction in mg/kg
Magnesium
Tungsten
Mass fraction in mg/kg
Aluminum
Calcium
3.2 ± 1.0
3.6 ± 2.1
157 ± 5
97 ± 8
Mass fraction in %
Free Carbon
Cobalt
0.39 ± 0.09
5.6 ± 1.2
2.2 ± 0.4
686 ± 22
10.4 ± 0.5
6.3 ± 0.9
8.0 ± 1.6
268 ± 22
96 ± 5
0.51 ± 0.12
Chromium
Copper
Iron
Manganese
Sodium
Nickel
Silicon
Titanium
Zirconium
Mass fraction in %
Total Carbon
Oxygen
range of the reported results. For Mg, W, and Cfree the data
were not good enough to certify their mass fractions because
of the high spreading among the laboratory mean values.
For Mg the reported mass fractions were between 1.3 and
48.9 ± 2.3
7
.3 mg/kg, i.e., the lowest result was only 18% of the high-
est. For W the range was between 1.1 and 5.3 mg/kg, nearly
a factor of 5. Furthermore, the final results for W as for Cfree
21.01 ± 0.28
0.10 ± 0.04
0.209 ± 0.026
78.47 ± 0.31
0.116 ± 0.013
0.075 ± 0.023
(
range between 0.39 and 0.66 mg/kg) are only based on five
Nitrogen
accepted laboratory datasets. Therefore, the mass fractions
of these three parameters were not certified but given for
information only (Table IV).
For all the other 19 parameters (elements and species) the
number of datasets and the quality of data were suitable for
certification (Table V). The certified mass fractions were cal-
culated as the mean values of the accepted laboratories’
means, derived from six independent determinations
Total Boron
Soluble Boron
Boron oxide
Amount fraction in %
10
†
Boron
19.907 ± 0.014
† 10
Amount fraction of Boron related to total amount of Boron.
performed for each parameter. The combined uncertainty u
of the certified mass fractions [Eq. (2)], calculated in accor-
dance with the Guide to the Expression of Uncertainty in
Measurement (GUM) and ISO-Guide 35, contains contri-
butions from the interlaboratory comparison for certification,
from the (potential) inhomogeneity of the samples (contribu-
tions from within and between the bottles) and from the
c
sffiffiffiffiffiffiffi
sffiffiffiffiffiffiffi
2
s
2
HS
4
ubb ¼
ꢁ
ð4Þ
nHS
v
2
HS
24
19
s
nHS: number of parallel measurements at homogeneous
sample.
ms : degrees of freedom for calculation of s²HS
In all cases in which the variance representing the contribu-
tion of the inhomogeneity s²inhom was greater than the variance
u²bb Eq. (2) was used for the calculation of the uncertainty. In
(
1
potential) time instability of the samples over a period of
0 years after certification.
2
HS
.
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
M
s
2
inhom
2
lts
uc ¼
þ s
þ u
ð2Þ
case of s²inhom ꢂ u²bb then u
c
was calculated using Eq. (5)
n
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
M
s
sM: standard deviation of the accepted laboratory mean
values of interlaboratory comparison for certification.
n: number of accepted data sets.
2
2
lts
uc ¼
þ u þ u
bb
ð5Þ
n
s
inhom: standard deviation resulting from (potential) inho-
mogeneity of the samples calculated using Eq. (3)
This was the case for Mn and Si only. For W (given only
1
0
for information) and the amount fraction of B (no inhomo-
geneity to be expected) no inhomogeneity contribution was
included into the respective combined uncertainties u . In
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
À
Á
À
Á
standard deviation of homogeneity investigation
2
b
2
HS
2
2
HS
c
sinhom ¼
s ꢀs
þ s ꢀs
ð3Þ
w
Table VI the specific contributions to the combined uncer-
tainties for each element are given. The expanded uncertain-
ties U of the certified mass fractions given in the certificate
were calculated by multiplication of u with a coverage factor
s
between the bottles” (which in principle contains a contribu-
tion of the standard deviation of the analytical procedure
used in homogeneity investigation).
b
:
“
c
2
4
of k = 2 according to GUM as
s
within the bottles” (which in principle contains a contribu-
w
:
standard deviation in homogeneity investigation
U ¼ 2uc
ð6Þ
“
tion of the standard deviation of the analytical procedure
used in homogeneity investigation).
sHS: standard deviation in homogeneity investigation of
III. Conclusion
®
The new reference material ERM -ED102 with certified
mass fractions of 18 parameters (elements and species) and
“
homogeneous sample.” The value of sHS is assumed
1
0
to represent the standard deviation of the analytical proce-
dure used for the homogeneity investigation. In case it is
higher than sb or sw the respective term in brackets is set
to zero.
of the amount fraction of the isotope B were developed
and is now available for laboratories working in the field
of advanced ceramic materials analysis. Careful material
selection and preparation as well as accurate homogeneity
ults: uncertainty contribution from potential long term
investigation proved that the material has a sufficient
instability of the corresponding parameter.
degree of homogeneity. The certified values are based on
analyses carried out by 33 participating laboratories world-
wide. The expanded uncertainties are low enough to assure
the aimed applicability of the CRM. Especially for the
main components boron and carbon the relative uncertain-
1
9
Following ISO Guide 35 the parameters ubb [see Eq. (4)]
and u²bb were calculated as well. u²bb represented the blind
part of the variances which could be masked by the variance
of the analytical procedure s²HS
.

November 2011
Certification of Boron Carbide Powder
4015
Table VI. Calculation of Combined Uncertainties (Explanation of Symbols See Text)
Al
Ca
Co
Cr
Cu
Fe
Mg
Mn
s
n
M
8.7882
22
17.3778
23
0.0818
9
1.3044
17
0.6260
13
44.2650
23
2.0095
18
0.9919
21
s
s
s
ubb
u
u
b
2.4750
1.7490
2.0820
0.265176
0.5345
2.3636
4.73
1.2560
1.2150
1.1170
0.142268
0.3284
3.714
7.43
0.0312
0.0145
0.0114
0.001574
0.001337
0.0410
0.0820
0.09
0.4390
0.3170
0.1410
0.018676
0.01916
0.595
1.19
0.2090
0.1590
0.1920
0.024454
0.00758
0.1921
0.384
5.6840
6.9300
5.3890
0.686376
2.3334
10.603
21.21
22
0.2050
0.0850
0.1370
0.018912
0.01091
0.4981
0.996
0.0830
0.1270
0.1430
0.018213
0.03527
0.2198
0.440
w
HS
lts
c
U
~
U
5
5
8
8
1.2
1.2
0.4
0.4
1.0
1.0
0.5
0.5
†
U
beg
0.09
21
Na
Ni
Si
Ti
W
Zr
C-total
C-free
sM
n
0.8012
10
1.6364
15
36.3096
13
4.6300
22
2.2509
5
4.0316
20
0.2589
22
0.1178
5
s
s
b
0.3408
0.2134
0.1512
0.032828
0.0214
0.4245
0.84910
0.9
0.6430
0.2980
0.2350
0.032440
0.02727
0.7555
1.51
9.8700
15.2600
17.0000
3.6910
0.91041
10.764
21.53
22
2.1060
1.8350
1.5180
0.19334
0.3261
2.068
4.14
1.2140
1.2040
1.1120
0.141631
0.16627
1.1362
2.27
0.0590
0.0440
0.0375
0.00814
0.11680
0.1388
0.278
0.0240
0.0225
0.0186
0.00604
0.01045
0.05721
0.114
0.12
w
sHSt
u
u
u
bb
lts
c
0.01222
1.0071
2.01
2.1
2.1
U
~
U
1.6
1.6
5
5
2.3
2.3
0.28
0.15
†
U
beg
0.9
22
0.12
O
N
B-total
B-sol.
B
2
O
3
B
isotope ratio
s
n
M
0.0174
12
0.0198
12
0.3304
16
0.0116
7
0.0162
9
0.0183
7
sb
sw
0.0036
0.0025
0.0027
0.0064
0.0047
0.0023
0.0004994
0.00928
0.01308
0.0262
0.026
0.2328
0.2257
0.1979
0.0285397
0.06914
0.1526
0.305
0.0031
0.0040
0.0019
0.0006169
0.00121
0.00626
0.0125
0.013
0.0014
0.0021
0.0015
0.0002828
0.00997
0.01143
0.0229
0.023
s
HSt
u
bb
ults
0.0005862
0.0166
0.01751
0.0350
0.035
u
c
0.00694
0.0139
0.014
U
~
U
Ubeg
0.31
0.28
†
0.011
0.018
0.012
0.011
0.014
†
U
beg = expanded uncertainty without the contribution of the potential long time instability.
4
ties of 0.4% and 1.3% respectively are low. The B/C-ratio
was found to be 4.15, i.e., not far from the theoretical
ratio of 4.3. The calculated combined uncertainties contain
contributions from the interlaboratory comparison, as well
as from potential inhomogeneity or aging of the material.
The CRM enables the users to calibrate their solid sam-
pling methods with a well-characterized material. It can
also be used for validation and verification in case of
H.-J. Heinrich and R. Matschat, “Investigations on Freon-Assisted Atomiza-
tionofRefractoryAnalytes(Cr,Mo,Ti,V)inMultielementElectrothermalAtomic
Absorption Spectrometry,” Spectrochimica Acta Part B, 62, 807–16 (2007).
5
H. Blumenthal, “Determination of Boron in Metal Borides,” Analytical
Chemistry, 23, 992–4 (1951).
M. Watanabe, “Precise Determination of Boron in Boron Carbide Powder
by Stepwise Pressure Alkali/Acid Decomposition/Mannitol Titration Method,”
Bunseki Kagaku, 50, 139–41 (2001).
S. Merchel and A. Berger, “Determination of Nitrogen in Boron Carbide
6
7
by Instrumental Photon Activation Analysis,” Analytical and Bioanalytical
Chemistry, 388, 385–9 (2007).
M. A. Amberger and J. A. C. Broekaert, “Direct Multielement Determina-
®
wet-chemical analyses. ERM -ED102 broadens the range
8
of CRMs in the field of high technology ceramic materials.
For the future the preparation of a high purity silicon
carbide reference material and a boron nitride material as
well as two graphite CRMs is planned. The material is
available from BAM Federal Institute for Materials
Research and Testing.
tion of Trace Elements in Boron Carbide Powders by Total Reflection X-Ray
Fluorescence Spectrometry and Analysis of the Powders by ICP Atomic Emis-
sion Spectrometry Subsequent to Sample Digestion,” Journal of Analytical
Atomic Spectrometry, 24, 1517–23 (2009).
9
Z. Wang, D. Qiu, G. Taoa, and P. Yang, “Direct Determination of Trace
Metals in Boron Carbide by Slurry Introduction Axial Viewed Inductively
Coupled Plasma Optical Emission Spectrometry,” Journal of Analytical Atomic
Spectrometry, 24, 1258–61 (2009).
1
0
¨
T. Kantor, J. Hassler, and O. Forster, “Determination of Trace Metals in
References
Industrial Boron Carbide by Solid Sampling Optical Emission Spectrometry.
Optimization of DC arc Excitation (Current, Atmosphere and Chemical Modi-
fier),” Microchimica Acta, 156, 231–43 (2007).
1
J. Greim and K. A. Schwetz, Boron Carbide, Boron Nitride, and Metal
Borides, Ullmann’s Encyclopedia of Industrial Chemistry, Vol. 5, Wiley-VCH,
Weinheim, 2006.
J. C. LaSalvia, J. Campbell, J. J. Swab, and J. W. McCauley, “Beyond
11
R. Matschat, J. Haßler, H. Traub, and A. Dette, “Multielement Trace
Determination in SiC Powders: Assessment of Interlaboratory Comparisons
Aimed at the Validation and Standardization of Analytical Procedures with
Direct Solid Sampling Based on ETV ICP OES and DC arc OES,” Analytical
and Bioanalytical Chemistry, 383, 1060–74 (2005).
2
Hardness: Ceramics and Ceramic-Based Composites for Protection,” JOM, 62,
1
6–23 (2010).
3
ASTM standard C791 – 04, Standard Test Methods for Chemical, Mass
12
E. Reitznerova, K. Florian, J. Hassler, and R. M. Barnes, “Direct and
Indirect Atomic Spectrochemical Analysis of Boron Carbide,” Acta Metallurg-
ica Slovaca, 3, 640–4 (1997).
Spectrometric, and Spectrochemical Analysis of Nuclear-Grade Boron
Carbide. ASTM International, West Conshohocken, PA, 2004, DOI: 10.1520/
C0791-04, http://www.astm.org.

4
016
Journal of the American Ceramic Society—Matschat et al.
Vol. 94, No. 11
1
3
19
R. Matschat and A. Dette, Certification Report of BAM-S003 “Silicon
Carbide Powder”. Federal Institute for Materials Research and Testing
BAM), 2004. (http://www.rm-certificates.bam.de/de/rm-certificates_media/rm_
cert_special_materials/bam_s003repe.pdf)
BAM Federal Institute for Materials Research and Testing, Guidelines for
the Production of BAM Reference Materials, 2006. (http://www.bam.de/en/
fachthemen/referenzmaterialien/referenzmaterialien_medien/bam_rm_guidelines.
pdf)
(
1
4
®
20
R. Matschat and A. Dette, Certificate of ERM -ED101 “Silicon Nitride
Powder”. Federal Institute for Materials Research and Testing (BAM),
001. (http://www.rm-certificates.bam.de/de/rm-certificates_media/rm_cert_special_
materials/erm_ed101e.pdf)
ERM Technical Committee, Technical Guidelines for the Production and
Acceptance of a European Reference Material, 2007 (http: //www.erm-crm.
org)
K. A. Schwetz and J. Hassler, “A Wet Chemical Method for the Determi-
2
2
1
1
5
®
R. Matschat and A. Dette, Certification report of ERM -ED102 “Boron
Carbide Powder”. BAM Federal Institute for Materials Research and Testing,
008 (http://www.rm-certificates.bam.de/de/rm-certificates_media/rm_cert_special_
materials/erm_ed102repe.pdf)
nation of Free Carbon in Boron Carbide, Silicon Carbide and Mixtures
Thereof,” J. Less-Common Met., 117, 7–15 (1986).
J. E Rein and R. M. Abernathey, “Rapid Mass-Spectrometric Determina-
2
2
2
tion of Boron Isotopic Distribution in Boron Carbide,” Talanta, 19, 857–62
(1972).
G. Bonas, M. Zervou, T. Papaeoannou, and M. Lees, “‘SoftCRM’: A
1
6
International Standards Organisation, ISO Guide 31, Contents of Certifi-
cates of Reference Materials, 1981, Geneva.
2
3
1
7
International Standards Organisation, ISO Guide 34, General Require-
ments for the Competence of Reference Material Producers, 2000, Geneva.
New Software for the Certification of Reference Materials,” Accred. Qual.
Assur., 8, 101–7 (2003).
International Standards Organisation, Guide to the Expression of Uncer-
1
8
24
International Standards Organisation, ISO Guide 35, Reference Materials –
General and Statistical Principles for Certification, 3rd edition, 2006, Geneva.
tainty in Measurement, 1993, Geneva.
h