2,2,5,5-Tetramethyltetrahydrofuran (TMTHF): A non-polar, non-peroxide forming ether replacement for hazardous hydrocarbon solvents

Article abstract of DOI:10.1039/c7gc01392b

An inherently non-peroxide forming ether solvent, 2,2,5,5-tetramethyltetrahydrofuran (2,2,5,5-tetramethyloxolane), has been synthesized from readily available and potentially renewable feedstocks, and its solvation properties have been tested. Unlike traditional ethers, its absence of a proton at the alpha-position to the oxygen of the ether eliminates the potential to form hazardous peroxides. Additionally, this unusual structure leads to lower basicity compared with many traditional ethers, due to the concealment of the ethereal oxygen by four bulky methyl groups at the alpha-position. As such, this molecule exhibits similar solvent properties to common hydrocarbon solvents, particularly toluene. Its solvent properties have been proved by testing its performance in Fischer esterification, amidation and Grignard reactions. TMTHF's differences from traditional ethers is further demonstrated by its ability to produce high molecular weight radical-initiated polymers for use as pressure-sensitive adhesives.

Full text of DOI:10.1039/c7gc01392b

Green Chemistry
View Article Online
View Journal
PAPER
2,2,5,5-Tetramethyltetrahydrofuran (TMTHF): a
non-polar, non-peroxide forming ether
Cite this: DOI: 10.1039/c7gc01392b
replacement for hazardous hydrocarbon solvents†
a
Fergal Byrne,a Bart Forier,b Greet Bossaert,b Charly Hoebers,b Thomas J. Farmer,
a
James H. Clarka and Andrew J. Hunt
*
An inherently non-peroxide forming ether solvent, 2,2,5,5-tetramethyltetrahydrofuran (2,2,5,5-tetra-
methyloxolane), has been synthesized from readily available and potentially renewable feedstocks, and its
solvation properties have been tested. Unlike traditional ethers, its absence of a proton at the alpha-posi-
tion to the oxygen of the ether eliminates the potential to form hazardous peroxides. Additionally, this
unusual structure leads to lower basicity compared with many traditional ethers, due to the concealment
of the ethereal oxygen by four bulky methyl groups at the alpha-position. As such, this molecule exhibits
similar solvent properties to common hydrocarbon solvents, particularly toluene. Its solvent properties
have been proved by testing its performance in Fischer esterification, amidation and Grignard reactions.
TMTHF’s differences from traditional ethers is further demonstrated by its ability to produce high mole-
cular weight radical-initiated polymers for use as pressure-sensitive adhesives.
Received 10th May 2017,
Accepted 4th July 2017
DOI: 10.1039/c7gc01392b
rsc.li/greenchem
and energy intensive due to their high boiling points, 176 and
177 °C respectively.11 Unless a product can be easily crystallized
Introduction
The volatile, non-polar (VNP) class of solvent is important and from the reaction medium, solvent removal by evaporation is
widely used, with applications as reaction media in synthetic often the only alternative. Volatility is a solvent property which
chemistry, in liquid–liquid extractions and in the coating can improve the greenness of a process by facilitating recovery.12
industry where rapid removal by evaporation is necessary. The energy demand to remove less-volatile solvents such as
However, many VNP solvents such as the hydrocarbons hexane limonene and para-cymene, can impair the environmental
and toluene have been recognized as hazardous air pollutants credentials of an otherwise green process. Until new processes
by the US EPA1 and both are suspected of damaging fertility as and products have been developed which do not require solvent
defined under REACH.2,3 Furthermore, hexane is currently on evaporation, green, volatile, and safe solvents are needed.
a CoRAP list for evaluation due to carcinogenicity, mutageni-
Esters are another attractive alternative as they generally do
city and reprotoxicity (CMR) concerns.4 VNP solvents have not possess the high toxicity of hydrocarbons which is demon-
been targeted for replacement with safer, non-volatile, and strated by their natural presence in foods and use as flavour
ideally renewable alternatives in many recent reviews and additives.13 However, at low molecular weights, esters have
solvent guides,5–7 but this has proven to be difficult. Although higher polarity compared to hydrocarbon solvents, impeding
many traditionally used VNP solvents can be made from their use as VNPs.14 At higher molecular weights the polarity
biomass,8,9 the same toxicity and flammability issues as their of esters decreases but with the consequence that their boiling
petroleum-based equivalents persist.2,10
Many of the candidates to replace VNP solvents have much ation energy intensive.15
point increases, making product isolation by solvent evapor-
higher boiling points than the solvent to be replaced. While
Many supposedly greener ethers such as cyclopentyl methyl
limonene and para-cymene have similar solvent properties to ether (CPME), 2-methyltetrahydrofuran (2-MeTHF) and 2,5-di-
toluene, product isolation by solvent removal is challenging methyltetrahydrofuran (DMTHF) have a suitable balance of
polarity and volatility to replace traditional VNP solvents, but
safety issues persist due to their explosive peroxide-forming
aGreen Chemistry Centre of Excellence, University of York, York YO10 5DD, UK.
potential, questioning their green credentials.16–18 Improved
E-mail: andrew.hunt@york.ac.uk
safety has been claimed in the case of CPME but an anti-
oxidant additive, butylated hydroxytoluene (BHT), is required
to prevent peroxidation.16,19 Issues regarding the peroxide
bNitto Belgium, NV Eikelaarstraat 22, Belgium
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c7gc01392b
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
ESI†) which simultaneously purified TMTHF by removing
small amounts of higher boiling point side-products. This
flow-type apparatus allowed the continuous production of
TMTHF by simply adding more DMHDL. Multiple reuse of the
zeolite catalyst was possible by filtering from the reaction
residue and calcining at 600 °C for 4 hours, resulting in no
noticeable loss of catalyst performance in 500 g scale (w.r.t.
DMHDL) batches over a six-month period. Fig. S1 (ESI†) shows
yields from three successive reactions using the same catalyst,
although many more were subsequently carried out with simi-
larly high performance.
Fig. 1 Synthetic route to TMTHF.
The starting material, DMHDL, is currently produced on a
large scale from the reaction of acetylene and acetone via 2,5-
dimethylhex-3-yne-2,5-diol for use in polyethylene copoly-
mers.21,22 The only by-products in the production of TMTHF
are potassium phosphate from the coupling of acetylene and
acetone, and water from the ring closure of DMHDL.
Potassium phosphate has significant potential for downstream
use as a fertilizer in agricultural applications. This results in a
highly efficient production process (atom economy (AE) = 96%,
reaction mass efficiency (RME) = 93%). Using bio-based drop-
in acetone and acetylene as the starting materials, renewable
TMTHF can be produced, securing its long term production
and sustainability (see ESI† for proposed synthetic routes).23–26
forming potential of ethers has recently been reflected in the
pharmaceutical industry CHEM21 solvent selection guide
where 2-MeTHF and CPME are classed as problematic, while
DMTHF has not yet been classified.6
Thus, there is a real need to develop a safer VNP solvent,
evidenced by limited peroxide formation and low mutageni-
city. Additionally, boundaries were set for a solvent with a
boiling point <115 °C (comparable to toluene), a melting point
<−15 °C (for storage in cold climates) and an ability to dissolve
polyisoprene rubber (as a verification of non-polarity). Herein,
the facile synthesis (Fig. 1) and testing of the non-peroxide
forming (without the use of anti-oxidant additives) ether,
2,2,5,5-tetramethyltetrahydrofuran (TMTHF, IUPAC = 2,2,5,5-
tetramethyloxolane), has been described. It was demonstrated
Peroxide tests
to be a safer, potentially renewable alternative to traditional Autoxidation to form peroxides is known to occur readily in
VNP solvents. TMTHF’s resistance to peroxide formation has many chemical compounds, and is particularly an issue with
been investigated and compared to current commercially avail- ethers.27–31 The subsequently formed low-molecular weight
able ether solvents. Finally, its viability as a replacement for peroxides are extremely unstable compounds and in many
VNP solvents has been demonstrated by its application in a cases are shock sensitive.27 While the rates of peroxide for-
series of synthetic reactions, radical-initiated polymerizations mation in some chemical classes in certain conditions have
and solubility tests, and finally its green credentials are been reported, little is known about the concentrations at
discussed.
which they become explosive.27,32
As such, standardized methods of measurement have not
been established and safety thresholds are vaguely defined as
100 ppm.27,32 Peroxide formation in ethers occurs when
protons on the alpha-carbon to the ethereal oxygen are
abstracted by low energy visible light or by a radical initiator
[eqn (1)], followed by oxidation by atmospheric oxygen to form
a radical (R–O–O•) [eqn (2)]. Propagation continues when a
newly formed radical abstracts an alpha-proton from an adja-
cent ether molecule, producing a hydroperoxide (R–O–O–H)
and another ether radical (R•) [eqn (3)].28
Results and discussion
Synthesis of TMTHF
TMTHF was prepared via the facile dehydration of readily avail-
able 2,5-dimethylhexane-2,5-diol (DMHDL). A moderate temp-
erature of 110 °C in the presence of a suitable heterogeneous
solid acid catalyst was found to be sufficient for the synthesis
of TMTHF. A catalyst screen of a range of solid acid catalysts
showed that H-beta-zeolites fully convert DMHDL with excel-
lent selectivity of >99% for TMTHF (Fig. S1, ESI†) at low cata-
lyst loadings.20 At 1 wt% catalyst loading, a reaction time of
2 hours was sufficient for full conversion of DMHDL. Silica/
alumina ratios of 25 : 1 and 150 : 1 were tested with no vari-
ation in yield observed between them (Fig. S1, ESI†). Other
hν
RH ꢀ! R• þ H•
R• þ O2 ! ROO•
ð1Þ
ð2Þ
ð3Þ
ROO• þ RH ! ROOH þ R•
Replacing the adjacent alpha-protons with stronger
solid acid catalysts were also found to produce TMTHF in good bonding groups, such as methyl groups, limits the peroxide
yields (Fig. S1, ESI†), but were discounted as a result of notice- forming potential of the ether molecule via the aforemen-
able side-product formation (1, 2, 3, 4, Fig. S2, ESI†).
tioned mechanism.
The high yielding beta-zeolite catalysed synthesis was easily
This is demonstrated by 1,8-cineol (eucalyptol), a naturally
scaled to 1 L, using a reactive distillation apparatus (Fig. S3, occurring monoterpenoid ether which contains two quaternary
Green Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Green Chemistry
Paper
applications where high volatility is essential. Many similarly
inherently non-peroxide forming ether structures which
contain two quaternary alpha-carbons can be designed but
only two structures, 2,2,5,5-tetramethyltetrahydrofuran
(TMTHF) and di-tert-butyl ether (DTBE), boast solubility pro-
perties and a boiling point within the limits set in this work
(Fig. 2 and Table 1). Adding further functionality to either of
these molecules increases their boiling points above the
threshold set, making further derivatives unsuitable for many
standard applications of VNP solvents. Although both candi-
dates look promising as non-peroxide forming ethers, DTBE
cannot be synthesized by the obvious acid catalysed route
using tert-butanol and isobutene.34,35 Alternative routes
produce by-products resulting in low atom economies, making
the use of DTBE as a solvent less attractive and as such, it was
not considered further in this work.34,36–38
As already demonstrated, TMTHF can be produced catalyti-
cally from readily accessible cheap and bio-available molecules
(acetylene and acetone) via an atom economic pathway. Its
resistance to peroxide formation has been shown experi-
mentally in comparison to other common ethers, in con-
ditions which accelerate the rate of peroxide formation. Air
was bubbled through TMTHF, CPME, 2-MeTHF, THF with
irradiation by UV light (254 nm) in the absence of antioxi-
dants. Peroxide formation was observed in varying amounts in
CPME, 2-MeTHF and THF upon irradiation for 3 hours
(Table S1, ESI†). In contrast, TMTHF did not form any detect-
able peroxides in the same conditions. When TMTHF was
placed under the even harsher conditions of bubbling air,
UV irradiation while under reflux, still no peroxide formation
was observed. Furthermore, the concentration of peroxides in
CPME, 2-MeTHF and THF increased over 6 months under
normal storage conditions (brown chemical bottles with
septum seal) whereas TMTHF did not form any detectable per-
oxides in the same time scale.
Fig. 2 Argus labs picture of quaternary and traditional ethers.
carbons at both alpha-positions to its ethereal oxygen (Fig. 2).
1,8-Cineol has previously been shown not to form hydroperox-
ides upon oxidation but instead breaks down innocuously into
ketones, acids and aldehydes.33 It can be utilized as a solvent,
but its high boiling point (176–177 °C) prevents its use in
Table 1 The physical properties of TMTHF shown in comparison to a range of hydrocarbon, traditional ether and other quaternary ether solvents
Hydrocarbons Quaternary ethers Traditional ethers
Solvent property
Hexane
C-Hexane
Toluene
1,8-Cineol
TMTHF
DTBE
THF
2-MeTHF
Diethyl ether
Mw/g mol−1
Boiling point/°C
Melting point/°C
Density/g ml−1
Mol. vol./cm−3 mol−1
AIT/°C
LEL/v/v%
δD/MPa0.5
δP/MPa0.5
δH/MPa0.5
δ/MPa0.5
86.18
69a
84.16
81a
92.14
111a
−95a
0.867a
106.6e
522f
1.1h
154.25
176a
128.25
112b
130.23
107c
72.11
66a
86.13
78d
74.12
35a
−95a
0.661a
131.4e
225a
1.1a
6a
2a
<−90b
0.802b
151.1e
417f
<−90b
0.762c
172.5e
No data
No data
14.0e
−108a
0.883a
81.9e
321a
2.0a
−136d
0.854d
100.2e
270g
−116a
0.713a
104.7e
180a
0.778a
108.9e
245a
1.3a
0.927a
167.5e
No data
No data
16.6e
2.5e
0.9h
1.5g
1.9a
14.9e
0e
16.6e
0e
18.0e
1.4e
15.4e
2.4e
16.8e
5.7e
16.9e
5.0e
14.5e
2.9e
2.5e
0e
0e
2.0e
2.5e
2.1e
1.4e
8.0e
4.3e
5.4e
14.9e
0.00i
0.00j
0.00k
4.00n
16.6e
0.00i
0.00j
0.00k
3.44n
18.2e
0.00i
0.10j
0.51k
2.73n
16.9e
0.00i
15.7e
0.00i
0.77j
0.35k
1.92b
14.3e
19.5e
0.00i
0.58j
0.59k
0.46n
18.1e
0.00i
0.58l
0.53l
0.77b
15.0e
0.00i
0.47m
0.27m
0.89n
A
B
0.00i
0.72j
0.41k
2.28b
0.51j
0.17k
2.57b
π*
log P(o/w)
a Pubchem. b This work. c Smutny et al.35 d Aycock et al.17 e Predicted by HSPiP. f Carried out by ITS testing services. g Pennakem MSDS.56 h Carried
out by Chilworth Technology. i Assumed value. j This work, using N,N-diethyl-4-nitroaniline and 4-nitroaniline dyes. k This work, using N,N-
diethyl-4-nitroaniline dye. l Jessop et al.41 m Kamlet et al.39 n Sangster.57
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
Physical properties of TMTHF
from which solvation power can be easily visualized relative to
other solvents. HSP theory follows the principle of “like dis-
solves like”, where solvents tend to be able to dissolve nearby
solutes on the HSP map and hence, two solvents with similar
HSP’s are more likely to have similar solvation power.40
Combining the square of the three HSP’s gives the square of
the Hildebrand parameter, δ [eqn (4)]. The Hildebrand para-
meter is a more primitive measure of overall polarity and is
equal to the square root of cohesive energy density (CED)
[eqn (5)]. CED is a measure of a substances self-association.
Increased molecular polarity and hydrogen-bonding results in
increased CED and therefore, a larger Hildebrand parameter.
The physical properties of TMTHF are shown in Table 1. Its
boiling point (112 °C) has been determined experimentally
and was found to be very close to that of toluene (111 °C).
Such a similar boiling point would help ease the substitution
from toluene to TMTHF in industrial processes such as in the
coating industry, where the boiling point of the coating solvent
is a vital consideration. TMTHF’s density has been measured
experimentally at 293 K to be 0.802 g mL−1. Importantly, for
use in industrial processes, its autoignition temperature is
417 °C, higher than the other ether solvents and the two ali-
phatic hydrocarbons, hexane (225 °C) and cyclohexane
(245 °C), which are commonly used in industrial applications.
Although its lower explosion limit (LEL), expressed as a
volume percent with the assumption of it being an ideal gas,
is below that of toluene (TMTHF = 0.9%, toluene = 1.1%), its
safety threshold in terms of mass is superior to toluene due to
its higher molecular weight and lower density (see ESI† for
example calculations).
δ2 ¼ δD þ δP þ δH
ð4Þ
ð5Þ
2
2
2
0:5
δ ¼ ½ðΔH ꢀ RTÞ=Vꢁ
It can be seen in Table 1 that HSP describes TMTHF’s sol-
vation power differently to KAT. While KAT describes TMTHF
to be like other ethers, HSP suggests that the solvation power
Solubility properties of TMTHF
TMTHF also fulfilled the final criteria of this work: it has of TMTHF is more like toluene, an aromatic solvent free of
shown an ability to dissolve non-polar synthetic polyisoprene heteroatoms (Fig. S4, ESI†). As expected, the hydrogen-
rubber, which only the most non-polar solvents are able to do. bonding ability, δH, of the traditional ethers is high, due to
Although TMTHF is an ether by definition, its solvent and their basic ethereal oxygens, which is consistent with KAT.
physical properties are more akin to those of toluene rather However, TMTHF has been assigned a low δH, which is more
than other widely used ethers such as tetrahydrofuran (THF) characteristic of hydrocarbons.
and diethyl ether. A comparison of its Kamlet–Abboud–Taft
(KAT) parameters39 and Hansen Solubility Parameters (HSP)40 (ESP) map can help to explain this difference between HSP
with other solvents is shown in Table 1.
and KAT (Fig. 2).45 The areas coloured red on the ESP map are
TMTHF’s molecular structure and electrostatic potential
KAT parameters consist of three parameters which are areas of high negative charge density, the areas coloured off-
determined by measuring the absorbance of dyes in the pres- white are positively charged while blue represents non-polar
ence of the test solvent.39,41 α is a measure of hydrogen-bond regions. Red areas can be seen on all the ethers shown. This is
donating ability (acidity);42 β is a measure of hydrogen-bond due to the lone-pairs of electrons on their ethereal oxygens,
accepting ability (basicity);43 and π* is a combined measure of which are the source of their high β. While red areas on
polarity and polarizability.44 All the solvents shown in Table 1 TMTHF indicate the presence of hydrogen-bond accepting
are aprotic which is reflected in their α values of zero. Like the lone pairs (in agreement with its high β in KAT), it can be seen
other ethers, TMTHF has high β (0.77) due to the lone pairs of that access to the lone pairs is sterically inhibited by the four
electrons on the ethereal oxygen, while the hydrocarbons in adjacent methyl groups. This results in reduced hydrogen-
contrast have low β values. TMTHF has a low π* (0.35), similar bond accepting ability which is consistent with HSP.
to diethyl ether, while toluene’s high polarizability due to its
The low Hildebrand parameter of TMTHF is as a result of
aromatic ring results in an elevated π* compared to the ali- both its low δH [from eqn (4)] and its low CED [from eqn (5)].
phatic hydrocarbons. Overall, KAT parameters suggest that Although CED has not been measured, the lack of polarity and
TMTHF should behave like a typical ether solvent but interest- the steric hindrance to hydrogen bonding ability would be
ingly, its HSPs and its performance in some synthetic reac- expected to be the main causes of its low value.
tions, described later, indicate that this is not the case.
Overall, while high hydrogen-bond accepting ability is
HSPs are determined by group contribution theory and present in TMTHF, shown by its high β, steric hindrance
describe the characteristic solubility of a molecule using three caused by the four methyl groups inhibits this interaction. To
parameters in a similar way to KAT. Instead of one parameter confirm its suitability as a replacement for hydrocarbon as
to represent both polarity and polarizability as in KAT, the dis- opposed to ether solvents, TMTHF has been tested as a solvent
persion forces (from which polarizability can be calculated) in several chemical reactions. Specific reactions were selected
and polarity are represented as two separate parameters, where the solvents basicity is known to contribute signifi-
δD and δP. In addition, there is only one parameter to represent cantly, both positively and negatively. It has transpired that
total hydrogen-bonding ability, δH. By plotting a molecule’s HSP’s description of TMTHF’s solvation power is more indica-
HSPs on three axes, a three-dimensional map is generated tive of its performance than KAT’s.
Green Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Green Chemistry
Paper
Fischer esterification and amidation reaction testing
The α and π* parameters were found not to be statistically
significant in these reactions. The use of TMTHF as the reac-
tion solvent breaks from the trend observed by Sherwood et al.
in both reactions.
It has previously been shown that the reaction rate of the cata-
lyst-free Fischer esterification of butanoic anhydride with
1-butanol to produce butyl butanoate (Fig. 3) and the catalyst-
free amidation of 4-phenylbutanoic acid with benzylamine to
produce N-benzyl-4-phenylbutanamide (Fig. 4) are enhanced
by solvents with a low basicity, β, by the use of linear solvation
energy relationships (LSER).11
It can be seen in Fig. 3 and 4 that the rates of both reactions
when using TMTHF as the solvent were more comparable to
that of high-performing toluene and para-cymene, than those
of either dimethylformamide (DMF), a solvent with a similarly
high β, or 1,4-dioxane, a traditional ether. This supports the
hypothesis that TMTHF’s high β is, in practice, sterically
impeded by the four methyl groups, resulting in a solvent
which is more similar to the non-polar solvents toluene and
para-cymene, both of which can promote Fischer esterification
and amidation reactions better than normal ethers.
Grignard reaction testing
TMTHF’s disparity from other ethers is further demonstrated
by its inability to form some Grignard reagents. This is shown
in the Grignard reaction between benzyl halides and 2-buta-
none (Table 2). These reaction were chosen as they are known
to proceed to varying degrees using different ethers.46
2-MeTHF was more selective for the Grignard product, 6, while
THF was more selective for the Wurtz product, 5, in reactions
A and B.
TMTHF would be expected to perform as well as the tra-
ditional ethers based on its KAT parameters. However, like
toluene, it was shown to be unable to promote the reaction.
Despite the strong Lewis basicity of TMTHF, the steric hin-
drance of its four bulky methyl groups likely inhibit the lone
pair on the ethereal oxygen from solubilizing the magnesium
halide of the Grignard reagent, preventing its formation and
the subsequent reaction from taking place. Although this is
not a favourable result for a chemist who wishes to perform
Grignard reactions in a new green solvent like TMTHF, it adds
Fig. 3 Fischer esterification reaction scheme and LSER graph.
Table 2 Results of the Grignard reactions in THF, 2-MeTHF, TMTHF and
toluene
Reaction A
Conv./%
Reaction B
Conv./%
Solvent
6 : 5
6 : 5
THF
99
100
0
18 : 82
87 : 13
—
99
100
0
33 : 67
97 : 3
—
2-MeTHF
TMTHF
Toluene
0
—
0
—
Fig. 4 Amidation reaction scheme and LSER graph.
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
weight to the claim that TMTHF behaves more like arene also observed in the final PSA. In contrast, low Mw’s were
hydrocarbons than common ether solvents.
observed when using 2-MeTHF as the solvent (9200 g mol−1).
This is assumed to be due to chain-transfer occurring between
the polymer and solvent by the radical abstraction of the
Radical-initiated vinyl polymerisation using TMTHF
It is not only in organic synthesis that TMTHF can be applied alpha-protons of the ether solvent. Hence, the desired high
as a solvent; TMTHF has also been successfully used in molecular weights could not be obtained. This is further evi-
radical-initiated polymerizations of vinyl monomers.
A
dence that TMTHFs unusual ether structure behaves signifi-
polymerization application which would particularly benefit cantly differently to traditional ethers.
from the use of a greener VNP such as TMTHF is the pressure-
Ames mutagenicity testing and nearest neighbour toxicity
prediction
sensitive adhesive (PSA) industry where high molecular weight
(Mw) polymers (>30 000 g mol−1) are required.47
Currently, VNP solvents such as toluene and hexane are The use of the Ames mutagenicity test to screen new molecules
used in the production of PSAs. However as already men- has recently been proposed as it is quick and cheap.
tioned, REACH restrictions are in place for the use of toluene Encouragingly, an Ames test of mutagenicity was carried out
in adhesives and spray paints (Annex XVII),48 while hexane is a on TMTHF using two Salmonella bacterial strains, TA98 and
suspected CMR.4 2-MeTHF has previously been used as a TA100, where it was shown not to be mutagenic at concen-
greener alternative to THF in the production of anhydride-con- trations up to 5 mg ml−1 (Fig. S5, ESI†).12,50 Full toxicity
taining polymers for radiation sensitive compositions by testing will be required however before it can be authorized by
radical polymerization.49 However, the use of such solvents REACH for production or use of more than 1 ton per year.51
produced polymers with Mw < 4000 g mol−1. Similarly low Mw’s REACH encourages the use of predictive methods for initial
were obtained by the same authors when using THF. While toxicity screening of new molecules to avoid animal testing
low Mw’s may be sufficient in photosensitive compositions, until it is necessary.52 As such, a “nearest neighbour” predic-
higher Mw’s are required for PSA compositions.
tion of toxicity is promising, as the nearest neighbour to
The absence of alpha-protons on TMTHF means appropri- TMTHF is 1,8-cineol. A nearest neighbour prediction makes
ately high Mw’s can be achieved. This has been demonstrated estimations of molecular properties by using chemicals in a
in the polymerization of butyl acrylate and acrylic acid training set that are most structurally similar to the test chemi-
(Table 3). The monomers and dibenzoyl peroxide as the cal.53 1,8-Cineol is a terpene and is often used in fragrances
radical initiator were mixed in TMTHF at 70 °C under an inert and flavorings,54 two applications where human exposure is
atmosphere. The use of TMTHF yielded polymers with com- high and where low toxicity is vital, making it an excellent
parable Mw’s to those obtained with toluene (501 000 and comparison for a potential VNP solvent.55 It has previously
509 000 g mol−1 respectively) as well as comparable solid been suggested as a green solvent, but only due to its source
content (27.25% and 31.00% respectively). Slight differences from biomass and not in the context of its resistance to per-
can be seen in the Mw distribution between TMTHF and oxide formation.41 TMTHF and 1,8-cineol share several mole-
toluene, with a narrower Mw distribution obtained in TMTHF cular properties: two quaternary alpha-carbons resulting in
(Fig. S14 and S15, ESI†). This demonstrates that TMTHF can resistance to peroxidation; low octanol/water partition coeffi-
influence the radical-initiated vinyl polymerization mechan- cients (log P(o/w)) of 1.92 and 2.28 respectively; similar KAT
ism. Improved tack and adhesion compared to toluene was parameters; and similar molecular volumes of 151.1 cm3
mol−1 and 167.5 cm3 mol−1 respectively (predicted by HSPiP),
all of which are shown in Table 1. The low log P(o/w) of both
Table 3 Results of radical polymerisation of butyl acrylate and acrylic
1,8-cineol and TMTHF compared to the alkanes and aromatics
acid using TMTHF, toluene and 2-MeTHF
suggests a reduced likelihood of bioaccumulation in living
organisms. In contrast, the significantly more chemically
active alpha-protons in traditional ethers, demonstrated by
their ability to form peroxides, means they will not give an
accurate nearest neighbour toxicity prediction for TMTHF.
Acid stability
Polymer property
TMTHF
Toluene
2-MeTHF
Acid stability tests were carried out on TMTHF using a range of
acids with varying pKa including glacial acetic acid, trifluoro-
acetic acid (TFA), para-toluenesulfonic acid monohydrate (PTSA),
sulfuric acid (98%) and hydrochloric acid (37%). In these tests,
TMTHF was stirred with a 1 mol% of the selected acid for
24 hours at room temperature or under reflux. The TMTHF used
in the tests was >99% pure by 1H NMR spectroscopy.
Mw/g mol−1 a
501 000
27.25
809
>10
291
509 000
31.00
757
>10
264
9200
31.00
n/af
Solid content/%b
Adhesion/cN 20 mm−1 c
Cohesion/daysd
Tack/ge
n/af
n/af
a Measured by GPC at Nitto Europe. b Percent of solid polymer dis-
solved in solvent. c Measured by the breaking load test at Nitto Europe.
d Measured by the shear adhesion test at Nitto Europe, values shown
are the time of failure in days. e Measured by the rolling ball test at
Nitto Europe. f Not measurable due to the low Mw obtained.
No noticeable colour change was observed after 24 hours at
room temperature and no change could be detected by 1H
Green Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Green Chemistry
Paper
NMR spectroscopy (Fig. S6, ESI†). When stirred under reflux, The process can be made sustainable with the use of renew-
minor colour change from a clear colourless solution to a pale able acetylene and acetone as drop-in replacements. Based on
gold solution was observed in samples mixed with PTSA, sulfu- its structure, TMTHF is classed as an ether and its Kamlet–
ric acid and hydrochloric acid. However, it is not certain that the Abboud–Taft parameters demonstrate high hydrogen-bond
colour change was a result of degradation of TMTHF or of some accepting ability like traditional ethers. However, its Hansen
of the impurities which were initially present in the test solvent. solubility parameters are more like those of toluene. These
1H NMR spectroscopy suggests the presence of some alkene similarities to toluene and disparities to traditional ethers
impurities (∼0.5%) in the test solvent prior to addition of acid, have been confirmed by its performance in model Fischer
which would also be susceptible acid catalysed polymerisation esterification, amidation and Grignard reactions. Furthermore,
via a carbocation intermediate and could result in colour change TMTHF was shown to be a suitable solvent for radical-initiated
(Fig. S7, ESI†). Work is underway to further test ultrapure polymerizations. It is inherently resistant to peroxide for-
TMTHF to give a better understanding of its stability in acidic mation without the assistance of antioxidants, unlike other
conditions, but early results are nevertheless promising.
ethers such as THF, 2-MeTHF and CPME, allowing it to be
These encouraging results suggest that TMTHF is suitable used safely in the chemical industry. Finally, an Ames test
for reactions in acidic conditions at room temperature and using TA98 and TA100 Salmonella bacterial strains found
that no significant degradation occurred when mixed with acid TMTHF not to be mutagenic, further evidence of it being a
under reflux for 24 hours, which opens up potential appli- safer solvent compared those it would replace. Preliminary
cations as a medium in acid catalysis. In addition, the dangers acid stability tests suggest TMTHF can resist acid attack at
associated with mixing acids with peroxides which form in tra- room temperature and results are promising even under
ditional ethers are not present in TMTHF due to its inability to reflux. Such a solvent will create new opportunities for industry
form peroxides. However, precautions must always be taken to improve the green credentials of a process with a safer
and so a peroxide test prior to the addition of any acid would renewable replacement for toluene and other widely used VNP
always be advisable.
solvents.
Green credentials
To date a solvent that is collectively green, volatile, non-polar
and safe has eluded the chemical industry. Our initial study
shows that TMTHF adheres to these criteria and is an ideal
candidate to fill this important gap in the solvent space.
TMTHF has a boiling point of 112 °C, a melting point below
−90 °C, a very useful low polarity and is resistant to peroxide
formation. The potentially renewable route to this molecule
that utilizes clean synthetic methodologies (including catalysis
and flow chemistry) make this a solvent with strong green cre-
dentials. This is further strengthened by the high AE of 96%
and RME of 93% for its synthesis. TMTHF has passed an
initial Ames test for mutagenicity, and nearest neighbour pre-
dictions of toxicity are promising due its structural similarity
to 1,8-cineol, a component of many fragrances and flavor-
ings.54 However, full toxicity testing would need to be con-
ducted before this solvent could be produced or used on a
commercial scale. TMTHF’s boiling point of 112 °C means it
can be removed by evaporation with relative ease, a property
which scores well in the recyclability segment of the GSK
solvent guide.7 It is immiscible with water, glycerol, dimethyl
sulfoxide (DMSO), ethylene carbonate and propylene carbon-
ate, thus opening up applications in liquid–liquid extraction.
TMTHF therefore has the potential to further improve some
processes in which a greener workup solvent is required.
Acknowledgements
The authors would like to thank Nitto Belgium NV for funding
the project; Dr James Sherwood and Dr Con Robert McElroy
for discussions throughout this work; the Biorenewable
Developments Centre (BDC) for the use of their lab where
Ames testing was carried out; Angélique Rat for work carried
out during a summer internship; and Clariant for the HBEA
zeolite samples which were used throughout this work.
Notes and references
1 US EPA – Initial List of Hazardous Air Pollutants with
Modifications, https://www.epa.gov/haps/initial-list-hazar-
dous-air-pollutants-modifications (accessed March 2017).
2 n-Hexane – Substance Information – ECHA, https://echa.
europa.eu/substance-information (accessed March 2017).
3 Toluene – Substance Information – ECHA, https://echa.
europa.eu/substance-information/-/substanceinfo/100.003.297
(accessed March 2017).
4 n-Hexane – Substance evaluation – CoRAP – ECHA, https://
echa.europa.eu/information-on-chemicals/evaluation/com-
munity-rolling-action-plan/corap-table/-/dislist/details/
0b0236e1807e4bd4 (accessed March 2017).
Conclusions
5 P. G. Jessop, Green Chem., 2011, 13, 1391–1398.
6 D. Prat, A. Wells, J. Hayler, H. Sneddon, C. R. McElroy,
S. Abou-Shehada and P. J. Dunn, Green Chem., 2015, 18,
288–296.
TMTHF can be produced catalytically in a highly atom econo-
mical process by the ring closure of DMHDL, a diol which is
currently produced industrially from acetylene and acetone.
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
7 C. M. Alder, J. D. Hayler, R. K. Henderson, A. M. Redman, 33 K. Geier, PhD Thesis, Technische Universität München,
L. Shukla, L. E. Shuster and H. F. Sneddon, Green Chem.,
2016, 18, 3879–3890.
2006.
34 J. L. E. Erickson and W. H. Ashton, J. Am. Chem. Soc., 1941,
8 E. R. Sacia, M. H. Deaner, Y. “Lin” Louie and A. T. Bell,
Green Chem., 2015, 17, 2393–2397.
9 B. O. de Beeck, M. Dusselier, J. Geboers, J. Holsbeek,
63, 1769–1769.
35 E. J. Smutny and A. Bondi, J. Phys. Chem., 1961, 65,
546–550.
E. Morré, S. Oswald, L. Giebeler and B. F. Sels, Energy 36 H. Masada and T. Sakajiri, Bull. Chem. Soc. Jpn., 1978, 51,
Environ. Sci., 2014, 8, 230–240. 866–868.
10 Methylcyclopentane – Substance Information – ECHA, 37 D. H. R. Barton, G. Page and D. A. Widdowson, J. Chem.
https://echa.europa.eu/substance-information/-/substanceinfo/
100.002.277 (accessed March 2017).
11 J. H. Clark, D. J. Macquarrie and J. Sherwood, Green Chem.,
2012, 14, 90.
12 S. Jin, F. Byrne, C. R. McElroy, J. Sherwood, J. H. Clark and
A. J. Hunt, Faraday Discuss., 2017, DOI: 10.1039/c7fd00049a.
13 R. K. Henderson, C. Jiménez-González, D. J. C. Constable,
Soc. D, 1970, 1466a–1466a.
38 G. A. Olah, Y. Halpern and H. C. Lin, Synthesis, 1975, 315–
316.
39 M. J. Kamlet, J. L. M. Abboud, M. H. Abraham and
R. W. Taft, J. Org. Chem., 1983, 48, 2877–2887.
40 C. M. Hansen, Hansen Solubility Parameters: A User’s
Handbook, CRC Press, 2nd edn, 2012.
S. R. Alston, G. G. A. Inglis, G. Fisher, J. Sherwood, S. P. Binks 41 P. G. Jessop, D. A. Jessop, D. Fu and L. Phan, Green Chem.,
and A. D. Curzons, Green Chem., 2011, 13, 854–862.
14 J. Sherwood, PhD Thesis, University of York, 2013.
15 M. Bandres, P. de Caro, S. Thiebaud-Roux and
M.-E. Borredon, C. R. Chim., 2011, 14, 636–646.
16 K. Watanabe, N. Yamagiwa and Y. Torisawa, Org. Process
Res. Dev., 2007, 11, 251–258.
2012, 14, 1245–1259.
42 R. W. Taft and M. J. Kamlet, J. Am. Chem. Soc., 1976, 98,
2886–2894.
43 M. J. Kamlet and R. W. Taft, J. Am. Chem. Soc., 1976, 98,
377–383.
44 M. J. Kamlet, J. L. Abboud and R. W. Taft, J. Am. Chem.
Soc., 1977, 99, 6027–6038.
17 D. F. Aycock, Org. Process Res. Dev., 2006, 11, 156–159.
18 E. A. Youngman, F. F. Rust, G. M. Coppinger and H. E. D. 45 ArgusLab, http://www.arguslab.com/arguslab.com/ArgusLab.
L. Mare, J. Org. Chem., 1963, 28, 144–148. html (accessed March 2017).
19 Cyclopentyl methyl ether (CPME) – Zeon Corporation, 46 A. Kadam, M. Nguyen, M. Kopach, P. Richardson,
http://www.zeon.co.jp/business_e/enterprise/spechemi/spe-
chemi5-13.html (accessed March 2017).
F. Gallou, Z.-K. Wan and W. Zhang, Green Chem., 2013, 15,
1880–1888.
20 F. Byrne, A. J. Hunt, T. J. Farmer, J. H. Clark, B. Forier and 47 F. Byrne, A. J. Hunt, T. J. Farmer, B. Forier and G. Bossaert,
G. Bossaert, NL Pat, P32827NL00/WZO, 2016 (patent
pending).
21 Dimethylhexanediol – BASF, http://www.windenergy.basf.
com/group/corporate/wind-energy/en/brand/2_5_DIMETHYL_
2_5_HEXANEDIOL (accessed March 2017).
NL Pat, P32826NL00/MKO, 2016 (patent pending).
48 Substances restricted under REACH – ECHA, http://echa.
europa.eu/addressing-chemicals-of-concern/restrictions/
substances-restricted-under-reach, (accessed March 2017).
49 S. Dilocker, S. Malik, B. De and A. Reddy, US Pat,
20030225233A1, 2003.
22 M. Maas-Brunner, J. Henkelmann, G. Kaibel, A. Kindler,
C. Knoll, H. Rust, C. Tragut, M. Stroezel, U. Rheude and 50 B. N. Ames, W. E. Durston, E. Yamasaki and F. D. Lee, Proc.
R. E. Lorenz, US Pat, 6956141B1, 2005. Natl. Acad. Sci. U. S. A., 1973, 70, 2281–2285.
23 G. Esposito, L. Frunzo, A. Panico and F. Pirozzi, Environ. 51 Regulation (EC) No 1907/2006 – REACH, https://osha.
Technol., 2012, 33, 2733–2740.
24 M. Görling, M. Larsson and P. Alvfors, Appl. Energy, 2013,
112, 440–447.
25 D. Stapf, P. Pässler, M. Bachtler, O. Scheidsteger and
B. Bartenbach, US Pat, 6365792B1, 2002.
26 Renewable, Bio-based Acetone, http://www.greenbiologics.
com/acetone.php (accessed March 2017).
27 D. E. Clark, Chem. Health Saf., 2001, 8, 12–22.
28 C. E. Frank, Chem. Rev., 1950, 46, 155–169.
29 C. E. Redemann, J. Am. Chem. Soc., 1942, 64, 3049–3050.
30 D. B. Sharp, L. W. Patton and S. E. Whitcomb, J. Am. Chem.
Soc., 1951, 73, 5600–5603.
europa.eu/en/legislation/directives/regulation-ec-no-1907-
2006-of-the-european-parliament-and-of-the-council.
52 What about animal testing? – ECHA, https://echa.europa.
eu/chemicals-in-our-life/animal-testing-under-reach.
53 US EPA, Toxicity Estimation Software Tool (TEST) (Version
4.2.1.).
54 Eucalyptol – WHO, http://apps.who.int/food-additives-
contaminants-jecfa-database/chemical.aspx?chemID=4037
(accessed March 2017).
55 1,8-Cineole – TOXNET, https://toxnet.nlm.nih.gov/cgi-bin/
sis/search2/r?dbs+hsdb:@term+@DOCNO+991 (accessed
March 2017).
31 V. Fábos, G. Koczó, H. Mehdi, L. Boda and I. T. Horváth, 56 2-Methyltetrahydrofuran
– Penn Speciality Chemicals,
Energy Environ. Sci., 2009, 2, 767.
http://www.pennakem.com/pdfs/methf5.pdf.
32 R. J. Kelly, Chem. Health Saf., 1996, 3, 28–36.
57 J. Sangster, J. Phys. Chem. Ref. Data, 1989, 18, 1111–1229.
Green Chem.
This journal is © The Royal Society of Chemistry 2017