Novel thione-thiol rearrangement of dithiocarbonic acid o-(2,2,6,6-tetramethylpiperidin- 1-yl) ester in the context of controlled radical polymerization

Article abstract of DOI:10.2533/chimia.2010.56

Addition of the lithium salt of N-hydroxy-2,2,6,6-tetramethylpiperidine 10 to carbon disulfide and subsequent methylation affords the rearranged dithiocarbonic acid S-methyl-S'-(2,2,6,6-tetramethylpiperidin-1-yl)- ester 9 rather than the expected xanthate ester S-methyl-O'-(2,2,6,6- tetramethylpiperidin-1-yl)-ester 8. n-utylacrylate could be polymerized with the novel compound 9 even though the polymerization was not controlled. Schweizerische Chemische Gesellschaft.

Full text of DOI:10.2533/chimia.2010.56

56ꢀ CHIMIAꢀ2010,ꢀ64,ꢀNo.ꢀ1/2ꢀ
doi:10.2533/chimia.2010.56ꢀꢀ
From ChemiCal researCh to industrial appliCations
Chimiaꢀ64ꢀ(2010)ꢀ56–58ꢀ ©ꢀSchweizerischeꢀChemischeꢀGesellschaft
Novel Thione-Thiol Rearrangement of
Dithiocarbonic Acid O-(2,2,6,6-Tetra-
methylpiperidin-1-yl) Ester in the Context
of Controlled Radical Polymerization
a
PeterꢀNesvadba* ,ꢀLucienneꢀBugnona,ꢀandꢀTrixieꢀWagnerb
DedicatedꢀtoꢀProfessorꢀDanielꢀBellušꢀonꢀtheꢀoccasionꢀofꢀhisꢀ70thꢀbirthday
Abstract:ꢀAdditionꢀofꢀtheꢀlithiumꢀsaltꢀofꢀN-hydroxy-2,2,6,6-tetramethylpiperidineꢀ10ꢀtoꢀcarbonꢀdisulfideꢀandꢀsub-
sequentꢀmethylationꢀaffordsꢀtheꢀrearrangedꢀdithiocarbonicꢀacidꢀS-methyl-S’-(2,2,6,6-tetramethylpiperidin-1-yl)-
esterꢀ9ꢀratherꢀthanꢀtheꢀexpectedꢀxanthateꢀesterꢀS-methyl-O’-(2,2,6,6-tetramethylpiperidin-1-yl)-esterꢀ8.ꢀn-Butyl-
acrylateꢀcouldꢀbeꢀpolymerizedꢀwithꢀtheꢀnovelꢀcompoundꢀ9ꢀevenꢀthoughꢀtheꢀpolymerizationꢀwasꢀnotꢀcontrolled.ꢀ
Keywords:ꢀAcyloxyaminesꢀ·ꢀAlkoxyaminesꢀ·ꢀControlled/livingꢀradicalꢀpolymerizationꢀ·ꢀ
Newman-Kwartꢀrearrangementꢀ·ꢀNitroxidesꢀ·ꢀThione-thiolꢀrearrangementꢀ·ꢀThionitroxides
1. Introduction
The intensive research work on NMP at perature slowly into N,N-dimethylcarbam-
Ciba resulted in the development of novel oyl-N’-methylbenzothiohydroxamates 3a.
Ciba has a long-standing tradition and highly efficient alkoxamines and nitrox- Synthetically important is the thermal
leadership in the chemistry of sterically ides[5–8] and Ciba was the first company Newman-Kwart rearrangement[15,16] of O-
hindered amines. These derivatives of worldwide to realize NMP on industrial arylthio carbamates 4 into S-arylthiocar-
2,2,6,6-tetramethylpiperidine (TMP) are scale. Several new products, e.g. pigment bamates 4a which enables the transforma-
used as stabilizers to protect synthetic dispersants, prepared by NMP have al- tion of phenols ArOH into the correspond-
polymers against degradation induced by ready been introduced to the market.[1] ing thiophenols ArSH. Depending on the
light and heat. The related nitroxide radi- The journey towards industrially feasible actual substrate, both ionic[11,12,16] or free
cals inhibit the lignin photo-oxidation and NMP was not always foreseeable but also radical[13,14] mechanisms were postulated
thus ensure that wood keeps its natural accompanied with unexpected results. One for these rearrangements.
color and attractive appearance. Recently, example of such a serendipity observation
the TMP-N-alkoxyamines emerged as en- is the topic of this short article.
vironmentally friendly halogen-free flame
During our work on nitroxide-mediated
controlled/living radical polymerization
Thione to thiol rearrangements are (NMP)[4] we have developed various classes
retardants for plastics.[1] The next opportu- well-known. For example O,S-dialkyl of novel N-alkoxyamines[5–8] which work as
nity to further develop the TMP chemistry xanthates 1 rearrange thermally[9,10] or efficient initiators/regulators for NMP. In the
and expand it beyond the aforementioned under base catalysis[11,12] to S,S-dialkyl same context we observed that the structur-
applications appeared with invention of dithiocarbonates 1a (Scheme 1). Upon ally closely related N-acyloxy-2,2,6,6-tetra-
controlled/living radical polymeriza- heating, ketoxime O-(S-methyl xanthates) methylpiperidines do not induce NMP but
tion (CRP)[2,3] some 20 years ago. From 2 rearrange[13] to S-imino-S’-methyl-di- rather behave as classical free radical initia-
the different CRP techniques known, the thiocarbonates 2a and N,N-dimethylthio- tors.[17] The principle of NMP is based on
nitroxide-mediated controlled/living radi- carbamoyl-N’-methylbenzohydroxamates the reversible, thermally induced homolysis
cal polymerization (NMP)[4] was the most 3 are transformed[14] already at room tem- of the weak (depending on the actual struc-
attractive for Ciba to pursue because the
initiators/regulators for NMP are N-alk-
oxyamines and nitroxides which belong to
the Ciba’s core competence.
*Correspondence:ꢀPDꢀDr.ꢀP.ꢀNesvadbaa
Tel.:ꢀ+41ꢀ61ꢀ636ꢀ24ꢀ12
Fax:ꢀ+41ꢀ61ꢀ636ꢀ23ꢀ32
E-mail:ꢀpeter.nesvadba@basf.com
aCibaꢀInc.ꢀ(nowꢀpartꢀofꢀBASF)
PerformanceꢀChemicalsꢀResearch,ꢀR-1059.6.05
Klybeckstr.ꢀ141
CH-4002ꢀBasel
bNovartisꢀInstitutesꢀforꢀBioMedicalꢀResearch
Lichtstrasseꢀ35
CH-4056ꢀBasel
Schemeꢀ1.ꢀExamplesꢀofꢀthione-thiolꢀrearrangements.

From ChemiCal researCh to industrial appliCationsꢀ
CHIMIAꢀ2010,ꢀ64,ꢀNo.ꢀ1/2ꢀ 57
ture ca. 115–135 kJ/mol)[6,18] NO–C bond
of alkoxyamines affording persistent ni-
troxide radicals and reactive C-centered
radicals, as exemplified on 2,2,6,6-tetra-
methyl-1-(1-phenyl-ethoxy)-piperidine 5
(Scheme 2). Our density functional (DFT)
calculations (Scheme 2 and details in the
experimental part) of bond dissociation
enthalpies (BDE) of the strategic NO–C
and N–OC bonds in 5 and in the model
compounds,
N-acetyloxy-2,2,6,6-tetra-
methylpiperidine 6 and N-methoxycar-
bonyloxy-2,2,6,6-tetramethylpiperidine 7
support this observation. Indeed, the cal-
culation predicts for the NO–C bond in
the alkoxyamine 5 a BDE of 122 kJ/mol
which is in excellent agreement with the
experimental value[18] (119 kJ/mol) and
which is less than the BDE of the N–OC
bond (200 kJ/mol). The situation is just
opposite in the N-acyloxyamines 6 and
7 which suggests that homolysis of these
compounds affords two reactive radicals,
namely 2,2,6,6-tetramethylpiperidinyl and
acetyloxy or methoxycarbonyloxy, respec-
tively. These radicals would directly or
indirectly initiate the uncontrolled polym-
erization. The same calculations predict
even weaker N–OC bond in the unknown
N-methylthiothiocarbonyloxy-2,2,6,6-
tetramethylpiperidine 8. On the other hand
the dithiolcarbonate isomer 9 should ho-
molyse along the NS–C bond to afford
2,2,6,6-tetramethylpiperidinyl-thionitrox-
ide and methylthiocarbonyl radicals. The
low calculated BDE value (114 kJ/mol) of
the N–OC bond in 8 led us to speculate that
this compound may behave as a free radical
initiator working at lower temperature than
derivatives of 6 or 7 and hence prompted
us to attempt its synthesis. The obvious
strategy appeared to be the addition of the
anion of N-hydroxy-2,2,6,6-tetramethylpi-
peridine 10 (Scheme 3) to carbon disulfide
followed by methylation. Surprisingly, we
have found that addition of the lithium
salt of 10 to CS2 in THF and subsequent
treatment with methyl iodide affords the
dithiocarbonic acid S-methyl-S’-(2,2,6,6-
tetramethylpiperidin-1-yl) ester 9 in 77%
yield instead of the expected S-methyl-O-
(2,2,6,6-tetramethyl-piperidin-1-yl) ester
8. This unexpected reaction is thus a new
example of a thione-thiol rearrangement.
Schemeꢀ2. Calculatedꢀbondꢀdissociationꢀenthalpiesꢀinꢀtheꢀcompoundsꢀ5–9ꢀ(kJ/mol,ꢀ298K,ꢀ1ꢀatm,ꢀ
seeꢀintroduction).ꢀ
Schemeꢀ3. Synthesisꢀofꢀtheꢀrearrangedꢀcompound 9.
MS spectra on Finnigan SSQ 710 apparatus.
Elemental analyses were obtained from the
Ciba Service Center Elemental Analysis.
The GPC analysis of the polymer
was performed on a set of 2 PL-Gel
300 × 7.5 mm (5 mm) mixed C columns +
1 PL-Gel guard column using a Spark Hol-
land Basic Marathon sampler and a Flux
RHEOS 4000 pump with inline ERC 1215
a degasser, coupled with ERC 7515A re-
fractometer and ERC7217 UV detector. The
columns temperature was 40 °C, the detec-
tor temperature 30 °C. Tetrahydrofuran was
used as eluent (1 ml/min). Polystyrene stan-
dards were used for the calibration.
The X-ray structure determination was
carried out at Novartis Institutes for Bio-
Medical Research, Basel, Switzerland.[19]
The quantum chemical calculations
were conducted with Spartan 06 (Wave-
function, Inc., USA) and Gaussian 03
(Gaussian Inc., USA) suites of programs.
The ROB3P86/6-311G(2d,2p)//B3LYP/6-
31Gd method, which was used to calculate
the bond dissociation enthalpies (including
zero point energy and corrections to 298 K,
1 atm), is a slight modification (geometry
optimization was performed at B3LYP/6-
31Gd instead of B3P86/6-31Gd level) of
the well-established procedure[20] of DiLa-
bio and coworkers. The relative energies
of 8 and 9 were calculated RB3LYP/6-
311+G(2d,p)//RB3LYP/6-31G(d) and
RMP2-fc/6-311+G(2d,p)//RMP2-fc/6-
31G(d) levels. The minima character of all
optimized species was confirmed by fre-
quency calculations at the same level, the
vibrational frequencies were scaled by the
appropriate scaling factors.[21]
2.1 Dithiocarbonic Acid S-Methyl
Ester S-(2,2,6,6-tetramethyl-piperi-
din-1-yl) Ester (9)
A
solution of 2,2,6,6-tetramethyl-
piperidine-1-oxyl (11, 23.43 g, 0.15 mol,
Fluka >96%) in THF (100 ml, dried over
molecular sieve) and platinum on charcoal
(5% loading) catalyst (450 mg) was hy-
drogenated at room temperature and 2 bar
pressure until the hydrogen uptake ceased
(45 min). The mixture was then rapidly
filtered into a 1 l three-necked flask filled
with argon, the catalyst was washed with
THF (150 ml) and the slightly yellow fil-
trate was cooled to 0 oC. Thereafter, 1.6 M
solution of n-butyl lithium in hexane (100
ml, 0.16 mol, Fluka) was added during 45
2. Experimental
Solvents and reagents are commercial-
ly available and were used as received.
1H and 13C NMR spectra were record-
ed on a Bruker Avance 300 spectrometer
(300 MHz for 1H and 75.47 MHz for 13C).
Chemical shifts (d) are given in ppm down-
field from internal Me4Si standard, J values
are in Hz. IR spectra were taken on Nicolet
Magna-IR 750 spectrometer in ATR mode,
o
min and at 0–2 C and the mixture was
o
stirred for another 30 min at 0 C under
argon. Carbon disulfide (12.56 g, 0.165
mol, Fluka >99%) was then added at
0–2 oC during 20 min. The solution, which

58ꢀ CHIMIAꢀ2010,ꢀ64,ꢀNo.ꢀ1/2ꢀ
From ChemiCal researCh to industrial appliCations
turned from slightly yellow into orange,
was stirred at 0 oC for 75 min. Methyl io-
dide (23.42 g, 0.165 mol, Fluka >99%) was
then added during 45 min while keeping
the temperature at 0–2 oC and the solution
was stirred for additional hour at 0 oC. The
resulting orange-yellow solution was then
evaporated on a rotary evaporator and wa-
ter (100 ml) and methyl-t-butyl ether (150
ml) was added to the residue. The organic
phase was separated, the aqueous layer
was diluted with 25% NaCl solution (50
ml) and extracted with methyl-t-butyl ether
(2 × 50 ml). The combined organic layers
were washed with 25% NaCl solution (2 ×
40 ml), dried over MgSO4 and evaporated.
The solid, slightly yellow residue (36.4
g) was dissolved in toluene (50 ml) and
high polydispersity (M /Mn = 1.9–4) and
is therefore not controllwed/living.
We were able to polymerize bulk n-
butylacrylate with 1 mol% of 9 at 140 oC.
However, the corresponding polymer, ob-
tained in 92% yield, was highly polydis-
perse (M /Mn = 8.6) which indicates an un-
controllewd polymerization in this case too.
4. Conclusions
Fig.ꢀ1.ꢀX-rayꢀstructureꢀofꢀ9.
We have reported a novel thione-thiol
rearrangement of dithiocarbonic acid O-
11 afforded in quantitative yield the hy-
droxylamine 10. Deprotonation of 10 with
n-butyllithium in THF followed by subse-
quent addition of CS2 and methyl iodide
yielded the rearranged compound 9 in
77% isolated yield instead of the expected
isomer 8. The TLC of the crude reaction
mixture before isolation showed that 9 is
the only product formed. The structure of 9
has been confirmed by elemental analysis,
(2,2,6,6-tetramethylpiperidin-1-yl) ester
affording the first example of the hitherto
unknown thionitroxide derivatives. The
new compound 9 is a radical initiator for
the thermal polymerization n-butylacrylate
even though the polymerization is not con-
trolled/living.
o
let crystallize at –20 C to afford 28.7 g
(77.3%) of 9 as almost white, malodorous
crystals, mp. 81–83 oC.
Anal.calcd.forC11H NOS2(247.42):C,
53.40; H, 8.56; N, 5.66; 2S1, 25.92. Found: C,
53.46; H, 8.35; N, 5.57; S, 25.94. MS(ESI):
m/z (%) = 248.2 (100) [M+H]+), 142.3 (60).
IR, cm–1: 1618 (C=O). 1H-NMR (300 MHz,
CDCl3): d = 2.33 (s, 3 H, SCH3), 1.8–1.4 (m,
6 H, 3 × CH ), 1.34 (s, 6 H, 2 × CH3), 1.17
(s, 6 H, 2 ×2CH3). 13C-NMR (75.38 MHz,
CDCl3): d = 201.2 (CO), 60.1, 40.5, 32.2,
25.3, 17.3, 12.4. Crystals for X-ray analysis
were grown from acetonitrile.
Received: December 18, 2009
[1] W. Rutsch, M. A. Cech, Chimia 2007, 61, 33.
[2] W. A. Braunecker, K. Matyjaszewski, Progr.
Polym. Sci. 2007, 32, 93.
[3] G. Moad, E. Rizzardo, S. H. Thang, Acc. Chem.
Res. 2008, 41, 1133.
[4] D. H. Solomon, J. Polym. Sci., Part A: Polym.
Chem. 2005, 43, 5748.
[5] P. Nesvadba, L. Bugnon, R. Sift, J. Polym. Sci.,
Part A: Polym. Chem. 2004, 42, 3332.
[6] H. Fischer, A. Kramer, S. R. A. Marque, P.
Nesvadba, Macromolecules 2005, 38, 9974.
[7] P. Nesvadba, Chimia 2006, 60, 832.
[8] S. Miele, P. Nesvadba, A. Studer,
Macromolecules 2009, 42, 2419.
[9] T. Taguchi, M. Nakao, Tetrahedron 1962, 18, 245.
[10] T. Taguchi,Y. Kawazoe, M. Nakao, Tetrahedron
Lett. 1963, 4, 131.
[11] H. Nakagawa, M. Eto, K. Harano, Tetrahedron
1998, 54, 9333.
[12] H. Nakagawa, M. Eto, K. Harano, Heterocycles
1999, 51, 51.
[13] F. Gagosz, S. Z. Zard, Synlett 1999, 1978.
[14] W. B. Ankers, R. F. Hudson, A. J. Lawson, J.
Chem. Soc., Perkin Trans. 2 1974, 1826.
[15] C. Zonta, O. De Lucchi, R. Volpicelli, L.
Cotarca, Top. Curr. Chem. 2007, 275, 131.
[16] G. C. Lloyd-Jones, J. D. Moseley, J. S. Renny,
Synthesis 2008, 661.
[17] M. Roth, R. Pfaendner, P. Nesvadba, M.-O. Zink,
WO Patent Publication No. WO2001090113,
2001.
[18] W. G. Skene, S. T. Belt, T. J. Connolly, P. Hahn,
J. C. Scaiano, Macromolecules 1998, 31, 9103.
[19] The supplementary crystallographic data
for 9 can be obtained free of charge from the
Cambridge Crystallographic Data Centre via
1
mass spectrometry, H and 13C-NMR and
IR-spectroscopy. Most striking is the pres-
ence of the strong >C=O group absorp-
tion at 1618 cm–1, the >C=S group of the
isomer 8 would not absorb in this region.
Moreover, the 13C-resonance at 201.2 ppm
is compatible with the structure 9 but not
8 whose >C=S signal is expected between
230–240 ppm. The ultimate proof of the
structure 9 was obtained by X-ray analysis
of single crystals grown from acetonitrile
(Fig. 1).
The detailed mechanism of this novel
rearrangement is not clear yet. However,
the observation that addition of one equiv-
alent of aqueous HCl instead of methyl io-
dide affords 2,2,6,6-tetramethylpiperidine
(TMP) and elemental sulfur instead of 10
and CS2 suggests that it is the Li-xantho-
genate 12 which rearranges into the Li-
dithiocarbonate 13. The methylation of 13
then affords 9. The formation of TMP and
sulfur via the mercapto amine 14 can be
plausibly explained (Scheme 3). Decom-
position of N-mercapto amines to amines
and sulfur is known.[24]
We suppose that the main contribution
to the thermodynamic driving force of the
novel rearrangement originates from the
transformation of a relatively weak C=S
bond into a strong C=O bond. Indeed, cal-
culations with DFT and Moeller-Plesset
theory (see Experimental) lead to the same
result indicating that the ground state en-
ergy (298 K, 1 atm) of 9 is 132.9 kJ/mol
resp. 132.5 kJ/mol lower than that of 8.
The compound 9 is a protected thio-
nitroxide, the novel rearrangement thus
opens a new access to thionitroxide de-
rivatives.[25] Bricklebank and Pryke report-
ed[26] that radical polymerization of styrene
in the presence of 2,2,6,6-tetramethylpi-
peridyl-1-thiyl affords polystyrene with
2.2 Bulk Polymerization of
n-Butylacrylate with (9)
A magnetically stirred solution of
9 (0.247 g, 1 mmol) in n-butyl acrylate
(12.82 g, 10 mmol) in a 50 ml round flask
was degassed by three vacuum/argon cy-
cles and was then heated at 140 oC for 90
min to afford colorless, very viscous poly-
1
mer. Integration of the H-NMR (CDCl3)
signals of the OCH2 groups of the unre-
acted monomer (4.16 ppm) and polymer
(4.0 ppm) indicated 92% conversion. GPC:
Mn = 67291, Mw = 580861, Mw/Mn = 8.6.
3. Results and Discussion
Reports on reactions of N,N-dialkyl
hydroxylamines with carbon disulfide are
very scarce. Haase and Wolffenstein ob-
served deoxygenation of diethyl- or diben-
zyl hydroxylamine upon heating with CS2,
however the structure of the products was
not conclusively elucidated.[22] Recently,
preparation of stable O-dialkylamino xan-
thates starting from diethyl- or benzylphe-
nyl hydroxylamine and their use as agents
for controlled/living radical polymeriza-
tion was described[23] in the patent litera-
ture, however, no analytical data were pro-
vided for these compounds.
www.ccdc.cam.ac.uk/data_request/cif
(Code
CCDC-290324)
[20] E. R. Johnson, O. J. Clarkin, G. A. DiLabio, J.
Phys. Chem. A 2003, 107, 9953.
[21] A. P. Scott, L. Radom, J. Phys. Chem. 1996,
100, 16502.
[22] F. Haase, R. Wolffenstein, Ber. Dtsch. Chem.
Ges. 1904, 37, 3228.
[23] D. Charmot, H.-T. Chang, V. Nava-Salgado,
US Patent Publication Application No. US
2004073042, 2004.
[24] F. M. Benitez, J. R. Grunwell, J. Org. Chem.
1978, 43, 2914.
[25] B. Maillard, K. U. Ingold, J. Am. Chem. Soc.
1976, 98, 520.
[26] N. Bricklebank, A. Pryke, J. Chem. Soc., Perkin
Trans. 1 2002, 2048.
Our attempted synthesis of 8 is depict-
ed in Scheme 3. Catalytic hydrogenation
of 2,2,6,6-tetramethylpiperidine-N-oxyl