An investigation on the synthesis of borazine

Article abstract of DOI:10.1016/j.ica.2010.10.030

Borazine is a promising precursor for boron nitride. A detailed investigation on the reaction of sodium borohydride and ammonium sulfate from 40 °C to 120 °C for synthesis of borazine was performed. The reaction was monitored by means of ¹¹B nuclear magnetic resonance (¹¹B NMR) and Fourier transform infrared spectroscopy (FTIR), X-ray diffraction (XRD), mass spectroscopy (MS). The reaction produces mainly ammonia borane (AB), but not borazine at temperatures below 60 °C. Increases of temperature promote yield of borazine, which reaches the maximum around 110 °C. Whereas further increased temperature causes severe polymerization of borazine, and hence holds back yields of borazine.

Full text of DOI:10.1016/j.ica.2010.10.030

Inorganica Chimica Acta 366 (2011) 173–176
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
An investigation on the synthesis of borazine
⇑
Jun-sheng Li , Chang-rui Zhang, Bin Li, Feng Cao, Si-qing Wang
State Key Laboratory of Advanced Ceramic Fibers and Composites, College of Aerospace and Materials Engineering, National University of Defense Technology,
Changsha, Hunan 410073, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 15 June 2010
Received in revised form 15 October 2010
Accepted 26 October 2010
Available online 4 December 2010
Borazine is a promising precursor for boron nitride. A detailed investigation on the reaction of sodium
borohydride and ammonium sulfate from 40 °C to 120 °C for synthesis of borazine was performed. The
reaction was monitored by means of 11B nuclear magnetic resonance (11B NMR) and Fourier transform
infrared spectroscopy (FTIR), X-ray diffraction (XRD), mass spectroscopy (MS). The reaction produces
mainly ammonia borane (AB), but not borazine at temperatures below 60 °C. Increases of temperature
promote yield of borazine, which reaches the maximum around 110 °C. Whereas further increased tem-
perature causes severe polymerization of borazine, and hence holds back yields of borazine.
Ó 2010 Elsevier B.V. All rights reserved.
Keywords:
Synthesis
Borazine
Ammonia borane
Boron nitride
Precursor
1. Introduction
ing from inexpensive materials and using standard laboratory
equipments. However, they didn’t study the effects of conditions
Boron nitride has a lot of important applications as high tem-
perature structural and functional materials, due to its unique
properties such as high thermal stability, excellent chemical,
thermal-shock, corrosion and oxidation resistances, low dielectric
constant and loss, high electrical resistivity and thermal conductiv-
ity [1]. However, it is not feasible to be prepared in forms of fibers,
coatings, and complicatedly shaped bodies by conventional high
temperature sintering process. Precursor derived ceramic routes
provide an approach to this issue [2–4]. Borazine (B3H3N3H3) is
reported to be a promising precursor for boron nitride [5–8].
Borazine was originally prepared in 1926 by pyrolysis of the
diammoniate of diborane. Subsequently it was observed as product
in several other reactions [9]. However, all those reactions involved
the use of diborane and gave borazine in small quantity. Later a
procedure suitable for laboratory preparation of borazine was re-
ported, which involved the initial preparation of trichloroborazine
(TCB) and its subsequent reduction by metal borohydrides [10].
Yet, this approach has some drawbacks such as difficult purifica-
tion and handling of diborane. In 1979 a method for synthesizing
borazine by thermal decomposition of ammonia borane in glymes
was described [11]. However, ammonia borane is expensive and
can be purchased only in small quantity. Thomas Wideman et al.
reported a procedure for the laboratory preparation of borazine
by the reaction of (NH4)2SO4 and (NaBH4) in tetraglyme at 120–
140 °C in 1995 [12]. This approach was a one-step procedure, start-
on the reaction.
Investigation on the reaction of (NH4)2SO4 and (NaBH4) will
help understanding of the formation of borazine, and hence
improve yields of borazine. However, there are no reported papers
to date about study on the reaction of (NH4)2SO4 and (NaBH4) for
synthesis of borazine. Herein, it makes sense to study the reaction
under various temperatures.
2. Experimental procedure
2.1. Raw materials
Sodium borohydride, ammonium sulfate and triglyme, all of
which were reagent grade, were purchased from commercial re-
source. Sodium borohydride was used as received. Ammonium sul-
fate was dried under vacuum at 120 °C for 12 h and grinded to pass
200 mesh sieve before use. Triglyme was vacuum distilled from
molten sodium shortly before use.
2.2. Preparation of borazine
All manipulations were carried out under an argon atmosphere
using standard Schlenk techniques. It is noteworthy that a signifi-
cant amount of hydrogen was generated during the reaction, which
could be a hidden danger of fire hazard. It is strongly recom-
mended to exhaust the reaction gases directly to a well-ventilated
hood.
⇑
Corresponding author. Tel./fax: +86 731 84576433.
E-mail address: charlesljs@163.com (J.-s. Li).
0020-1693/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2010.10.030

174
J.-s. Li et al. / Inorganica Chimica Acta 366 (2011) 173–176
Grinded ammonium sulfate (150 g, 1.14 mol) and sodium boro-
hydride (60 g, 1.59 mol) were added to a 5 L four-neck round-bot-
tom flask. The flask was fitted with a thermometer, a reflux
condenser, a rubber stopper and an additional funnel, which was
filled with 500 mL triglyme. The rubber stopper was used to sam-
ple the reaction mixture by a syringe for 11B NMR and FTIR analysis
during the reaction. The exit of the condenser was connected
through a liquid-nitrogen trap series to vacuum pump. The system
was vacuumed and flushed with nitrogen gas several times. The
solvent triglyme was added to the flask, and the contents were stir-
red by a magnetic stir bar. The reaction system was warmed to
specified temperatures in 30 min and held at those temperatures
until the cease of gases release. After the completion of reaction,
the reaction mixture was distilled several times for complete sep-
aration of borazine, under high vacuum using a vacuum-jacked
Vigreux column. Borazine and other volatile condensable
compounds were collected in the liquid-nitrogen traps. The
liquid-nitrogen trap was connected to a À45 (for condensation of
tetraglyme), À78 (for condensation of borazine), and À196 °C (for
condensation of other more volatile compounds such as diborane)
trap-series for fractionation under vacuum. There was a nonvola-
tile white solid left in the original liquid-trap after vacuum distilla-
tion. Purified borazine was condensed in the À78 °C trap.
(b)
(a)
30
20
10
0
-10
-20
-30
Chemical Shift (ppm)
Fig. 1. 11B NMR spectra of the reaction mixture (a) at 40 °C, and (b) at 60 °C. It
indicates that the reaction formed mainly ammonia borane at relatively low
temperature.
H3BNH3
2.3. Characterization
The 11B NMR Spectra were obtained at 128 MHz using a Bruker
Avance 400 spectrometer, referenced to BF3ÁEt2O (0 ppm) as exter-
nal standard. The FTIR spectra were recorded by a Thermo Nicolet
Avatar 360 FTIR spectrometer. X-ray diffraction measurements
were carried out at a wavelength of 1.5418 Å (Cu Ka) with a Bruker
ADVANCED D8 diffractometer. MS measurements were performed
using a Agilent 5975 MS spectrometer.
3. Results and discussion
3.1. Reactions at temperatures below 60 °C
The yields of borazine at various temperature were summarized
in Table 1, which showed that the temperature has a significant
influence on the yields of borazine. No borazine was collected at
temperatures below 60 °C. Fig. 1 shows the 11B NMR spectra of
the reaction mixture at 40 °C and 60 °C, respectively after comple-
tion of the reaction. Both of the spectra show a strong signal
(d = À25.2 ppm, quartet, JBH = 94 Hz) of ammonia borane [13].
There appears no signal of borazine in the spectrum of the reaction
at 40 °C, and only a weak resonance of borazine (d = 28.5 ppm) for
that at 60 °C [14]. The spectrum of the reaction at 60 °C also shows
a weak signal around À11 ppm, which is possible due to cyclotri-
borazane [BH2NH2]3 [15]. The reaction mixture was filtered after
completion of the reaction at 40 °C, and the insoluble residue
was washed with triglyme. A white powder was obtained as the
residue after distillation of the filtrate. The white powder was con-
firmed as ammonia borane by XRD analysis (Fig. 2), which is also
consistent with the 11B NMR spectroscopic results [16].
10
20
30
40
50
60
2θ (degree)
Fig. 2. XRD pattern of a white powder obtained as the residue by evaporation of the
filtrate of the reaction mixture at 40 °C. It shows that the reaction at 40 °C produced
ammonia borane.
nia borane by reacting NaBH4 and (NH4)2SO4 in tetrahydrofuran or
dioxane at room temperature [17].
3.2. Reactions at temperatures between 80 and 110 °C
When the reaction of NaBH4 and (NH4)2SO4 was performed
above 80 °C, borazine was apparently observed in the 11B NMR
spectra. Yields of borazine increases with the increase of tempera-
ture, and reaches maximum at 110 °C. The process of the reaction
at 110 °C was carefully monitored by means of 11B NMR spectros-
copy (Fig. 3).
The quintet at À45.2 ppm (J = 82 Hz) suggests the presence of
NaBH4. As the reaction proceeded, the peaks of NaBH4 decreased,
and finally disappeared, indicating complete consuming of NaBH4.
A strong doublet at 28.2 ppm (JBH = 136 Hz) of borazine was ob-
served. Besides, there appeared some other weak resonances at
Herein it is concluded that the reaction of NaBH4 and (NH4)2SO4
at 40 °C formed mainly ammonia borane rather than borazine. This
is in agreement with the previous reports on preparation of ammo-
Table 1
Yields of borazine at various temperatures.
À28.9 ppm, À24.8 ppm, À13.5 ppm resulting from
l-aminodibora-
Reaction temperature (°C)
Yield (%)
40
À
60
À
80
14
100
27
110
34
120
18
ne (( -NH2)B2H5), ammonia borane and cyclotriborazane ([H2B–
l
NH2]3), respectively. Previous research also reported the appear-
ance of cyclotriborazane as an intermediate in the thermal decom-
‘‘À’’ Means no borazine was isolated.

J.-s. Li et al. / Inorganica Chimica Acta 366 (2011) 173–176
175
indicated by 11B NMR suggests that the reaction at 110 °C is a com-
plex system.
Moreover, a nonvolatile white powder was collected in the li-
quid-nitrogen trap after evaporation of borazine and some other
volatile products. XRD analysis of the nonvolatile white powder
showed it was amorphous. Elemental analysis showed it has a for-
mula of [BH2NH2]x. It was reported that the thermal decomposition
of ammonia borane produced a reactive intermediate monomeric
aminoborane H2B–NH2, which was only stable and detected at
temperatures below À196 °C and converted to polymeric amin-
oborane [BH2NH2]x on warming to room temperature [19,20].
Hence the nonvolatile white powder in the liquid-nitrogen trap
was possibly formed from monomeric aminoborane H2B–NH2,
which was produced and condensed into the liquid-nitrogen traps
at À196 °C. However, since monomeric aminoborane H2B–NH2 is
probably formed in small quantity and is highly reactive, it is not
observed in 11B NMR and MS analysis.
(a)
(b)
(c)
11B NMR analysis also shows the signal of borazine is broadened
compared to that for purified borazine, indicating that some poly-
merization of borazine occurred. However, the polymerization oc-
40
30
20
10
0
-10
-20
-30
-40
-50
curred only in a slight degree suggested from the shapes of the 11
B
Chemical Shift (ppm)
NMR peaks [14]. H2 (2.95 equivalent) per mole of NaBH4 was re-
leased overall during the reaction, which further confirmed the
low degree of polymerization of borazine.
Fig. 3. 11B NMR spectra of the reaction mixture at 110 °C. (a) After 3 h, (b) after 8 h,
and (c) purified borazine. The spectra show strong signals of borazine (d = 28.2 ppm,
doublet, JBH = 136 Hz), as well as weak signals of ammonia borane (H3BNH3,
À24.8 ppm), cyclotriborazane ([BH2NH2]3, À13.5 ppm),
l-aminodiborane ((l-
NH2)B2H5, À28.9 ppm) and NaBH4 (À45.2 ppm).
3.3. Reactions at temperatures up to 120 °C
Yields of borazine decreased sharply when the temperature of
the reaction was elevated to as high as 120 °C (Table 1). The reac-
tion mixture became more and more viscous as the reaction pro-
ceeded, so that the stirring became very difficult. This was the
result of polymerization of borazine. Increased viscosity of the
reaction mixture also restrained the effective contact between
(NH4)2SO4 powder and the NaBH4 solution. Accordingly, the reac-
tion rate was reduced, which further promoted the polymerization
of borazine in turn. Fig. 5 shows the 11B NMR spectrum of the reac-
tion mixture at 120 °C. after completion of the reaction There is a
very broad featureless resonance around 30 ppm, which confirms
the polymerization of the borazine [21]. The reaction mixture
was filtered after reaction. The filtrate was distilled under vacuum
at room temperature, which gave a transparent viscous glue as the
residue. Elemental analysis of the transparent viscous glue gave a
position of ammonia borane in ether solvents [18]. In addition, it
was reported that the thermal decomposition of cyclotriborazane
in glymes caused dehydrogenation leading to the formation of
borazine as the main product [14].
Hence, the reaction of NaBH4 and (NH4)2SO4 mostly proceeds as
follows: the reaction firstly forms ammonia borane, which subse-
quently undergoes thermal dehydrogenation to form cyclotribo-
razane. Cyclotriborazane further dehydrogenates to form
borazine. However alternative reaction sequences are not ruled
out.
MS analysis was carried out for the compounds collected in the
liquid-trap when the liquid trap was warmed to room temperature
(Fig. 4). It shows the occurrence of diborane and borazine [19]. The
occurrence of diborane indicated by MS and
l-aminodiborane
60
50
40
30
20
10
0
10
20
30
40
50
m/z
60
70
80
90
Chemical Shift (ppm)
Fig. 4. Mass spectrum of the compounds collected in the liquid-nitrogen trap. It
shows the occurrence of diborane and borazine.
Fig. 5. 11B NMR spectrum of the reaction mixture at 120 °C. It indicates that severe
polymerization of borazine occurred at temperature as high as 120 °C.

176
J.-s. Li et al. / Inorganica Chimica Acta 366 (2011) 173–176
4. Conclusions
In summary, the reaction of NaBH4 and (NH4)2SO4 in triglyme
(b)
(a)
developed very differently depending on the temperature. The
reaction produced mainly ammonia borane (AB), but not borazine
at temperatures below 60 °C. Yields of borazine increased with the
increase of temperature, and reached the maximum (34%) around
110 °C. However, further increase of temperature (up to 120 °C)
caused severe polymerization of borazine, and hence decreased
the yield of borazine. Hydrogen, ammonia borane and cyclotribo-
razane were found as the main intermediates, together with trace
amounts of l-aminodiborane and diborane.
Acknowledgment
We are grateful for the support by Chinese National Natural Sci-
ence Foundation (Nos. 50902150 and 90916019).
References
[1] J. Eichler, C. Lesniak, J. Eur. Ceram. Soc. 28 (2008) 1105.
[2] P. Colombo, J. Eur. Ceram. Soc. 28 (2008) 1389.
[3] S.G. Lee, J. Fourcade, R. Latta, A.A. Solomon, Fusion Eng. Des. 83 (2008) 713.
[4] V.M. Gudapati, V.P. Veedu, M.N. Ghasemi-Nejhad, Compos. Sci. Technol. 66
(2006) 3230.
[5] S. Seghi, J. Lee, J. Economy, Carbon 43 (2005) 2035.
[6] S. Seghi, B. Fabio, J. Economy, Carbon 42 (2004) 3043.
[7] B. Li, C. Zhang, F. Cao, S. Wang, Y. Cao, J. Feng, B. Chen, Adv. Appl. Ceram. 107
(2008) 1.
[8] B. Li, C. Zhang, F. Cao, S. Wang, Y. Jiang, J. Inorg. Mater. 23 (2008) 229.
[9] (a) A. Stock, E. Wiberg, H. Martini, Ber. Dtsch. Chem. Ges. 63 (1930) 2927;
(b) A. Stock, E. Pohland, Ber. Dtsch. Chem. Ges. 59 (1926) 2215;
(c) A. Stock, E. Pohlan, Ber. Dtsch. Chem. Ges. 62 (1929) 90.
[10] R. Schaeffer, M. Steindler, L. Hohnstedt, H.S. Smith, L.B. Eddy, H.I. Schlesinger, J.
Am. Chem. Soc. 76 (1954) 3303.
4000
3500
3000
2500
2000
1500
1000
500
Wavenumber (cm-1)
Fig. 6. FTIR spectra of (a) the transparent viscous glue isolated from the reaction
mixture at 120 °C, and (b) purified borazine. The characteristic broad band between
1400–1600 cmÀ1 indicates that borazine underwent polymerization partially at
temperature as high as 120 °C.
[11] W.V. Hough, C.R. Guibert, G.T. Hefferan, US Patent 415 0097, 1979.
[12] T. Wideman, L.G. Sneddon, Inorg. Chem. 34 (1995) 1002.
[13] R.P. Shrestha, H.V.K. Diyabalanage, T.A. Semelsberger, K.C. Ott, A.K. Burrell, Int.
J. Hydrogen Energy 34 (2009) 2616.
[14] R. Schellenberg, J. Kriehme, G. Wolf, Thermochim. Acta 457 (2007) 103.
[15] C.A. Jaska, K. Temple, A.J. Lough, I. Manners, Chem. Commun. (2001) 962.
[16] E.W. Hughes, J. Am. Chem. Soc. 78 (1956) 502.
[17] P.V. Ramachandran, P.D. Gagare, Inorg. Chem. 46 (2007) 7810.
[18] J.S. Wang, R.A. Geanangel, Inorg. Chim. Acta 148 (1988) 185.
[19] F. Baitalow, J. Baumann, G. Wolf, K. Jaenicke-Röbler, G. Leitner, Thermochim.
Acta 391 (2002) 159.
[20] R. Komm, R.A. Geanangel, R. Liepins, Inorg. Chem. 22 (1983) 1684.
[21] P.J. Fazen, E.E. Remsen, J.S. Beck, P.J. Carroll, A.R. McGhie, L.G. Sneddon, Chem.
Mater. 7 (1995) 1942.
formula of [B3.0N2.9H4.2]x. FTIR spectrum of the transparent viscous
glue was shown in Fig. 6, together with that of purified borazine.
The absorption peaks at 3490 cmÀ1 and 915 cmÀ1 are attributed
to the stretching and bending of N–H, respectively. The absorption
peaks at 2525 cmÀ1 and 718 cmÀ1 are attributed the stretching and
bending of B–H, respectively. The absorption peaks at 1464 cmÀ1
are attributed to the deformation of [B–N]3 ring. The absorption
peak around 1400 cmÀ1 of the transparent viscous glue is appar-
ently broader than that of purified borazine, which is due to the
polymerization of borazine. It is reasonable to consider the trans-
parent viscous glue to be polyborazylene [B3.0N2.9H4.2 x
] formed
[22] T. Wideman, P.J. Fazen, K. Su, E.E. Remsen, G.A. Zank, L.G. Sneddon, Appl.
Organomet. Chem. 12 (1998) 681.
from polymerization of borazine [22].