Fusion

SECTION 2 CHAPTER 4
 Fusion

Introduction
Borate fusion is an extremely effective method of preparing
cement, refractories, ceramics, rock, and similar materials for ele-
mental analysis by XRF, AA, and ICP. The samples are mixed in
powdered form with a flux, either lithium tetraborate or a mixture
of lithium tetraborate and lithium metaborate. The sample must be
ground to a fine powder before the fusion; the more consistent the
particle size, the more reproducible and accurate the fusion will be.
The sample mixture is heated until the flux melts and the sample
dissolves in it, yielding a homogenous melt. The sample forms
borate salts and this eliminates particle size and mineralogical
effects. The use of borate fusions will also minimize matrix effects
which are seen in XRF analysis. The melt can be cast as a glass disk
for XRF, or quickly dissolved in dilute nitric acid or hydrochloric
acid for analysis in solution by AA or ICP. Recent advances also
allow borate fusion of samples containing sulfides, ferroalloys, etc.
One of the advantages of borate fusion is the short preparation
time, typically ten to fifteen minutes to make glass disks or solutions
with automated fluxers. For samples prepared as liquids, borate
fusion can be quicker overall than microwave dissolution in pressure
vessels, and the use of hazardous acids (e.g. HF) is avoided.
SPEX SamplePrep offers two approaches to borate fusion:
the Katanax K1 and K2 Automated Electric Fluxers for rapid,
reproducible fusions, and graphite crucibles for smaller-scale opera-
tions with muffle furnaces. SPEX SamplePrep also provides the full
line of SPEX CertiPrep fusion fluxes.
FUSION

When to Employ Borate Fusion

Borate fusions are widely used for samples which are either
difficult to prepare as homogenous pressed powder (e.g. cement),
hard to dissolve in acid (e.g. zirconia and alumina), or both (e.g.
metal ores and silicate rocks). For borate fusions to be successful,
the sample when fused must be in the form of an oxidized, inorgan-
ic compound. Cement is usually a blend of carbonates and silicates;
zirconia and alumina are oxides; and so forth. Compounds without
oxygen such as sulfides, carbides, chlorides, etc, must be oxidized
before being fused. Reduced metals must also be oxidized. Organic
compounds must be ashed. (An example of this is the analysis of

214 Additional Technical Advice at www.spexcsp.com

 Fusion

metallurgical coal. The coal sample must be ashed and the ash is
then fused. The sulfur in the coal will volatilize in the process and
therefore sulfur cannot be measured). Once fusion is complete, the
melt can be cast as a glass disk or poured into a dilute acid solution
and dissolved. Platinum-group metals cannot be fused with borates
because these compounds reduce during fusion and the metals will
not only remain insoluble in the flux but can also alloy with the 95%
platinum/5% gold crucibles and damage or destroy them.
Many users of the K1 and K2 Automated Electric Fluxers are in
the following industries: cement, glass, ceramics, mining and miner-
als. Samples analyzed include not only raw materials like dolomite,
sand, basalt and iron ore, but also their industrial products and by-
products such as cement, building materials and mining concentrates.
Additional industrial samples include pigments such as TiO2, and
slags from smelters, blast furnaces, refineries and glass plants.
Most of these samples are naturally oxygen–rich and do not
require chemical transformation prior to borate fusions. However,
hybrid oxidation/fusion techniques have been developed for reliable
borate fusions of sulfides, carbides and some ferro-alloy materials
formerly considered out-of-bounds for the technique.
Borate fusion has become increasingly popular as a preparation
technique for XRF sample disks. Because fusion eliminates particle
size and mineralogical effects and produces a homogeneous sample,
it has proven to be the best method for materials that have these
characteristics when X-ray fluorescence analysis is the method of
choice. The fusion method will reduce matrix effects but not
eliminate them. Borate fusion may not be the method of choice if
FUSION

the analyst is interested in trace analysis, since the sample is diluted

during fusion. There have been recent advances in improving the
performance of fusions for trace analysis.
The pressed powder method can be highly accurate when care-
fully done with multi-phase samples such as cements, rock, sand and
ore but such samples are subject to segregation during grinding and
pressing, and to matrix effects (e.g. particle size and mineralogical
effects). Borate glass disks are easier to preserve than pressed powder
disks because they are stable if carefully stored in desiccators.
Synthetic standards for XRF can also be made from pure oxides
with the borate fusion method, as borate glass is essentially a solid
solution with few matrix-matching problems.

Additional Technical Advice at www.spexcsp.com 215

 Fusion

In preparing samples for AA, ICP, and other liquid-analyzing

techniques, the major advantage of borate fusion is that it is often
simpler and quicker than dissolution with acid in a microwave pres-
sure vessel. A complete fusion/solution procedure, from ignition of
the heating elements to decanting of a clear solution, can take fif-
teen minutes or less in an automated fluxer. While borate fusions do
require some caution in evacuating heat and fumes, and the use of
dilute HCl or HNO3 to dissolve the melt, hazardous reagents such
as HF and other concentrated mineral acids are not necessary.
Borate fusion methods offer a wide range of applications but
may not be suitable for all materials. Fusion destroys the original
form of the sample, so structural and molecular information should
be measured before the fusion is made. The high temperature of
borate fusion (1000˚ to 1150˚ C) drives off compounds of volatile
metals such as Hg, Sn, and Sb, while other compounds form during
fusion. Extra steps necessary to prepare organic materials and
reduced inorganics for fusion can extend turnaround time but still
may be the most accurate method to choose. For many samples
borate fusion is the simplest, quickest, and most accurate analytical
approach.
Both fusion and pulverizing/pressing are important and
widely used sample preparation methods, each with its own
advantages. SPEX SamplePrep has a full range of equipment for
either approach. Please consult our applications specialists to help
determine which method is more suitable for your laboratory.

How to Prepare a Fused Sample

FUSION

Fusions are accomplished in several steps. First the sample is

mixed with a flux in an appropriate ratio (usually between 1:2 and
1:10), with the addition of a non-wetting agent to prevent the flux
from sticking to the crucible and the mold. Typical amounts of
flux/sample mixture are 6 to 7 grams for a 31 mm glass disk, and
1 to 2 grams for a solution. The sample is heated past the melting
point of the flux in an inert, heat-resistant crucible. Most borate
fusions are performed in crucibles made of 95% platinum and 5%
gold, a standard non-wetting alloy. Some borate fusions are done in
graphite crucibles.

216 Additional Technical Advice at www.spexcsp.com

 Fusion

During the fusion, heating is maintained and the crucible regu-

larly agitated until the sample has completely dissolved in the molten
flux. At this point the melt is either poured into a mold and annealed
to form a glass disk for XRF, or poured into dilute mineral acid (e.g.
10% HNO3) and stirred until the glass flux dissolves. In some cases
(notably pyrosulfate fusions) the melt is left to harden in the crucible,
and the crucible and the glass together are placed in an acid solution
to dissolve the glass.
It is possible to achieve great sample throughput with an auto-
mated, programmable fluxer. The Katanax K1, when programmed
for cement samples, can produce up to seven fused glass disks in an
hour, using 0.6 grams of cement mixed with 6 grams of flux and
about half a percent of LiBr, a non-wetting agent. The 5-position
K2, when programmed for similar samples can produce up to 25
fused disks per hour.
The same borate fusion procedures can be carried out manually
in a muffle furnace with SPEX SamplePrep graphite crucibles. The
larger flat-bottomed SPEX SamplePrep crucibles can be used to cast
glass disks as well as perform fusions, while the crucibles are handled
and agitated with SPEX SamplePrep tongs. The time per sample
with the muffle furnaces is usually much longer than with an auto-
mated fluxer.
Each approach has its advantages. Again, please contact our
applications specialists to determine the optimum equipment for
your requirements.
FUSION

Additional Technical Advice at www.spexcsp.com 217

 Fusion

How to Select a Fusion Flux

Most fusions involve the use of lithium tetraborate (Li2B4O7,
M.P. 920˚ C), lithium metaborate (LiBO4, M.P. 845˚ C), or a mixture
of the two. As a rule lithium tetraborate is better suited for the disso-
lution of basic oxides, and is preferred for cement and most ores.
Lithium metaborate or “met/tet” mixtures are more suitable for
acidic oxides such as silicate rocks and silica-alumina refractories.
Individually or together, these lithium borates will dissolve oxides,
carbonates, silicates, sulfates, etc. Metals, sulfides, nitrates, carbides,
phosphides, etc. cannot be fused directly in lithium borate fluxes,
and will often attack platinum-gold crucibles, or alloy with them.
However, many of these materials can be first oxidized with standard
techniques and then successfully fused. Methods have been devel-
oped for fusing sulfide-rich material such as copper ore. The sample
is mixed with lithium or sodium nitrate and preheated to oxidize the
sulfide to sulfate. When this has been done the fusion can proceed as
normal without any loss of sulfur from the fusion.
Other fluxes include sodium tetraborate (Na2B4O7), sodium
metaphosphate (NaPO3), and potassium pyrosulfate (K2S2O7). These
have lower melting points than lithium borate fluxes and more spe-
cialized applications. Alternative fluxes and fluxing techniques are
discussed in Bock (1979) and in Sulcek and Povondra (1989), both
cited in the bibliography at the back of this Handbook. Melting
point may be a factor in the selection of a flux, as the higher temper-
ature of a fusion, the greater the degree of volatilization. However
the utility of lithium tetraborate and lithium tetraborate/metaborate
mixtures is so great that most analytical fusions are carried out with
FUSION

these fluxes at temperatures between 1000˚ and 1150˚ C.

SPEX SamplePrep offers the full line of SPEX CertiPrep fusion
fluxes. This includes lithium tetraborate, lithium metaborate, several
lithium tetraborate/metaborate blends, sodium tetraborate, and a
lithium tetraborate/lithium carbonate mixture. Many of these fluxes
are available in pure and ultra-pure grades. All are supplied with a
certificate of analysis for trace metal impurities.
The most popular blended flux, 2:1 “tet:met,” is also available
with added LiBr non-wetting agent for greater convenience. LiBr
solution is available separately for quick addition of a non-wetting
agent to a flux.

218 Additional Technical Advice at www.spexcsp.com

 Fusion

Most SPEX CertiPrep fusion fluxes are of the “glass bead”

type, pre-fused and coarse-grained for maximum homogeneity and
ease of use, and minimum absorption of moisture. The exceptions
are the lithium borate/carbonate flux (a physical mixture) and so-
called “fluffy” lithium tetraborate. The latter is fine-grained and less
dense than the “glass bead” fluxes, but is available for those who
find its physical characteristics superior for their application.
All fusion fluxes sold by Spex SamplePrep are manufactured to
the highest standards available. Spex SamplePrep guarantees both
the purity and the quality of the fluxes and additives supplied to our
customers.

Additives for Fusions: Non-Wetting Agents,

Fluidizers, and Oxidizers
Non-wetting agents (NWA) are iodides and bromides which
can be added in small quantities to a fusion so the molten flux will
not stick to the crucible or mold. The visible effect of a non-wetting
agent is to increase the surface tension of the melt. A fused disk
with too little NWA will have a concave upper surface and may be
difficult to remove from the mold, whereas a molten flux bead with
excessive NWA will ball up when poured and not form a complete
disk.
When glass disks for XRF are being made, a non-wetting agent
is mixed with the flux and the sample before fusion starts. Typically
the amount of NWA is about 0.2% of the weight of the flux, e.g.
12 mg of NWA for 6 grams of flux. Certain samples such as iron
ores, which greatly increase the “stickiness” of a melt, require addi-
FUSION

tional NWA. As non-wetting agents gradually volatize during a

fusion, somewhat longer fusions (as for some technical ceramics)
may also need greater amounts of NWA. The ideal amount of
NWA for a specific procedure is usually determined by experiment.
When making solutions by pouring the molten flux into a
dilute mineral acid, it is desired to have complete transfer from the
crucible to the beaker. This can require a much higher proportion
of non-wetting agent than is necessary to pour a glass disk. A quan-
tity of flux plus sample not exceeding 2 grams might require 50 to
100 mg of NWA.

Additional Technical Advice at www.spexcsp.com 219

 Fusion

Lithium iodide and bromide are popular non-wetting agents

because they do not add an impurity to the flux. However lithium
bromide is hydroscopic, so it is usually made into a saturated solu-
tion and added to the flux from a dropper bottle. Lithium iodide,
sodium iodide, and cesium iodide are somewhat more air-stable,
and easier to use as solids. While it is simpler to add a drop or two
of NWA than it is to weigh out 10 or 20 mg of a solid, liquid NWA
cannot be added to a hot crucible while a fusion is in progress. Note
that non-wetting agents should be used with care when copper-
bearing samples are being fused, as copper halides are extremely
volatile.
Lithium fluoride can be used as a fluidizer, lowering the
melting point of a flux and making it flow far more easily. At
10% by weight, it lowers a flux’s melting point by about 100˚ C.
Oxidizers such as lithium nitrate and sodium nitrate can be
useful in eliminating unoxidized components from a sample that
will not fuse. Graphite, often present in cement mix, is relatively
harmless but can leave a black film on a glass disk or even cause it to
crack. Graphite can be oxidized to CO2. Other sample components
such as phosphides and sulfides may be corrosive enough to damage
or wreck a crucible in a single fusion. If they are oxidized to phos-
phates and sulfates they will be comparatively harmless, and their
cations will be present in the fused glass disk for analysis. As oxidiz-
ers have much lower melting points then borate fluxes, any fusion
including them should proceed at a low temperature until oxidation
is complete.

Graphite Crucibles and Handling Equipment

FUSION

SPEX SamplePrep graphite crucibles

are a cost-effective alternative to the metal
crucibles (platinum/gold, zirconium, etc.)
used to contain samples during fusions. SPEX
SamplePrep graphite crucibles are disposable,
eliminating both the need for time-consuming
cleaning procedures and the possibility of
cross-sample contamination. Chemically inert and heat-resistant,
graphite will not combine with samples during fusion. And SPEX
SamplePrep’s unique nickel-chromium rack and tongs make han-
dling easy.

220 Additional Technical Advice at www.spexcsp.com

 Fusion

SPEX SamplePrep graphite crucibles are available in two

purities: high-purity, for trace-level analytical work, and economical
regular-purity, good for most fusions. SPEX SamplePrep high-
purity graphite has total metallic impurity limits of 2 ppm. SPEX
SamplePrep regular-purity graphite typically has from 10 to
100 ppm each of Mg, Fe, Al, Si, and Cu; as these impurities are
distributed throughout the graphite, contamination of the sample
is minimal.
SPEX SamplePrep Cat. No. 7152 is the standard 9 mL fusion
crucible found in most laboratories; long-time users should note the
introduction of the 7152HP, the same crucible in purified graphite.
For fusions where the 7152 is too small, SPEX SamplePrep Cat. No.
7156 is a standard graphite crucible with about three times the capaci-
ty of the 7152. (Cat. No. 7155 is the high-purity version of the
7156).
SPEX SamplePrep flat-bottomed graphite crucibles (Cat. No.
7157– 7162) are designed both as fusion crucibles and as molds for
XRF cast glass disks; fusion of the sample and annealing of the disks
can be completed in the same container. These crucibles have
tapered walls to facilitate removal of the glass disk. The three
different crucible sizes (31 mm, 33.6 mm, and 40 mm) produce
disks which will fit the standard XRF masks of 31 mm, 35 mm,
and 40 mm. The crucible bottoms are flat but not polished, so that
for greatest analytical accuracy the disks should be flat-lapped.
While graphite is an ideal crucible material for many applica-
tions, withstanding temperatures well in excess of any required for
even the most difficult borate fusions, it does oxidize slowly above
430° C; over a period of hours some erosion of the crucible can
FUSION

occur, the graphite converting to CO2. Because of this, graphite is

not recommended for extremely lengthy fusions, or for fusions
where the sample might be reduced.
SPEX SamplePrep also offers a complete line of borate fusion
fluxes and non-wetting agents.

Additional Technical Advice at www.spexcsp.com 221

 Fusion

Katanax K1 and K2 Automated Electric Fluxers

Features and Benefits
• K1 – Single-position automated
electric fluxer.
• K2 – Five-position automated
electric fluxer.
• Prepares glass disks for XRF and
solutions for AA and ICP.
• Fusion without Flames™—uses no
propane or LPG.
• Automated control of fusion
parameters including heating time
and temperature, mixing time and
rate, and cooling time for glass disks
or stirring time for solutions.
• Fusion wizard for creating
customized recipes.
• Fuse up to 5-7 samples an hour in
Katanax K1 and Katanax K2 the K1.
Automated Electric Fluxers
• Fuse up to 20-24 samples an hour
in the K2.
• Multi-lingual, digital interface.
• Easy to use and maintain.

Sold and serviced by SPEX SamplePrep, a company with over

FUSION

50 years of service to spectroscopists.

SPEX SamplePrep provides all of the necessary supplies for

a problem-free fusion experience including fluxes, additives
and 95% Pt/5% Au crucibles and molds. See page 94 in this
Handbook for a detailed product listing or call SPEX SamplePrep
at 1-800-LAB-SPEX x465 for all your fusion needs.

222 Additional Technical Advice at www.spexcsp.com

 Fusion

FUSION

Additional Technical Advice at www.spexcsp.com 223