Orthoamides by selective borohydride reduction of N,N′,N″-triacyl- and N,N′,N″-tri(alkoxycarbonyl)-guanidines

Article abstract of DOI:10.1016/j.tetlet.2018.12.060

N,N′,N″-Triacylguanidines and N,N′,N″-tri(alkoxycarbonyl)guanidines were prepared and reduced with borohydride salts in a mixture of tetrahydrofuran and acetic acid to give triacyl and tri(alkoxycarbonyl) orthoamides in yields of 40–85%. However, similar

Full text of DOI:10.1016/j.tetlet.2018.12.060

Tetrahedron Letters 60 (2019) 427–431
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Orthoamides by selective borohydride reduction of N,N0,N00-triacyl- and
N,N0,N00-tri(alkoxycarbonyl)-guanidines
⇑
Matthew C. Davis , Thomas J. Groshens
Naval Air Warfare Center, Michelson Laboratory, Chemistry Division, China Lake, CA 93555, USA
a r t i c l e i n f o
a b s t r a c t
N,N0,N00-Triacylguanidines and N,N0,N00-tri(alkoxycarbonyl)guanidines were prepared and reduced with
borohydride salts in a mixture of tetrahydrofuran and acetic acid to give triacyl and tri(alkoxycarbonyl)
orthoamides in yields of 40–85%. However, similar reduction of N,N0,N00-tri(t-butoxycarbonyl)guanidine
did not give orthoamide but the aminal di(t-butyl) methylenedicarbamate.
Article history:
Received 16 November 2018
Revised 20 December 2018
Accepted 24 December 2018
Available online 27 December 2018
Ó 2018 Elsevier Ltd. All rights reserved.
Keywords:
Orthoamide
Sodium cyanoborohydride
Reduction
Guanidine
Carbamate
Aminal
Introduction
Orthoformamides (1), or simply orthoamides, are organic com-
pounds with three nitrogen atoms bonded to a single carbon atom,
Chart 1 [1,2]. Hexaalkyl-substituted orthoamides are surprisingly
stable molecules, for example 1,1,1-tris(dimethylamino)methane
has an atmospheric boiling point of 145–146 °C [3]. Orthoamide
can be made by nucleophilic substitution of trialkyl orthoformate
or trihalomethane by lithium amides [4,5]. Peralkylated orthoa-
mides are lacking in functionality for transformation to other use-
ful products, however, such as nitrolysis to the unknown 1,1,1-tris
(nitramino)methane. Bredereck et al. developed a method to syn-
thesize 1,1,1-tris(alkylcarboxamido)- and 1,1,1-tris(alkoxycar-
bonylamino)-methanes (2a) by acid catalyzed condensation of
triethyl orthoformate with carboxamide and carbamate in reflux-
ing toluene [6]. Although this method is certainly rapid and gen-
eral, it would be difficult to directly prepare orthogonally
substituted orthoamides without a complicated separation from
a statistical mixture. Therefore, we decided to try to identify an
alternative synthesis route which might eventually allow the syn-
thesis of orthoamides with two or even three different acyl and/or
alkoxycarbonyl groups on the nitrogen atoms (2b).
Chart 1. Structures of orthoamides and N,N0,N00-triacyl-orthoamides.
made the chemistry of guanidine rich and comprehensively stud-
ied. Guanidine is simply dehydro-orthoamide where a molecule
of hydrogen has been removed leaving an sp2 carbon–nitrogen
bond. The C@N bond in guanidine might be viewed as a form of
imine, the dehydration product of an amine and carboxaldehyde.
Thus, we hypothesized that appropriately substituted guanidines
might be reduced by methods similar to those employed in reduc-
tion of imines [7], effectively adding hydrogen across their carbon–
nitrogen sp2-bond yielding orthoamides, Chart 2. There was some
chemical precedence in the literature that supported the concept.
Hexaalkylguanidinium salts were reduced by Red-Al (NaH2[Al
(OCH2CH2OMe)2]) to the orthoamides 1,1,1-tris(dialkylamino)
methanes [8]. Kim et al. [9] and Alder et al. [10] reduced tricyclic
or bridged guanidinium salts to bridged orthoamides by reduction
with lithium aluminum hydride (LAH) or sodium borohydride
Derivatives of guanidine (HN@C(NH2)2) are widely found in
natural products (e.g. arginine) and pharmaceuticals which has
⇑
Corresponding author.
E-mail address: matthew.davis@navy.mil (M.C. Davis).
https://doi.org/10.1016/j.tetlet.2018.12.060
0040-4039/Ó 2018 Elsevier Ltd. All rights reserved.

428
M.C. Davis, T.J. Groshens / Tetrahedron Letters 60 (2019) 427–431
Chart 2. Retrosynthetic analysis of triacyl- or tri(alkoxycarbonyl)-orthoamides by
reduction of triacyl- or tri(alkoxycarbonyl)-guanidines.
(NaBH4). Stearns and Rapoport showed that certain N-acyl-N0, N0,
N00, N00-tetraalkylguanidines when reduced by LAH gave an acylam-
inal (RCONHCH2NR2) where the hydride reagent had reacted at the
C@N bond preferentially [11].
Results and discussion
Fig. 1. X-ray crystal structure (50% probability thermal ellipsoids) of TAG (3).
N,N0,N00-Triacylated guanidines appeared to be the simplest
derivatives to test the reduction hypothesis although they are
not commercially available. Surprisingly, N,N0,N00-triacetylguani-
dine (TAG, 3) was made as late as 1959 by Greenhalgh and Bannard
and apparently not studied further [12]. Guanidine carbonate was
an effective replacement of guanidine acetate in a reaction with
excess acetic anhydride at 100 °C for one hour, Scheme 1.
dride (NaCNBH3) was used as it has less aggressive character than
NaBH4 [16]. A solution of TAG in tetrahydrofuran and acetic acid
was reacted with one molar equivalent of NaCNBH3 at room tem-
perature, Scheme 2. A precipitate formed from the mixture shortly
after the hydride was added which was anticipated since triacyl
orthoamides are known to have poor solubility in organic solvents.
The precipitate was filtered (57% yield) and its high melting point
(273 °C) was very different from TAG (104 °C) but close to that
reported for 1,1,1-tris(acetamido)methane (6) by Bredereck et al.
(261 °C).[6] The proton nuclear magnetic resonance (NMR) spec-
trum of the product (d6-DMSO) was further conclusive of orthoa-
The desired product was isolated in 50% yield. At higher tem-
perature and/or longer reaction times, the reaction mixture
becomes pale yellow colored indicating the product has undergone
intermolecular cyclization to form 2-methyl-4,6-di(acetamido)-
1,3,5-triazine. Recrystallization of TAG from ethanol gave x-ray
quality crystals (Fig. 1), although longer and more vigorous heating
in this solvent was reported to cause deacetylation [12,13]. The
method was also used to make N,N0,N00-tripropionylguanidine (4),
but high vacuum was required to remove the excess propionic
anhydride and propionic acid after the reaction was complete.
With pivalic anhydride, the temperature had to be raised to
130 °C in order to completely dissolve the guanidine carbonate
salt. The product isolated from the reaction was not N,N0,N00-tripi-
valoylguanidine or a 1,3,5-triazine but N,N0-di(pivaloyl)urea (5)
instead. It’s not clear how this deamination occurred but Baster-
field and Paynter also obtained urea products in their synthesis
of N,N0-di(ethoxycarbonyl)guanidine from guanidine and ethyl
chloroformate in ethanol [14]. These by-products indicate that
the C@N bond of these guanidine derivatives are susceptible to
attack probably by oxygen nucleophiles forming short-lived urea
acetal intermediates with loss of ammonia or ammonium ion.
Borohydride salts were chosen over aluminumhydride salts
since the latter would likely not be selective in reduction of C@N
over C@O, both present in TAG [11]. Imines typically will not
reduce with NaBH4, or only very slowly, without an acid present
to protonate the nitrogen atom forming an electrophilic iminium
species. So it was anticipated that an acid should be present to
form a protonated guanidine species similar to an N-acyliminium
ion [15]. For the first attempt to reduce TAG, sodium cyanoborohy-
mide
6 including the characteristic coupling between the
methine proton (quartet at 6.52 ppm) with the three adjacent
amide protons (doublet at 8.46 ppm) and a singlet for the methyl
groups (1.7 ppm) with an integral ratio of 1:3:9, respectively
[17]. Unfortunately, recrystallization from a variety of solvents
did not yield x-ray quality crystals of this orthoamide, or any other
made in this report.
In the synthesis of the higher tripropionyl (7) and tributyryl (8)
orthoamides, the method was changed slightly without deleterious
effect, Scheme 3. After formation of N,N0,N00-triacylguanidine from
guanidine carbonate was complete, one equivalent of NaCNBH3
was added to the room temperature reaction mixture of N,N0,N00-
triacylguanidine, carboxylic acid and carboxylic acid anhydride.
Again, the orthoamides precipitated from the reaction mixtures
in modest yield (30–40%) and had spectroscopic data and elemen-
tal analysis consistent with orthoamide structure. There was not
sufficient time to attempt the synthesis of 1,1,1-tris(formy-
lamino)methane [18], useful as a building block for nitrogen hete-
rocycles, by this new route. Though it may be possible to
synthesize the unknown N,N0,N00-triformylguanidine from guani-
dine salts and acetic formic mixed anhydride [19].
Next was to expand the scope of the reduction and test N,N0,N00-
tri(alkoxycarbonyl)guanidines but these are also not commercial
products, save for the expensive N,N0,N00-tri(t-butoxycarbonyl)
Scheme 1. Synthesis of triacylguanidines or diacylurea from guanidine carbonate reaction with carboxylic anhydride.

M.C. Davis, T.J. Groshens / Tetrahedron Letters 60 (2019) 427–431
429
these difficulties, TCBzG was also made by the ‘two-step method’
and its improvement by Cesco-Cancian et al. [22,23].
With four guanidine tricarbamates (9–12) in hand, reduction
experiments could be started. In one reaction, TMG was reacted
with two molar equivalents of NaCNBH3 (six equivalents hydride)
in tetrahydrofuran and acetic acid solution at room temperature,
Scheme 5. A precipitate did not form and work up gave a crys-
talline solid which was shown to be dimethyl methylenedicarba-
Scheme 2. Sodium cyanoborohydride reduction of TAG (3) to triacetyl orthoamide
(6).
mate, apparently from reductive deamination of
a carbon–
nitrogen bond. Its proton NMR spectrum in deuterochloroform
had a triplet at 4.5 ppm for the methylene group, a broad singlet
at 5.8 ppm for carbamate hydrogen and a singlet at 3.7 ppm for
methyl of carbamate, with an integral ratio was 2:2:6, respectively
[24,25]. A quick survey of the stability of authentic 1,1,1-tris
(ethoxycarbonylamino)methane to the same conditions was made
[26]. After incubating one hour at room temperature, the mixture
was worked up and gave a residue containing both starting mate-
rial and diethyl methylenedicarbamate according to proton NMR
analysis. The experiment showed the sensitivity of tri(alkoxycar-
bonyl) orthoamide to reduction by this hydride and that excess
reducing reagent should be avoided. Returning to TMG, when the
amount of NaCNBH3 was reduced to 0.5 equivalents (1.5 equiva-
lents hydride), a precipitate formed from the reaction which was
isolated and fully characterized as the desired 1,1,1-tris(methoxy-
carbonylamino)methane (17), made in 47% yield, Scheme 5. The
orthoamides 1,1,1-tris(benzyloxycarbonylamino)methane (18)
[27] (86%) and unreported 1,1,1-tris(phenyloxycarbonylamino)
methane (19) (39%) were both made from their respective guanidi-
nes by analogous procedures to the above.
Scheme 3. One pot synthesis of higher triacyl orthoamide.
guanidine (TBG, 9). Many N,N0,N00-tri(alkoxycarbonyl)guanidines
have been made by Goodman and coworkers in a ‘two-step
method’, Scheme 4 [20,21]. In the first step, which will be referred
to as the ‘aqueous method’, guanidine hydrochloride is reacted
with four equivalents each of alkoxycarbonyl chloride and sodium
hydroxide in a biphasic mixture of methylene chloride and water
at 0 °C from which guanidine dicarbamate is isolated. In the second
step, the guanidine dicarbamate is reacted with two equivalents
sodium hydride in cold tetrahydrofuran followed by treatment
with one equivalent of alkoxycarbonyl chloride. So it was interest-
ing to discover that methyl chloroformate gave a 25% yield of N,N0,
N00-tri(methoxycarbonyl)guanidine (TMG, 10) directly by the
‘aqueous method’. This compound has apparently not been
reported so it was fully characterized to make certain of the result.
In fact, the ‘aqueous method’ also gave guanidine tricarbamates
from phenyl chloroformate (11) and, in one experiment, benzyl
chloroformate (TCBzG, 12). However, with ethyl chloroformate
the only products isolated by the ‘aqueous method’ after several
runs were either N,N0-di(ethoxycarbonyl)guanidine (13) or N,N0-
di(ethoxycarbonyl)urea (14) [14]. There are experimental details
yet to be defined that can drastically change the course of reaction
in the ‘aqueous method’ because in another experiment with ben-
zyl chloroformate, the product isolated was not TCBzG or N,N0-di
(benzyloxycarbonyl)guanidine (16) but N,N0-di(benzyloxycar-
bonyl)urea (15). The proton NMR spectra, melting points and solu-
bility’s of 15, 16 and desired TCBzG were very similar that
confusion was only cleared up by elemental analysis. Owing to
The diagnostic proton and carbon-13 resonances for the triacyl
and tri(alkoxycarbonyl) orthoamides are collected in Table 1. In
deuterated dimethylsulfoxide, the only practical NMR solvent for
these compounds, the proton signals for these derivatives of
orthoamide were between 6.1 and 6.6 ppm while the carbon-13
signals ranged from 60 to 67 ppm. In comparison, these car-
bonyl-substituted orthoamide proton signals were downfield and
the carbon-13 signals were upfield from those of 1,1,1-tris
(dimethylamino)methane [HC(NMe2)3; deuterochloroform] at
3.10 ppm and 100 ppm, respectively [28].
The commercial TBG was also studied in the reduction reaction
with sodium cyanoborohydride. Using the best conditions devel-
oped for the other guanidine tricarbamates, a solid did not precip-
itate from the reaction mixture. The product isolated after work up
was the aminal di(t-butyl) methylenedicarbamate (20) in 75%
yield, presumably from over reduction, Scheme 5. Compound 20
was reported recently [29,30], synthesized by condensation of
Scheme 4. ‘Aqueous method’ and two-step synthesis of guanidine tricarbamates from guanidine hydrochloride.

430
M.C. Davis, T.J. Groshens / Tetrahedron Letters 60 (2019) 427–431
Scheme 5. Sodium cyanoborohydride reduction of guanidine tricarbamates to orthoamide and aminal.
Table 1
Proton and carbon-13 NMR shifts (ppm) for triacyl and tri(alkoxycarbonyl)
orthoamides [(RCONH)3CH) in DMSO d6] synthesized from guanidine.
R
Compound
1Ha
13Cb
Me
Et
n-Pr
OMe
OCH2Ph
OPh
6
7
8
17
18
19
6.52
6.20
6.55
6.19
6.3
60.77c
61.27
61.08
66.49
66.93
67.34
6.35
a
b
c
200 MHz.
125 MHz.
75 MHz.
t-butyl carbamate and paraformaldehyde, and now the substance
is completely characterized. The compound crystallized in large
gem-like rhombohedrons from acetonitrile, an example of which
is shown, along with an x-ray crystal structure, in Fig. 2 [13].
Further attempts to reduce TBG to orthoamide failed. When the
reduction of TBG was run starting at ice bath temperature, the over
reduced product 20 was obtained. If the reduction was attempted
without any acetic acid present, no reaction took place. Similarly,
no reaction took place when the reduction of TBG was run in a mix-
ture of methanol and tetrahydrofuran. Even when TBG was reacted
with precisely one equivalent of acetic acid and 0.33 equivalent of
NaCNBH3 (one equivalent hydride), the proton NMR spectrum of
the crude product only showed resonances for the methylenedicar-
bamate 20. Alternate conditions known to reduce imines were
tried on TBG: triethylsilane and zinc in methanol; [31] and hydro-
genation (40 psi) in acetic acid and tetrahydrofuran with 5% palla-
dium on carbon catalyst (33 wt% loading). However, these
reactions also gave back starting material. It was already estab-
lished that t-butyloxycarbonyl-protecting groups are stable even
to boiling acetic acid so the formation of di(t-butyl) methylenedi-
carbamate probably does not involve deprotection of TBG [32].
Since it was shown that even tri(methoxycarbonyl) and tri(ethoxy-
carbonyl) orthoamides can undergo carbon–nitrogen bond cleav-
age by NaCNBH3, there is evidently far greater propensity for this
to happen in the case of 1,1,1-tris(t-butoxycarbonylamino)
methane.
Chart 3. Possible reaction mechanism for the reduction of triacyl guanidines by
acetic acid and borohydride reagents.
Lastly, NaCNBH3 could be replaced by NaBH4 which worked
equally well in the reduction of TAG and TMG to their correspond-
ing orthoamides by the typical conditions described above. A
mechanism is proposed for this reduction of substituted guanidi-
nes which is depicted in Chart 3. The sp2-hybridized nitrogen atom
of the substituted guanidine becomes protonated by the acetic
acid, forming a species similar to N-acyliminium salts [15]. The
now electrophilic carbon–nitrogen sp2 bond is attacked by hydride
from the borohydride reducing agent.
Conclusion
In summary, a new method to synthesize orthoamides by
reduction of tri(acyl) and tri(alkoxycarbonyl) guanidines with
borohydride salts has been developed. Further work towards the
synthesis of orthogonally acylated and alkoxycarbonylated orthoa-
mides using this chemistry, in addition to results from reactions
Fig. 2. Large single crystal of di(t-butyl) methylenedicarbamate (20) (left) and its x-ray crystal structure (50% probability thermal ellipsoids) (right).

M.C. Davis, T.J. Groshens / Tetrahedron Letters 60 (2019) 427–431
431
[8] W. Kantlehner, P. Speh, H.J. Braeuner, Synthesis (1983) 905–906.
with borodeuteride salts to further elucidate the reaction mecha-
nism, will be described in a future communication.
[9] R.D.H. Kim, M. Wilson, J. Haseltine, Synth. Commun. 24 (1994) 3109–3114.
[10] R.W. Alder, R.W. Mowlam, D.J. Vachon, G.R. Weisman, J. Chem. Soc. Chem.
Commun. (1992) 507–508.
[11] J.F. Stearns, H. Rapoport, J. Org. Chem. 42 (1977) 3608–3614.
[12] R. Greenhalgh, R.A.B. Bannard, Can. J. Chem. 37 (1959) 1810–1812.
[13] Crystallographic data (excluding structure factors) for the structures in this
paper have been deposited with the Cambridge Crystallographic Data Centre
as supplementary publication nos. CCDC 1879215 (3) and CCDC 1879216 (20).
Copies of the data can be obtained, free of charge, on application to CCDC, 12
Union Road, Cambridge CB2 1EZ, UK, or e-mail: deposit@ccdc.cam.ac.uk.
[14] S. Basterfield, L.E. Paynter, J. Am. Chem. Soc. 48 (1926) 2176–2179.
[15] B.E. Maryanoff, H.-C. Zhang, J.H. Cohen, I.J. Turchi, C.A. Maryanoff, Chem. Rev.
104 (2004) 1431–1628.
Acknowledgments
This research was made possible by support from the Office of
Naval Research, U.S. Department of the Navy. Thanks to Michael
D. Garrison (China Lake) for collecting vibrational analysis on the
new compounds. Thanks to Andy Chafin (China Lake) for the sam-
ple of 1,1,1-tris(ethoxycarbonylamino)methane. Thanks to Jacque-
line Parks of our Technical Library (China Lake) for finding several
of the literature citations. Thanks to Tom Czibovic and Michelle
Wade (China Lake) for gathering the reagents, supplies and
analyses.
[16] R.F. Borch, M.D. Bernstein, H.D. Durst, J. Am. Chem. Soc. 93 (1971) 2897–2904.
[17] S.S. Novikov L.I. Kheml’nitskii S.N. Shvedova E.V. Shepelev O.V. Lebedev Izvest
Akad. Nauk. SSSR, Ser. Khim. 1971 888 889
[18] W. Kantlehner, Tris(formylamino)methane, in: L.A. Paquette (Ed.),
Encyclopedia of Reagents for Organic Synthesis, John Wiley
Chichester, 1995, pp. S. 5443–5445.
& Sons,
[19] L.I. Krimen, Org. Synth. 50 (1970) 1–2.
[20] C.W. Zapf, M. Goodman, J. Org. Chem. 68 (2003) 10092–10097.
Appendix A. Supplementary data
[21] K. Feichtinger, H.L. Sings, T.J. Baker, K. Matthews, M. Goodman, J. Org. Chem. 63
(1998) 8432–8439.
[22] S. Cesco-Cancian, H. Chen, J.S. Grimm, N.S. Mani, C.M. Mapes, D.C. Palmer, D.J.
Pippel, K.L. Sorgi, T. Xiao, US Patent 9 359 327B2, Chem. Abstr. 152 (2010)
119671.
[23] Experiments to increase the yield of TCBzG and other tricarbamates of
guanidine by the ‘aqueous method’ will be reported in a future technical note.
[24] R.H. Barker, S.L. Vail, G.J. Boudreaux, J. Phys. Chem. 71 (1967) 2228–2233.
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.tetlet.2018.12.060.
References
[1] W. Kantlehner, The chemistry of orthoamides of carboxylic acids and carbonic
acid, in: S. Patai (Ed.), The Chemistry of Functional Groups, Supplement B, Part
1, The Chemistry of Acid Derivatives, John Wiley & Sons, Chichester, 1979,
Chapter 9, pp. 533–600.
[2] W. Kantlehner, Science of Synthesis, Thieme Verlag, Stuttgart, 2005, pp. 795–
841.
[3] H. Bredereck, F. Effenberger, T. Brendle, German Patent DE 1 217 391, Chem.
Abstr. 65 (1966) 29127.
[4] D.H. Clemens, E.Y. Shropshire, W.D. Emmons, J. Org. Chem. 27 (1962) 3664–
3670.
[25] H NMR
–
Bio-Rad Sadtler. Bio-Rad Laboratories, Philadelphia, PA;
methylenedicarbamic acid, dimethyl ester; HB_0254.
[26] V. Grakauskas, J. Org. Chem. 36 (1971) 3251–3252.
[27] Hinshaw, J. C. Polycyclic high energy oxidizers. Defense Technical Information
Center ADA128309, April 1983. http://www.dtic.mil/dtic/tr/fulltext/u2/
a128309.pdf. (accessed 16.09.18).
[28] Spectral Database for Organic Compounds, SDBS; National Institute of
Advanced Industrial Science and Technology. Spectra WHPM00534 and
WCDS09401. https://sdbs.db.aist.go.jp. (accessed 7.10.18).
[5] T. Cuvigny, H.C.R. Normant, Acad. Sci. Paris C 270 (1970) 2146–2149.
[6] H. Bredereck, R. Gompper, F. Effenberger, H. Keck, H. Heise, Chem. Ber. 93
(1960) 1398–1401.
[7] R.O. Hutchins, M.K. Hutchins, Reduction of C=N to CHNH by metal hydrides, in:
B.N. Trost, I. Fleming (Eds.), Comprehensive Organic Synthesis, Pergamon
Press, New York,, 1991.
[29] J.-A. Song, T.-L. Choi, Macromolecules 50 (2017) 2724–2735.
[30] B.S. Thyagarajan, K.C. Majumdar, J. Heterocycl. Chem. 11 (1974) 937–942.
[31] K.A. Kumar, T.S. Sreelekha, K.N. Shivakumara, K.C. Prakasha, D.C. Gowda,
Synth. Commun. 39 (2009) 1332–1341.
[32] M.C. Davis, T.J. Groshens, Tetrahedron Lett. 53 (2012) 4154–4155.