Synthesis and structural studies of tris-2-chlorobenzylamine and tris-2-bromobenzylamine

Article abstract of DOI:10.1007/s10870-005-2954-5

The tris-2-chloro and 2-bromotribenzylamines are prepared from aqueous ammonia and 2-chlorobenzyl chloride and 2-bromobenzyl bromide, respectively, in ethanol. Recrystallization yielded colorless cubes of each product. The crystal structures are each solved in space group P1, and are isostructural. The tris-2-chloro compound, 1, has a = 7.4226(5) A, b = 9.0825(7) A, c = 14.529(1) A, α = 78.279(1)°, β = 82.389(1)°, γ = 84.661(1)°, and V = 948.41(12) A³ with Z = 2, and d _calc = 1.368 Mg/m³. The tris-2-bromo analog, 2, has a = 7.6569(11) A, b = 9.0922(13) A, c = 14.614(2) A, α = 79.286(2)°, β = 81.777(2)°, γ = 85.401(2)°, and V = 987.9(2) A³ with Z = 2, and d_calc = 1.762 Mg/m ³. Lithium-halogen exchange experiments conducted in tetrahydrofuran at -78°C using n-butyl lithium revealed that no exchange occurred for the tris-2-chloro compound, but did occur for the tris-2-bromo analog to yield tribenzylamine upon quench and work-up.

Full text of DOI:10.1007/s10870-005-2954-5

Journal of Chemical Crystallography, Vol. 35, No. 3, March 2005 (ꢀC 2005)
DOI: 10.1007/s10870-005-2954-5
Synthesis and structural studies of
tris-2-chlorobenzylamine and tris-2-bromobenzylamine
Qing Chen,(1) Carrie E. Buss,(2) Victor G. Young, Jr.,(2) and Stephen Fox(1)∗
Received June 12, 2004; accepted October 17, 2004
The tris-2-chloro and 2-bromotribenzylamines are prepared from aqueous ammonia and
2-chlorobenzyl chloride and 2-bromobenzyl bromide, respectively, in ethanol. Recrystal-
lization yielded colorless cubes of each product. The crystal structures are each solved in
¯
˚
space group P1, and are isostructural. The tris-2-chloro compound, 1, has a = 7.4226(5) A,
◦
◦
◦
˚
˚
b = 9.0825(7) A, c = 14.529(1) A, α = 78.279(1) , β = 82.389(1) , γ = 84.661(1) , and
V = 948.41(12) A with Z = 2, and dcalc = 1.368 Mg/m3. The tris-2-bromo analog, 2, has
3
˚
◦
◦
˚
˚
˚
a = 7.6569(11) A, b = 9.0922(13) A, c = 14.614(2) A, α = 79.286(2) , β = 81.777(2) ,
γ = 85.401(2) , and V = 987.9(2) A with Z = 2, and dcalc = 1.762 Mg/m3. Lithium–
halogen exchange experiments conducted in tetrahydrofuran at −78◦C using n-butyl lithium
revealed that no exchange occurred for the tris-2-chloro compound, but did occur for the
tris-2-bromo analog to yield tribenzylamine upon quench and work-up.
◦
3
˚
KEY WORDS: Tribenzylamine; triclinic; 2-chloro; 2-bromo; lithium–halogen exchange.
Introduction
Tetradentate
shown tremendous scope in the synthesis of
transition-metal complexes, especially in the
bioinorganic genre. Perhaps the most extensively
studied examples are those comprising tris-(2-
pyridylmethyl)amine,1 (referred to herein as
tpa, Fig. 1), which has been widely exploited in
complexes of copper2 and iron.3
from in situ lithium–halogen exchange of tris-2-
chlorobenzylamine and its 2-bromo counterpart.
tripodal
ligands
have
The synthesis of the latter was first reported over
100 years ago,4 and recently found utility in the
preparation of tripodal benzenethiols.5 The for-
mer, perhaps surprisingly, has no record in the
literature. The ensuing report gives high-yielding
syntheses of tris-2-chlorobenzylamine and tris-2-
bromobenzylamine, the crystal structures of both
compounds, routine characterizations, and the re-
sults of lithium–halogen exchange endeavors.
We are interested in exploring the organ-
ometallic chemistry of complexes of the tris-
phenylide analog (Fig. 1) of tpa; we rationalized
that this trianionic ligand should be accessible
Experimental
(1) Department of Chemistry, University of Louisiana at Monroe,
Monroe, Louisiana, 71209-0530.
Materials and methods
(2) Department of Chemistry, University of Minnesota, Minneapolis,
Minnesota, 55455.
The reagents 2-chlorobenzylamine (95%), 2-
chlorobenzyl chloride (98%), and 2-bromobenzyl
bromide (98%) were purchased from the Aldrich
∗ To whom correspondence should be addressed; e-mail: sfox@
ulm.edu.
177
ꢀ
C
1074-1542/05/0300-0177/0 2005 Springer Science+Business Media, Inc.

178
Chen, Buss, Young, and Fox
7.65 (d, 1H, phenyl-H, J = 7.5 Hz). Mass spec-
trometry: (70 eV, electron impact) m/z 393, 391,
389, 280, 278, 264, 127, 125. Anal. Calcd for
C21H18NCl3: C, 64.55; H, 4.64; N, 3.58. Found:
C, 64.28; H, 4.76; N, 3.64.
Fig. 1. Representations of tpa and its tris-phenylide
counterpart.
Synthesis of tris-2-bromobenzylamine (2)
Chemical Company. The triethylamine and aque-
ous ammonia were purchased from the Fisher
Company. All solvents were reagent grade and
were used without further purification. Melting
points were recorded using an electrothermal
EM-6 apparatus, and were not corrected. Nu-
clear magnetic resonance spectra were recorded
using a Jeol Eclipse 400 MHz spectrometer. Mass
spectrometry measurements were recorded on a
Finnigan Mat GCQ instrument. Elemental analy-
ses were determined using an Exeter Analytical
CE440 SHA Elemental Analyzer.
The combination of 2-bromobenzyl bromide
(5.29 g, 21.0 mmol) with 29% aqueous ammo-
nia (1.76 g, 30.0 mmol) in ethanol at room tem-
perature produced a white solid during stirring
over 24 h. The precipitate was collected by filtra-
tion, washed with ethanol, and allowed to air-dry
to yield 3.55 g (97%) of 2 as white microcrys-
tals. Recrystallization was effected from hot ace-
tonitrile to yield colorless blocks. Melting point:
122.0–122.5◦C. 1H-NMR (CDCl3, δ): 3.79 (s, 2H,
N CH2), 7.08 (t, 1H, phenyl-H, J = 7.7 Hz),
7.27 (t, 1H, phenyl-H, J = 7.1 Hz), 7.51 (d, 1H,
phenyl-H, J = 8.1 Hz), 7.65 (d, 1H, phenyl-H,
J = 7.2 Hz). Mass spectrometry: (70 eV, elec-
tron impact) m/z 527, 525, 523, 521, 370, 368,
366, 356, 354, 352, 171, 169. Anal. Calcd for
C21H18NBr3: C, 48.13; H, 3.46; N, 2.67. Found:
C, 48.06; H, 3.43; N, 2.72.
Synthesis of tris-2-chlorobenzylamine (1)
The combination of 2-chlorobenzylamine
(1.49 g, 10.0 mmol) with 2-chlorobenzyl chloride
(3.29 g, 20.0 mmol) and triethylamine (2.02 g,
20 mmol) in acetonitrile (20 mL) in a 100 mL
round-bottom flask yielded a colorless solution.
After heating and stirring at reflux for 36 h, the
clear solution was allowed to cool to deposit
white crystals; these were collected by filtration,
washed with acetonitrile, and allowed to air-dry to
yield2.70g(69%)oftris-2-chlorobenzylamine, 1.
The same product was also obtained from the
combination of 2-chlorobenzyl chloride (4.83 g,
30.0 mmol) with 29% aqueous ammonia (2.64 g,
45.0 mmol) in ethanol at room temperature. A
white solid precipitated during stirring over 24 h.
The precipitate was collected by filtration, washed
with ethanol, and allowed to air-dry to yield 1.67 g
(43%) of white microcrystals. Recrystallization
was effected from hot acetonitrile to yield col-
orless blocks. Melting point: 89–91◦C. 1H-NMR
(CDCl3, δ): 3.82 (s, 2H, N CH2), 7.15 (t, 1H,
phenyl-H, J = 7.5 Hz), 7.22 (t, 1H, phenyl-H,
J = 7.5 Hz), 7.31 (d, 1H, phenyl-H, J = 8.0 Hz),
Reactivity with n-butyl lithium
In a 50 mL round-bottom flask equipped
with magnetic stirrer, 1.00 mmol of either 1 or
2 was dissolved in freshly distilled tetrahydrofu-
ran (5.0 mL). Under nitrogen flow, the colorless
solution was cooled to −78◦C while stirring. Af-
ter a few minutes, 1.2 mL of 2.5 M n-butyllithium
in hexane (3.0 mmol) was injected by syringe and
the solution turned peach-pink. After stirring for
10 min, the reaction was quenched by addition of
water (0.20 mL, 11 mmol) to yield a colorless so-
lutionandwhiteprecipitate. Afterafewmoremin-
utes, the solution was filtered and the filtrate al-
lowed to evaporate. The residual white solid from
the reaction of 1 was recrystallized from acetoni-
trile to yield white crystals of starting material
1; the residual white solid from the reaction of 2
was recrystallized from acetonitrile to yield white

tris-2-chlorobenzylamine and tris-2-bromobenzylamine
179
Table 1. Crystal Data and Structure Refinement for tris-2-Chlorobenzylamine, 1, and tris-2-Bromobenzylamine, 2
1
2
Empirical formula
CCDC deposit number
Formula weight
C21H18N Cl3
CCDC-241081
390.71
C21H18N Br3
CCDC-241082
524.09
Temperature (K)
173(2)
173(2)
˚
Wavelength (A)
0.71073
0.71073
Crystal system
Space group
Triclinic
P1
Triclinic
P1
¯
¯
Unit cell dimensions
˚
a (A)
7.4226(5)
7.6569(11)
˚
b (A)
9.0825(7)
9.0922(13)
˚
c (A)
14.529(1)
78.279(1)
82.389(1)
84.661(1)
14.614(2)
79.286(2)
81.777(2)
85.401(2)
α (◦)
β (◦)
γ (◦)
3
˚
Volume (A )
948.41(12)
987.9(2)
Z
2
2
Density (calculated) (Mg/m3)
Absorption coefficient (mm−1
F(000)
1.368
0.486
404
1.762
6.129
512
)
Crystal color, morphology
Crystal size (mm)
Cube, Colorless
0.25 × 0.25 × 0.21
1.44 to 25.10
−8 ≤ h ≤ 8
−10 ≤ k ≤ 10
0 ≤ l ≤ 17
Cube, Colorless
0.25 × 0.25 × 0.25
2.28 to 27.51
−9 ≤ h ≤ 9
−11 ≤ k ≤ 11
0 ≤ l ≤ 18
Theta range for data collection (◦)
Index ranges
Reflections collected
4750
10457
Independent reflections, R (int)
Observed reflections
3236, 0.0234
2579
95.5%
4442, 0.0364
3535
98.3%
Completeness to theta = 25.10◦, 27.51◦
Absorption correction
Multi-scan
Multi-scan
Max. and min. transmission
Refinement method
Data/restraints/parameters
Goodness-of-fit on F2
0.9048 and 0.8881
Full-matrix least-squares on F2
3236/0/226
1.024
0.3095 and 0.3095
Full-matrix least-squares on F2
4442/0/226
1.040
Final R indices [I > 2sigma (I)]
R indices (all data)
R1 = 0.0438, wR2 = 0.1146
R1 = 0.0576, wR2 = 0.1221
0.268 and −0.340
R1 = 0.0347, wR2 = 0.0820
R1 = 0.0496, wR2 = 0.0883
0.646 and −0.759
−3
˚
Largest diff. peak and hole (e.A
)
crystals of tribenzylamine. The products of the
lithium–halogen exchange experiments were de-
termined based on comparative melting point and
NMR measurements.
SMART CCD area detector diffractometer; both
datasets were collected at 173 K, using Mo Kα ra-
diation (graphite monochromator). Intensity data
were corrected for absorption and decay using
SADABS6 for both 1 and 2. The structures were
solved using direct methods and refined using
Bruker SHELXTL7 for 1 and SHELXL V6.128
for 2. Hydrogen atoms were placed in ideal po-
sitions and treated with a riding model. Table 1
contains complete details of data collection and
refinement for both compounds.
X-ray data collection, structure
solution and refinement
Data collection for 1 was performed on a
Siemens Smart Platform CCD diffractometer, and
data collection for 2 was performed on a Bruker

180
Chen, Buss, Young, and Fox
Results and discussion
Compounds 1 and 2 were synthesized from
2-chlorobenzyl chloride and 2-bromobenzyl bro-
mide, respectively, in ammoniacal ethanol solu-
tion; the substantially different yields (43%, 1, vs.
97%, 2) under these conditions perhaps reflect the
greater susceptibility of 2-bromobenzyl bromide
to nucleophilic attack, as well as the lesser sol-
ubility of 2 versus 1. A better yield (69%) for
1 was achieved in acetonitrile solution from 2-
chlorobenzyl chloride and 2-chlorobenzylamine.
The recrystallized compounds gave elemental
analyses for C, H, N, consistent with high purity
of the bulk materials. Furthermore, the GC–MS
data is consistent with the predominant formation
of M-C6H5X, M-C6H5XCH2, and C6H5XCH2
cations (X = Cl, Br) during fragmentation of 1
and 2.
The molecular compounds melted quite
sharply; as might be expected, 2 melts at a higher
temperature (122–123◦C) than 1, consistent with
enhanced intermolecular forces resulting from the
larger, more polarizable bromine vs. chlorine. Cu-
riously, 1 melts at a slightly lower temperature
(89–91◦C) than the unsubstituted tribenzylamine
(91–93◦C). In the absence of hydrogen-bonds,
normal dispersion forces would dictate that the
more massive, substituted analog would melt at a
significantly higher temperature, of course.
Not unexpectedly, the 1H-NMR spectra of 1
and 2 are very similar in profile and chemical shift.
The aromatic region for each is 7.00–7.70 δ, and
a profile of two sets of doublets (at lower field)
and two sets of triplets can be grossly assigned to
the four unique protons of each phenyl ring; the
magnetic equivalence of each of the three ben-
zylic groups per compound is supported by the
observation of a sharp singlet for the methylene
protons which also concurs with an integration of
2:1 relative to each of the phenyl protons.
Fig. 2. Molecularstructureshowingatom-numberingscheme
of tris-2-chlorobenzylamine, 1, at 50% probability.
finement parameters appears in Table 1. On go-
ing from 1 to 2, the slight unit cell volume in-
crease and calculated density difference naturally
reflect the substitution of the three chlorine atoms
by bromines; fractional atomic coordinates, of
course, show very little difference for 1 vs. 2. The
atom-numbering scheme for 1 and 2 is given in
Figs. 2 and 3, and important bond distances and
angles are found in Table 2.
Inspection of the pertinent torsion angles re-
veals that each phenyl ring is an essentially pla-
nar unit, as would be expected; the rings are
canted off perpendicular, however (as shown by
the N1 C1 C2 C7 torsion angles of 132.5(2)◦
for 1 and 130.9(3)◦ for 2, for example) to impart
a propeller-shaped molecular geometry grossly
The structure of each compound is solved
¯
in the space group P1—the space group being
determined from systematic absences and inten-
sity statistics. The compounds are isostructural;
a complete list of crystal data and structure re-
Fig. 3. Molecularstructureshowingatom-numberingscheme
of tris-2-bromobenzylamine, 2, at 50% probability.

tris-2-chlorobenzylamine and tris-2-bromobenzylamine
181
◦
a.9 Indeed, structure comparison reveals very lit-
tle difference between 1 and 2: analogous bonds
and angles are essentially identical within calcu-
lated error limits, with the natural exception of
the C Cl vs. C Br bonds. The difference here
˚
Table 2. Selected Bond Lengths (A) and Angles ( ) for 1 and 2
1
2
N1 C1
N1 C15
N1 C8
C1 C2
C8 C9
C15 C16
C7 Cl1
C14 Cl2
C21 Cl3
1.465(3) N1 C1
1.467(3) N1 C15
1.462(3) N1 C8
1.505(3) C1 C2
1.510(3) C8 C9
1.466(3)
1.471(4)
1.471(4)
1.508(4)
1.512(4)
1.512(4)
1.908(3)
1.903(3)
1.907(3)
˚
is ∼0.16 A, on average, which is consistent with
the significant covalent radius difference between
1.510(3) C15 C16
1.749(2) C7 Br1
11
˚
˚
chlorine (0.97 A) vs. bromine (1.14 A).
1.743(2) C14 Br2
1.742(2) C21 Br3
110.92(17) C1 N1 C15
110.96(18) C1 N1 C8
111.12(17) C15 N1 C8
110.87(18) N1 C1 C2
110.54(18) N1 C8 C9
The combinations of 1 and 2 with n-
butyllithium at −78◦C revealed that the tris-
chloro gave back itself, whereas the tris-bromo
compound yielded tribenzylamine upon work-up.
These observations are consistent with lithium–
halogen exchange occurring only for the tris-
bromo compound 2 under the reaction condi-
tions used, probably a consequence of the longer
carbon–bromine bonds rendering the bromine
atoms more accessible to attack by the lithium
reagent than their smaller chlorine counterparts
in 1. This result implicates the in situ formation
of the desired tris-phenylide trianion at −78◦C
from the reaction of 2, and bodes well for the
eventual fulfillment of our objective to investigate
organometallic complex analogs of tpa (Fig. 1).
C1 N1 C15
C1 N1 C8
C15 N1 C8
N1 C1 C2
N1 C8 C9
111.2(2)
111.1(2)
110.8(2)
110.6(2)
110.4(2)
110.9(2)
N1 C15 C16 111.52(17) N1 C15 C16
Note. Symmetry transformations used to generate equivalent
atoms: #1 −x + 1, −y + 1, −z + 1.
consistentwiththepointgroupC3. ThemeanC C
distance within the phenyl rings is 1.383(4) A
for 1, and 1.386(4) A for 2, compared with
1.374(19) A for tribenzylamine ; these slight dif-
ferences are nevertheless consistent with the in-
creased steric demands of Br vs. Cl vs. H. The
mean sp –sp C C distances are 1.508(3) A for 1,
˚
˚
9
˚
3
2
˚
˚
˚
and 1.510(4) A for 2, compared with 1.525(17) A
Acknowledgments
for tribenzylmine,9 and the accepted value of
10
˚
1.501 A.
S.F. wishes to thank Dr. H. O’Neil Brother-
ton of the University of Louisiana at Monroe for
performing elemental analysis and GC-MS deter-
minations.
The angles at the pyramidal nitrogen,
C N C, give average values of 111.0(2)◦ for 1,
111.0(2)◦ for 2, and 110.0(9)◦ for tribenzylamine;
again, the slightly larger angles for 1 and 2 might
reasonably be considered to result from the greater
steric demands of Br and Cl vs. H to slightly flatten
the tetrahedron visualized from C1, C8, C15, and
N. By the same token, the average N C C angle
diminishes slightly for 2 (110.6(2)◦) relative to 1
(111.0(2)◦), and for tribenzylamine (111.5(11)◦),
is the largest of all.
References
1. Anderegg, G.; Wenk, F. Helv. Chim. Acta 1967, 50, 2330.
2. Fox, S.; Nanthakumar, A.; Wikstro¨m, M.; Karlin, K.D.;
Blackburn, N.J. J. Am. Chem. Soc. 1996, 118, 24.
3. Me´nage, S.; Zang, Y.; Hendrich, M.P.; Que, L., Jr. J. Am. Chem.
Soc. 1992, 114, 7786.
4. Jackson, C.L.; White, J.F. Am. Chem. J. 1880–1881, 2, 315.
5. Koch, S.A.; Welch, M.J.; Cutler, C.S.; Quarless, D.A., Jr. PCT
Int. Appl. 1999, WO 9956791.
6. Blessing, R. Acta Crystallogr. Sect A 1995, 51, 33.
7. SHELXTL V6.10; Bruker Analytical X-Ray Systems: Madison,
WI, 2000.
8. SHELXTL V6.12; Bruker Analytical X-Ray Systems: Madison,
WI.
For 1 and 2, two of the three rings of each
molecule appear to display stronger intermolec-
ular stacking interactions than those of the third.
Threefold crystal symmetry is also precluded in
tribenzylamine (although in space group P21/a),
where stronger stacking is displayed between
phenyl rings along the b and c axes than along
9. Iwasaki, F.; Iwasaki, H. Acta Crystallogr. B 1972, 28, 3370.
10. Stoicheff, B.P. Tetrahedron 1962, 17, 135.
11. Cotton, F.A.; Wilkinson, G.; Gaus, P.L. Basic Inorganic
Chemistry; Wiley: New York, 1994; pp. 60–61.