Synthesis of Boron Nitride from Oxide Precursors

Marriage of Carbon's Neighbors

This is only one mechanism explaining how the things in and around boron nitride might be moving - one of many possible.
 <-  A BN rose


Review of the Synthetic Methods

Formation of Boron Nitride
Chemical Reactions in Boron Nitride
Utilization of the Model


Review of the Synthetic Methods

Despite of the electronic similarity between boron nitride (BN) and carbon, there have been neither graphite-like seams nor diamond-like pits of BN found on our planet and there has therefore been no other way for a man but to fabricate the stuff by himself. Due to a strong oxidative character of the Earth's environment, one can find boron bound just with oxygen. Moreover, to stabilize boron oxide, water or alkaline oxides have to complete the composition to get most of boron-containing resources.

To prepare boron nitride, three consequent processes have to be regarded:

  1. Purification of the ore from alkaline metals and other impurities to get initial materials for boron nitride synthesis (typically boric acid and boron sesquioxide)

  2. Conversion of the raw materials onto substances directly preceding formation of boron nitride (boron nitride precursors)

  3. Completing the process by synthesizing BN powder or polycrystalline solids as ceramics or CVD layers

There have been many synthetic methods described leading to boron nitride. But if they are compared, a common pattern fits to all of them, as shown in the Figure.

If you cannot find your own synthetic method, please do not dismiss this scheme, just abandon the suggested ligands and replace them for your own ones!



Formation of Boron Nitride

So-called "amide method" will be used to model the formation of hexagonal boron nitride planar networks using the ammonia - boric acid system.

In the first step, an ammonia molecule has to approach the surface of a boric acid grain. Due to the redistribution of electron densities in the system, new bonds are created. HO-B-N-H monomer is formed and water is released a s the consequence of the redistribution (scheme A). Afterwards, the monomer is stabilized by the trimerization (scheme B).

The trimer actually is the precursor of boron nitride - it contains the same number of the newly formed B-N and remaining B-O chemical bonds. The growth can progress by reacting of the trimer with other HO-B-N-H monomers and other molecules of water are released, scheme C.

After the third step you can get a macromolecule as shown in the left Figure. The growth can continue till the size of the macromolecule reaches that which can be found in turbostratic BN powders (t-BN). The size is mostly being expressed as the crystallite width, La, and approximately equals 10 nm. A hexagonal macromolecule with the formula HO-B27N27-H corresponds to this size.

Real crystals of BN are constructed from particular monolayers by their stacking each above other. The dipole-dipole interaction mediating stacking the layers is a chemical of the second order; so that each and every monolayer (a macromolecule) is a the chemical unit of hexagonal BN. The width of the BN crystallite is then identical to the size of a macromolecule.

As shown in the previous figures, oxygen and hydrogen must not be understood as impurities, but they represent constitutional elements terminating the two-dimensional h-BN macromolecules. Since the contents of these elements are inversely proportional to the circumference-to-area ratio, it can be assumed that the larger are the BN crystals, the higher is the purity of BN. This relationship is actually the principle of the purification process, where BN is first crystallized by the thermal treatment in nitrogen gas, and oxygen impurities converted to boron oxide are washed out. By excluding the oxygen content we can get a direct dependence of crystallite size on treatment temperature, as shown on the right.

The growth of BN crystals is not linear with temperature, but shows a sharp increase at around 1380°C. Above this temperature the growth becomes milder and practically stops above 1800°C. The crystallization process can be divided into two stages:

  1. In the first stage the crystals quite willingly triple their size, and

  2. in the second stage they triple their size again, although this process is more sluggish, so that the final product consists of crystal roughly 8-10 times larger when the starting powder. In contrast to the "snow ball" mechanism, by heating far above 1000ºC, the crystals multiply their size. The growth can be attributed to coalescence.

Seen from the chemical side, the "snow ball" growth is characteristic with a radical reaction, when the periphery of a BN macromolecule is fed with monomer or oligomer substances, and the growth rather slowly. Above around 1300°C this source is exhausted and the first stage of the coalescence comes up; resulting in so-called mesographitic BN (mBN). In this phase the mobility of BN crystallites is the highest. However, with proceeding the growth, the crystallites gradually loose their mobility and additionally, even the steric conditions for the growth worsen. Despite the obstacles, the crystallites triple their size again to achieve the size typical for well-crystallized ("graphitized") boron nitride (hBN). One might expect two inflection points in the dependence of crystallite size on temperature. The reasons why only one such a point was found might be given as follows:

  • BN is not monophasic but it consists of a continuous spectrum of structural sub-polytypes differing in the size:

  • the first stage (tBN -> mBN) is far more spontaneous then the second one

  • both the coalescence processes occur simultaneously

  • the second stage (mBN -> hBN) interferes with reactions between tBN and mBN

It follows from the above reasons that the average crystallite size of the final product is less then expected 90 nm. Another conclusion is that defining the crystallite size in the mBN is very speculative, since the size is very sensitive to the temperature.



Chemical Reactions in Boron Nitride

A ternary B-N-O diagram will be utilized for graphical description of reactions in boron nitride , although the studied system consists of five elements (H, B, C, N, O), so that the diagram should be rather four-dimensional. To adjust the system to your monitor, the number of the elements was reduced, and the concentrations of the remaining elemental trinity were recalculated to fit 100% in total. The formation of BN from four subspecies of the amide system is indicated with numbers:

  1. boric acid - urea

  2. boric acid - ammonia

  3. boron oxide - urea

  4. boron oxide - ammonia

It was shown in the previous section that the formation of hexagonal boron nitride in the amide system actually is a process of the recombination of -OH and -H leaving groups located on the periphery of planar BN monolayers. After recombining, molecules of water are liberated, and the growth of BN crystallites occurs. This is an ideal case to get pure BN, however, in real systems hydrolytic effect of the liberated water on boron nitride must be considered.

Additional information can be read out from the ternary B-N-O diagram:

  • A pair of the abscissas connecting the points corresponding to the starting substances cross each other in a point expressing the molar ratios between the substances, or in other words, to the composition of the starting mixture ensuring quantitative transformation to the precursor.

  • The distance between the point and that corresponding to the HNBOH precursor is proportional to the amount of gaseous exhausts produced in the process.

  • The vector connecting the precursor and final pure BN is determined by its orientation, characterizing chemical nature of the process (here, the dehydration), and by the magnitude, which is given by the actual yield of the process. It was shown in the second section that the content of constitutional oxygen is determined by the size of BN monolayers so that if a BN material with zero oxygen concentration would be considered, it should be composed from infinitely large crystals. The HNBOH-BN abscissa then expresses the polymorphism of hexagonal BN, the grains of which are composed from a continuous spectrum of polytypes infinitesimally differing in the size and correspondingly even in the oxygen content.

Besides the formation in the amide process, other reactions in the BN system can be displayed in the B-N-O diagram, Fig.3-2. The particular reactions are described below the figure.

1 - formation of the precursor in the amide process; formation (crystal growth) of BN
2 -pressureless sintering
3 - direct synthesis of boron with nitrogen
4 - carbothermal synthesis of BN
5 - hot pressing
6 - hydrolysis of BN
7 - oxidation of BN in dry air/oxygen gas
8 - high-temperature peritectic decomposition of BN in vacuo

1. Formation


1. Formation of the Precursor in the Amide Process

 The process is essentially described in the Section 2. We can understand the precursor as the smallest chemical individual already having hexagonal symmetry with balanced molar fractions of new-born B-N bonds and residual B-O bonds.

2. Crystal Growth and Pressureless Sintering of BN

Again, see the Section 2. Chemically, the growth is based on increasing the B-N / B-O molar ratio, which leads to enlargement of the HO-(BN)n-H macromolecular planes, and consequently to the growth of entire BN paracrystals during the thermal exposure in neutral environment.

Pressureless sintering (PLS) is such a process, where a green body cold-pressed from BN powder is heated. During the annealing the BN crystals further grow and water is released. If the temperature is sufficiently high (close to 2000°C), the sintered product contains a very low amount of oxygen and hydrogen, since due to the openness of the system, water vapors can be released without its harmful hydrolytic effect. However, in the core of the sintered specimen, the highest oxygen content can be found, gradually decreasing towards the surface. This inhomogeneity has its origin in the kinetics of the water releasing: probability of the hydrolytic effect must be proportional to the dwelling time of the released water in the heated body.

For the PLS process, powder with a proper crystallinity has to be chosen. When the crystallinity is very poor, formation of cracks in the body can be expected due to excessive evolution of water. The reverse problem arises, if you use a very well crystallized starting powder. In such a case, you can expect a substantial reduction in the formation of the ceramic skeleton, since you loose the chemical potential for it, primarily represented by the difference in the crystallinity of BN in starting powder and final sintered body.

If we have a look back at the ternary diagrams, we can easily derive the specific amount of released water from the distance between the points representing the composition (crystallinity) of the raw BN powder and final product. If the sintering temperature is high enough, we can neglect the constitutional water in the product and the chemistry of the sintering runs according to the following simplified reaction, Eq. 3-1.

HO - (BN)n - H = n BN + H2O [Eq. 3-1.]

3. Direct Synthesis of Boron With Nitrogen

This process stands out of the center of our attention. However, it should be mentioned here that the direct synthesis requires higher temperatures (by ca. 500 K) necessary for nitrogen atomization, and the product obtained from a pure system has poor crystallinity due to the lack of oxygen. The size of BN grains does not exceed the grain size of raw powder of elemental boron.

4. Carbothermal Synthesis of BN

This process actually is a gradual substitution of nitrogen for oxygen. Although this story is not focused on this process, it might be interesting to investigate this system from the viewpoint of the kinetic role of volatile boron suboxides on the formation of BN. The classical summary equation when boron sesquioxide with carbon gives boron nitride and carbon monoxide is being often used in literature, but the process itself  is not that simple.

5. Hot Pressing

One might think that this is just a physical process like, say, in the alumina system. As described in the Section 3.2., we had better to consider the reaction quantified in Eq. 3-1. as the fundament even for hot pressing (HP). The main difference between the PLS and HP processes is that in the latter the material is trapped in-between the pressing rams and the inner wall of the matrix. The thermally released water is thus forced to decompose boron nitride. If the hydrolysis runs quantitatively, we have to consider one more reaction additionally to the previously mentioned, Eq. 3-2.

3n BN + 3 H2O = B2O3 + (3n-2) BN + 3 H2 + N2 [Eq. 3-2.]

It is obvious that the yield of boron oxide in the product is reversally proportional to the crystallite size of BN and always you will find a higher oxygen content in the sintered body than in the raw powder, see again Figure 3-2., line No. 5. Fortunately, even the BN homologue with the highest oxygen fraction (n = 1) cannot be converted to the oxide completely.

6. Hydrolysis of BN

Hexagonal boron nitride is very sensitive to moisture. The hydrolytic process is the faster the higher is the surface area of the substance. Since the surface area, especially in the case of powders, reflects the size of BN monolayers, it can be said that the hydrolytic rate is proportional to the circumference-to-area ratio. Chemically is the process reverse to that described in the Section 3.2. If the amount of water is sufficient, we can get pure orthoboric acid after ages. Ammonia is then another product of the hydrolysis. Just try to snuff to an unsealed bottle with a turbostratic powder.

7. Dry Oxidation of BN

In any textbook you can find that boron sesquioxide and molecular nitrogen are the products of the oxidation in dry air or pure oxygen gas... The experimental data tell, however, something different: If the oxidation is carried out carefully, a ceramic sample's surface turns yellow to black, which colors are typical for lower oxides of boron (suboxides), and the reaction atmosphere contains a visible amount of NO2. This might suggest that the oxidation process is more complex than it has been admitted, and the formation of final boron sesquioxide runs through several intermediates. Again, this process, in which the fission of the O-O bonds occurs simultaneously with the extinction of the B-N bonds on the periphery of planar BN macromolecules, has Lewisian character.
In the simplest case, the vector characterizing dry oxidation of BN is opposite to that describing the carbothermal process (see again Figure 3-2.), as carbon is not regarded in the B-N-O ternary diagram. Generally, however, the composition of the solid products can be characterized by any point within the triangle defined by the BN, B2O3, and B vertices. This is the reason why a formation of boron suboxides even during the carbothermal reduction of boron sesquioxide, can be expected as suggested in Section 3.4.

8. Peritectic Decomposition of BN

Chemically the easiest, thermally the hardest process: The B-N bond in hBN is strong enough to survive thermal exposure in vacuo at very high temperatures. That's the reason for BN's popularity among the furnace constructors. After all, even this survival has its limit, as close to 3000°C boron nitride starts to decompose. Existence of lower nitrides should be admitted.

Utilization of the Mechanism 

Finally, you might ask what is it all for. Here goes the answer: Knowledge of the mechanism enables helps control the process and thus obtain some non-traditional forms of it. Just have a look:

Hexagonal Boron Nitride Single Crystals

  • The optical micrograph shows a pair of hBN single crystals with their typical hexagonal symmetry. The set of parallel wrinkles show how the growth was proceeding.

  • Experimental conditions: 1700°C, atmospheric pressure.

  • The future prospect: manufacturing of transparent perfectly plane-parallel BN plates with controlled width and thickness as the window material for high-energy radiation sources.


Transparent Glassy Boron Nitride

  • The optical micrograph shows a fraction of a curved transparent glassy platelet. The optical properties of the material can be seen as the difference in the sharpness of the scale below and beside it. Thanks to a strong cross-linking of the BN layers, the material keeps its transparency, glassy appearance and shape stability up to at least 2000°C.

  • Experimental conditions: 1800°C, atmospheric pressure.

  • The future prospect: manufacturing of optical elements with controlled curvature for high-energy radiation optics.


Boron Nitride Nanotubes

  • The SEM micrograph on the left shows a cluster of fibers having smooth surface, circular cross-section, and composition corresponding to that of pure boron nitride. The picture on the right shows that beside the BN tubes, wasp nest-like, apparently hollow spherules were formed in the reaction system, composition of which can be attributed to the pure BN, as well.

  • Experimental conditions: 1050°-1200°C, atmospheric pressure

  • The future prospect: continuous production (up to refining).


Boron Nitride with "Nanosurface"

  • The upper micrograph shows grains of turbostratic boron nitride. The original specific surface area of 3-5 m2/g exceeded 1,000 m2/g after modification of the surface by the selective etching. The originally smooth surface became decorated with islets with the diameter of tens nm.

  • Experimental conditions: 1200°-1400°C, atmospheric pressure

  • The future prospect: utilizing as a carrier for HT catalysts for non-aqueous systems. A method to dope some metals to the material in statu nascendi of the substrate has also been developed.



List of Papers

BN linx  


Is your site missing?



Launched: 10 April 1996

Milan Hubacek

Last update: 18 December 2005