Synthesis and thermoelectric properties of 2- and 2,8-substituted tetrathiotetracenes

Article abstract of DOI:10.1039/c8tc00073e

Reaction of elemental sulfur with 2-R¹ and 2,8-R¹,R²-substituted tetracenes (2) in refluxing DMF affords 5,6,11,12 tetrathiotetracenes (1) in good yields (74-99%) for a range of substituents where R¹,R² are: H,H (a); Me,H (b); MeO,H (c); Ph,H (d); Me,Me (e), iPr,Me (f, iPr = iso-propyl, CHMe₂), Me,MeO (g); MeO,MeO (h). The reaction rate is limited only by the solubility of the tetracene (2); 2g-h being both the least soluble and slowest reacting. At partial conversion recovered single crystalline 2g led to its X-ray structure determination. Vacuum deposited (substrate deposition temperature 300 K, pressure 7 × 10^-6 mbar, source temperature 500 K) thin films from 1 (of initial 88-99% purity) show final electrical conductivities, σ_{(in plane)} from 1.40 × 10^-5 S cm^-1 (1g) to 3.74 × 10^-4 S cm^-1 (1b) for the resultant near pristine films; while 1d proved too involatile to be effectively sublimed under these conditions. In comparison, initially 95% pure TTT (1a) based films show σ_(in-plane) = 4.33 × 10^-5 S cm^-1. The purities of 1a-h are highly upgraded during sublimation. Well defined micro-crystallites showing blade, needle or mossy like habits are observed in the films. The Seebeck coefficients (S_b) of the prepared 1 range from 374 (1c) to 900 (1f) μV K^-1 (vs. 855 μV K^-1 for identically prepared 95% pure TTT, 1a). Doping of films of 1f (R¹ = iPr, R² = Me) with iodine produces optimal p-type behaviour: σ_(in-plane) = 7.00 × 10^-2 S cm^-1, S_b = 175 μV K^-1. The latter's power factor (PF) at 0.33 μW m^-1 K^-2 is more than 500 times that of the equivalent I₂-doped TTT films (1a, R¹ = R² = H), previously regarded as the optimal material for thin film thermoelectric devices using acene radical cation motifs.

Full text of DOI:10.1039/c8tc00073e

View Article Online
View Journal
Journal of
Materials Chemistry C
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: S. Woodward, M.
R. Garrett, M. J. Durán-Peña, W. Lewis, K. Pudzs, J. Uzulis, I. Mihailovs, B. Tyril, J. Shine, M. Rutkis and E. F.
Smith, J. Mater. Chem. C, 2018, DOI: 10.1039/C8TC00073E.
This is an Accepted Manuscript, which has been through the
Volume
4 Number 1 7 January 2016 Pages 1–224
Royal Society of Chemistry peer review process and has been
accepted for publication.
Journal of
Materials Chemistry C
Accepted Manuscripts are published online shortly after
Materials for optical, magnetic and electronic devices
www.rsc.org/MaterialsC
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 2050-7526
PAPER
~
Nguyên T. K. Thanh, Xiaodi Su et al.
Fine-tuning of gold nanorod dimensions and plasmonic properties using
the Hofmeister effects
rsc.li/materials-c

Page 1 of 11
Journal of Materials Chemistry C
Manuscript for J. Matter. Chem. C., Garrett, et al.
View Article Online
DOI: 10.1039/C8TC00073E
Synthesis and Thermoelectric Properties of 2- and
2,8-Substituted Tetrathiotetracenes†
Mary Robert Garrett,a María Jesús DuránꢀPeña,a William Lewis,b Kaspars Pudzs,c Janis
Uzulis,c Igors Mihailovs,c Björk Tyril,a Jonathan Shine,a Emily F. Smith,d Martins
Rutkis,*c and Simon Woodward*a
a
GSK Carbon Neutral Laboratories for Sustainable Chemistry, Jubilee Campus, University of
Nottingham, Nottingham NG7 2TU, United Kingdom.
b School of Chemistry, University Park Campus, University of Nottingham, Nottingham NG7
2RD, United Kingdom.
c
Institute for Solid State Physics, University of Latvia, 8 Kengaraga Street, LVꢀ1063 Riga,
Latvia.
d
Nanoscale and Microscale Research Centre (nmRC), School of Chemistry, University Park
Campus, University of Nottingham, Nottingham NG7 2RD, United Kingdom.
Abstract: Reaction of elemental sulfur with 2ꢀR1 and 2,8ꢀR1,R2ꢀsubstituted tetracenes (2) in
refluxing DMF affords 5,6,11,12 tetrathiotetracenes (1) in good yields (74ꢀ99%) for a range of
substituents where R1,R2 are: H,H (a); Me,H (b); MeO,H (c); Ph,H (d); Me,Me (e), iPr,Me (f,
iPr = isoꢀpropyl, CHMe2), Me,MeO (g); MeO,MeO (h). The reaction rate is limited only by the
solubility of the tetracene (2); 2g-h being both the least soluble and slowest reacting. At
partial conversion recovered single crystalline 2g led to its Xꢀray structure determination.
Vacuum deposited (substrate deposition temperature 300 K, pressure 7 × 10ꢀ6 mbar, source
temperature 500 K) thin films from 1 (of initial 88ꢀ99% purity) show final electrical
conductivities, σ(in
from 1.40 × 10ꢀ5 S cmꢀ1 (1g) to 3.74 × 10ꢀ4 S cmꢀ1 (1b) for the
plane)
resultant near pristine films; while 1d proved too involatile to be effectively sublimed under
these conditions. In comparison, initially 95% pure TTT (1a) based films show σ(in plane) = 4.33
× 10ꢀ5 S cmꢀ1. The purities of 1a-h are highly upgraded during sublimation. Well defined microꢀ
crystallites showing blade, needle or mossy like habits are observed in the films. The Seebeck
coefficients (Sb) of the prepared 1 range from 374 (1c) to 900 (1f) ꢁV Kꢀ1 (vs. 855 ꢁV Kꢀ1 for
identically prepared 95% pure TTT, 1a). Doping of films of 1f (R1 = iPr, R2 = Me) with iodine
produces optimal pꢀtype behaviour: σ(inꢀplane) = 7.00 × 10ꢀ2 S cmꢀ1, Sb = 175 ꢁV Kꢀ1. The latter’s
Power Factor (PF) at 0.33 ꢁW mꢀ1 Kꢀ2 is more than 500ꢀtimes that of the equivalent I2ꢀdoped
TTT films (1a, R1 = R2 = H), previously regarded as the optimal material for thin film
thermoelectric devices using acene radical cation motifs.
Introduction
Organic thermoelectric (OTE) materials offer new opportunities for sustainable waste heat
recovery, personalised power (via body heat use) and microcooling (via the Peltier effect).1
Present OTE development is dominated by the use of polymeric thiophenes,
primarily PEDOT [poly(3,4ꢀethylenedioxythiophene)] based materials. For such
polymers only formulation approaches are appropriate for optimising TE
performance as opportunities for structural modification of the PEDOT itself are
†
Electronic supplementary information (ESI) available.
Page 1 of 11

Journal of Materials Chemistry C
Manuscript for J. Matter. Chem. C., Garrett, et al.
Page 2 of 11
View Article Online
DOI: 10.1039/C8TC00073E
highly limited. Recently, radical cation salts of electronꢀrich acene tetrathiotetracenes (TTTs)
(Scheme 1) emerged as a new class of materials for OTE device preparation.2 Such TTT cores
are highly suited to derivatisation (via position 1ꢀ4 or 7ꢀ10 functionalisation). Such
derivitisation is expected to strongly affect the thermoelectric properties of the conducting
states formed upon oxidation of the TTT core potentially leading to improved or tunable OTE
performance. However, derivatives of the parent TTT molecule 1a are rare and their
thermoelectric performance [especially Seebeck coefficients (Sb)] remain essentially unknown.3
We aim to synthesise new tetrathiotetracene derivatives 1 by convenient, easily implemented,
scalable, reproducible procedures. While the unsubstituted TTT (1a, R1 = R2 = H) has been
known for over 70 years and has attracted widespread use in organic electronics,4 extensive
investigations of even its simplest derivatives are scant (being limited to early Soviet literature
or partial claims in industrial patents).5 Only a handful of 2ꢀ, 2,3ꢀ and 2,3,8,9ꢀderivatives
(often with limited or no data supporting the structures) are available (due to difficulties in
their preparation and purification) and none has had its thermoelectric behaviour described.3,5
We proposed to prepare both known and unknown 2 and 2,8ꢀsubstituted tetrathiotetracenes
(1a-g) and to establish a highꢀthroughput screening protocol for comparison of the
thermoelectric properties of new derivatives to the parent TTT 1a. Our aim was to identify
new optimal TTT derivative cores with high Sb values for future use in energy recovery devices
o
operating at lower temperatures (ca. 150 C and below).2 There is a clear reason to target
such materials: the power output of a OTE device is directly proportional to (Sb)2. However,
most present materials, when doped for optimal OTE behaviour, show relatively low Sb values
(e.g. 10ꢀ50 ꢁV Kꢀ1).1a
Scheme 1. Known and new substituted tetrathiotetracenes (TTT is reserved for the parent 1a;
most known derivatives are from the patent literature of 1985ꢀ2000; XꢀZ = alkyl derivatives,
ether or amine derivatives, carbonyl derivatives, halide, CF3, thiopyridyl).5 Many are only
patent claimed structures and are at best partially characterised species. The small grey
numbers show the approved IUPAC numbering for substituted tetracenes.
Results and Discussion
Page 2 of 11

Page 3 of 11
Journal of Materials Chemistry C
Manuscript for J. Matter. Chem. C., Garrett, et al.
View Article Online
DOI: 10.1039/C8TC00073E
Tetrathiotetracene synthesis. Direct reaction of readily available tetracenes (2a-h)6 with
elemental sulfur in DMF at reflux, in the manner of PerezꢀAlberne,7 afforded the
tetrathiotetracenes (1a-h) (Table 1). All are readily isolated in good yields (74ꢀ99%) as very
dark green powders by simple filtration of the hot reaction mixtures giving products with
purities of ca. 90%. While many tetrathiotetracenes 1 can be sublimed under very high
vacuums (<10ꢀ6 mbar), at the more modest pressures attainable in most synthetic laboratories
(ca. 0.1 mbar) significant decomposition results at the higher temperatures (>270 oC) needed
to engender significant volatility. However, simple removal of residual solvents and S8 at 140
oC/0.1 mbar is easily achieved for directly precipitated samples of 1a-h, whose spectroscopic
properties are entirely consistent with the proposed structures. Only traces of nonꢀvolatile
inorganic byꢀproducts remain for such ‘precipitated’ samples of 1, which typically assay as 88ꢀ
99% pure by CHN analytical techniques. Such easily prepared samples are ideal for initial thin
film thermoelectric property screening as only 1 sublimes during film deposition leading to
enrichment to near pristine (>99%) material upon sublimation (see later). A significant
advantage of this ‘direct precipitation’/S8 removal/sublimation approach is that many more TTT
derivatives can be prepared and screened per unit time compared the traditional approaches
(where multiple purifications by either recrystallisation or sublimation are normally required)
prior to measurement of the TTT thermoelectric properties.3 The lowest purity TTT derivatives
(1) we attained were consistently associated with structures 1c (R1 = MeO, R2 = H) and 1h
(R1 = R2 = MeO). Tetrathiotetracenes containing methoxy functions are somewhat more air
sensitive than other derivatives (bulk samples of 1h were fully degraded within three months
under ambient conditions in the solid state). However, all our TTT derivatives 1a-h slowly
(over >10 days) underwent partial aerial degradation when bulk samples were stored at
ambient temperature in the dark in air. Because of this it is recommended that all of the
o
substituted tetrathiotetracenes 1b-h are kept under an inert atmosphere at ꢀ20 C, although
they may all be handled in air over a few hours without ill effect. We had previously seen that
thin films of TTT (1a, R1 = R2 = H) showed related aerial oxidative modification on extended
exposure to air.2c
Preparative details and solution electroꢀoptic properties of the tetrathiotetracenes (1) are
given in Table 1. Typically the reactions were conducted at 0.2 to 1 mmol scales (ca. 50ꢀ200
mg) but these could be scaled up to gram amounts without deleterious effects.
Table 1. Synthesis of tetrathiotetracenes (1) and their optical properties.a
No.
[2] (M)
R1
R2
Time (h) Yield 1 (%) λmax (nm)
Eg(opt) (eV)
1a
1b
1c
1d
1e
1f
0.15
0.15
0.15
0.15
0.15
0.15
H
Me
MeO
Ph
Me
iPr
H
H
H
5
5.5
5
86
86
80
695
691
687
707
686
684
1.53
1.61
1.59
1.59
1.51
1.59
H
Me
Me
5
18
18
>99
89
74
Page 3 of 11

Journal of Materials Chemistry C
Manuscript for J. Matter. Chem. C., Garrett, et al.
Page 4 of 11
View Article Online
DOI: 10.1039/C8TC00073E
1g
1h
0.07
Me
MeO
MeO
18
88
75
682
671
1.54
1.62
0.045
MeO
168
a
Typically reactions were carried out on ca. 0.2ꢀ1 mmol scales in DMF at the nominal
concentrations given (initially 2 is insoluble in DMF) under inert atmospheres.
Interestingly, the rate of formation of 1 from 2 is essentially independent of the electronic
donor/acceptor properties of the substituents within 1 in the eight cases tried. The solubility of
2 in refluxing DMF appears to be the more important factor. For the least soluble tetracenes
(2g-h) the reactions were sluggish and needed an extended heating period and sometimes the
use of more dilute solutions. It was found that at partial conversion (5 h) Xꢀray quality crystals
of 2g could be recovered from experiments aimed at providing the TTT derivative 1g. The
structure of 2g (Figure 1A) shows packing motif features that are classical herringboneꢀtype
interactions. In particular, the closest intermolecular distances are CꢀH…Cipso contacts that
range 2.94ꢀ2.95 Å In comparison, CꢀH…Cipso contacts are absent in a representative literature
TTT (1a) Xꢀray crystal structure (CCDC code: NPDTOL).8
The closest intermolecular
interactions in this structure are between the SꢀS units and the phenylene units (S…CꢀH 3.45ꢀ
3.68 Å) and between the disulfide units themselves (SꢀS…SꢀS 3.75ꢀ5.11 Å); leading to a
greater tendency for faceꢀface interaction (Figure 1B). Attempts to attain single crystals of any
of the TTT derivatives 1b-h have, thus far, been unsuccessful. The anomalous slow reactivity
of 1h appears to be due to initial formation of an intermediate 1:1 adduct between the parent
tetracene and tetrathioꢀderivative (1hꢂ2h) that has very low solubility in DMF resulting in very
slow subsequent formation of 1h. Despite its low solubility, 1H NMR spectra of the adduct
(1hꢂ2h) confirm its 1:1 constitution. The chemical shift differences of the signals in the adduct
vs. the pure components 1h and 2h also supports the formation of a new material. While TTT
(1a) is widely used in organic electronic applications, one limitation is its very low solubility
(ca. 0.3mM in toluene and ca. 3mM in CS2). We hoped that some of the derivatives prepared
here would have higher solubility, but in fact only marginal gains were made testifying to the
high degree of πꢀπ stacking in these systems. Compound 1f showed the highest solubility
about 2ꢀ3 times that of TTT (1a) depending on solvent choice, while 1h showed the lowest
(about a quarter of that of TTT, 1a).
Figure 1. Packing diagrams for (A) tetracene 2g and (B) TTT 1a (structure NPDTOL from the
Cambridge Crystallographic Database);8 hydrogen atoms are not shown for clarity.
(A)
(B)
UVꢀvis spectra of forest green CH2Cl2 solutions of 1 reveal the presence of 4ꢀ5 bands; the
intensity of the lowest energy band maximises at ~700 nm. Onset (Tauc) plots of this band
indicate optical bandgaps of in the range 1.53ꢀ1.62 eV which are only slightly affected by the
Page 4 of 11

Page 5 of 11
Journal of Materials Chemistry C
Manuscript for J. Matter. Chem. C., Garrett, et al.
View Article Online
DOI: 10.1039/C8TC00073E
nature of the substituents within 1 (see Table 1). The optical band gap in all of the
tetrathiotetracenes 1 is significantly lower than the Eg(opt) in the parent tetracenes 2a-g
which range 2.43ꢀ2.54 eV (see ESI). To further confirm that the substituents in our substituted
tetrathiotetracenes 1b-h do affect the electronic states of compounds, representative
examples had their ionisation energies (IEs)9 determined by photoemission spectroscopy (on
deposited films, see ESI).10 Values of ꢀ4.55 eV (TTT, R1 = R2 = H, 1a), ꢀ4.59 eV (R1 = R2
=
Me, 1e), ꢀ4.52 eV (R1 = iPr, R2 = Me, 1f), 4.60 eV (R1 = R2 = OMe, 1h) were determined.
Calculated values (by ∆SCF, mediumꢀpolarized; see ESI) of the IEs were within 2ꢀ6% of the
experimentally determined values. Additionally, transport band gaps Eg(T) were estimated to
be 2.32ꢀ2.51 eV (see ESI) in these calculations. As is typical, values of Eg(T) are greater than
Eg(opt).10
Thin film studies. Thin films of TTT (1a) and the substituted tetrathiotetracenes 1b-h were
prepared by thermal evaporation of freshly prepared samples of 1 under reduced pressure
(see ESI) as described by us previously.2b This procedure results in a significant upgrade in the
purity of the deposited tetrathiotetracenes 1a-h. In order to confirm the validity of this
assumption a number of control experiments (ESI, Section 6) were carried out: (i) The
behaviour of already pristine 1a (previously attained by vapour transport crystallisation2a) was
compared to the 95% purity 1a prepared here by our high throughput ‘direct precipitation’
procedure (Table 2). The inꢀplane electrical conductivity (entries 1ꢀ2, Table 2) of the pristine
and ‘direct precipitated’ 1a were within a factor of 1.8 of each and within a factor of 1.4 on
Seebeck coefficient (Sb). (ii) Initially 88% pure 1h (the TTT material most sensitive to aerial
oxidation and at the lowest purity used herein) was subjected to sublimation under conditions
identical to those used to prepare thin films. The postꢀsublimed dark green microꢀcrystalline
material was confirmed to be >99% pure by 1H NMR, IR, MS and elemental analysis. (iii) IR
studies on the deposited TTT films are identical to those of the preꢀdeposited (single organic
component by 1H NMR spectroscopy) materials indicating that only the TTT derivatives 1 are
deposited. (iv) XPS analysis (O 1s adsorption) of films of 1e indicate no evidence of significant
aerial oxidation (0.05ꢀ0.3 ± 0.2 atomic %), even after handling in air for several hours (Figure
S7a, ESI). Although the very minor residual impurities in the TTT derivatives 1a-h prepared
here do cause some increase in the σ(inꢀplane) conductivity and degradation in Sb performance,
this is small compared to the key substituent effects modifying the behaviour of the parent TTT
to be detected by our screening approach. Identification of improved TTT materials is therefore
straightforward (e.g. the highlighted Sb performance of 1f) using films from the ‘direct
precipitation’ samples prepared above. Based on the validations above, the behaviour of 1b-h
(from samples of 88ꢀ98% purity) was compared to TTT (1a). Three archetypical film
morphologies were observed: blade like morphologies (as previously reported for TTT 1a, but
also seen in films of 1b, 1d and 1h, Figure 2), mossyꢀlike behaviour, shown by 1c (R1 = OMe,
R2 = H, Figure, 2) and needle like habits as exemplified by 1e (R1 = R2 = Me, Figure 2) and
also 1f [for images of all 1a-h see ESI]. All of the 1a-h films show lower electrical
conductivities, as expected, for an insulator phase ꢀ the highest being 1b.
Table 2. Electrical properties of thin films of tetrathiotetracenes (1).a
No.
R1
R2
Film morphology
Film
thickness
(µm)
σ(inꢀplane)
(S cmꢀ1)
Sb
(ꢀV Kꢀ1)
Pristine
H
H
blades
ca. 1
2.37 × 10ꢀ5
1170
1a
Page 5 of 11

Journal of Materials Chemistry C
Manuscript for J. Matter. Chem. C., Garrett, et al.
Page 6 of 11
View Article Online
DOI: 10.1039/C8TC00073E
1a
1b
1c
1d
1e
1f
H
Me
H
H
blades
blades
1.13
0.56
2.89
var.[b]
1.89
0.68
1.51
0.53
4.33 × 10ꢀ5
3.76 × 10ꢀ4
1.34 × 10ꢀ4
1.55 × 10ꢀ6
1.26 × 10ꢀ4
1.45 × 10ꢀ4
1.40 × 10ꢀ5
2.95 × 10ꢀ5
855
464
MeO
Ph
H
H
mossy
blades
374
n.d.[b]
732
Me
iPr
Me
Me
MeO
MeO
needles
needles
mossy needles
blades
900
1g
1h
Me
MeO
501
870
a
Films of 1 were grown on glass (ISOLAB microscope slides) as previously described2b at a
substrate deposition temperature of 300 K, a pressure of 7 × 10ꢀ6 mbar, a deposition rate of
45 ng cmꢀ2 sꢀ1 and a source temperature of 500 K. The reproducibility across duplicates,
including different sample batches, was ±10% for in plane conductivity and Seebeck constant
measurement. For Seebeck measurements a temperature difference of 10 degrees: tcold = 305
K, thot = 315 K was applied (see ESI). The actual thickness of films obtained was measured
b
with profilometer Dektak 150 using stylus with diameter 12.5 µm and stylus force 0.1 mg.
Compound had poor volatility under conditions determined and only a partial film was cast
(variable thickness, up to 2.02 µm).
Figure 2. Representative mossy (morphology A for 1c; R1 = MeO, R2 = H), needle
(morphology B for 1f; R1 = iPr, R2 = Me) and blade (morphology C for 1h) film morphologies
in substituted TTT derivatives. Other members of 1 showed morphologies akin to these types
(see ESI).
(A)
(B)
(C)
Films of the TTT derivatives 1a-h were subjected to post deposition doping with iodine (1
atm., room temperature) as we have previously described (see ESI).2b Doping was continued
until a maximum value of σ(inꢀplane) was attained. The resultant films were air stable and the σ(inꢀ
and Sb values attained are shown in Table 3. The electrical conductivity rose in all cases
plane)
by factors ranging from
6
to >104 times. Reproducibly the 2,8ꢀdisbustituted
tetrathiotetracenes 1e-h led to films showing higher Seebeck coefficients (results highlighted)
than the parent material TTT (1a) or other monosubstituted TTTs (1b-c). Normally pꢀtype
doping of organic thermoelectric materials promotes a severe collapse in Seebeck performance
as electrical conductivity is successively increased by increased pꢀdoping.1a For the parent TTT
1a this behaviour significant limits the efficiency of devices built on TTT/I2 films,2 even when
their doped stoichiometry of the TTT is carefully controlled. Clearly the highlighted Sb
performance of I2 doped 1e-h significantly exceeds that of the parent (1a) and the monoꢀ
substituted analogues (1b-c) trialled here. Because of the (Sb)2 term in the Power Factor (PF)
Page 6 of 11

Page 7 of 11
Journal of Materials Chemistry C
Manuscript for J. Matter. Chem. C., Garrett, et al.
View Article Online
DOI: 10.1039/C8TC00073E
the present studies strongly suggest that tetrathiotetracenes 1 having small electron releasing
substituents at the 2,8 positions will be rather good candidates for future OTE devices; e.g. the
PF of 1f in Table is greater than 500x that of 1a used previously.2
Table 3. Electrical properties of iodine doped thin films of tetrathiotetracenes (1).a
No.
R1
R2
S/I ratiob
σ(inꢀplane)
(S cmꢀ1)
Sb
(ꢀV Kꢀ1)
PF
(ꢀW mꢀ1 Kꢀ1)
1a
1b
1c
1e
1f
H
H
H
H
2.86
2.06
1.95
2.85
2.92
2.08
1.95
9.52 × 10ꢀ3
7.22 × 10ꢀ2
8.30 × 10ꢀ4
4.42 × 10ꢀ3
7.00 × 10ꢀ2
2.25 × 10ꢀ2
6.44 × 10ꢀ3
25
20
17
5.95 × 10ꢀ4
2.89 × 10ꢀ3
2.52 × 10ꢀ5
3.20 × 10ꢀ3
3.27 × 10ꢀ1
5.27 × 10ꢀ3
1.97 × 10ꢀ2
Me
MeO
Me
iPr
Me
Me
Me
MeO
MeO
85
216
48
1g
1h
MeO
175
a
The precursor films 1 are as described in Table 2. Iodine doping was carried out by cyclic
sample exposure to the iodine vapor until the electrical conductivity measured by four contact
technique ceased to rise. The cycle lengths varied from 1ꢀ3 minutes, depending on the
increase of electrical conductivity (see ESI). Such cycling technique has been shown to reduce
cracking of TTT thin films (Ref. 2b). The reproducibility across duplicates, including different
sample batches, was ±10% for in plane conductivity and Seebeck constant measurement. For
Seebeck measurements a temperature difference of 10 degrees: tcold = 305 K, thot = 315 K was
b
applied (see ESI). Sulfur:iodine ratio determined from XRF analysis. Based on instrument
count statistics, error ±5%.
For the parent TTT (R1,R2 = H, 1a) the following single crystalline phases are known3: TTTI
(sulfur:iodine S/I ratio = 4.00), TTT2I3+δ (δ ~ 0.1 giving an S/I ratio = 2.58) and TTTI~2.8 (S/I
ratio = 1.43) these show highest reported single crystal electrical conductivities of: 70 S cmꢀ1,
103 S cmꢀ1 and 80 S cmꢀ1 respectively.3 Xꢀray fluorescence studies of the films of Table 3
revealed that 1a, 1e and 1f provided doping with S/I = 2.89±0.04 while 1b-c and 1g-h led to
S/I =2.02±0.06. The former corresponds to a stoichiometry of approximately ‘TTT3I4+δ’ (δ ~
0.1 equivalent to TTTI + TTT2I3+δ, calculated S/I = 2.93) and the latter to ‘TTTI2’. The ATRꢀIR
spectra of iodine doped films of were distinctly different in accord with substituent effects
causing two different types of doping regime (see ESI). Xꢀray photoelectron spectroscopy
(XPS) of iodine doped 1e (R1,R2 = Me) provided an S/I ratio of 2.85±0.05 (Figure S7b, ESI)
broadly in agreement with the fluorescence data. Attempts at fitting the broad iodine (3d)
doublet observed in the XPS of doped 1e were unsuccessful and consistent with the presence
of a mixture of iodine oxidation states. This is expected for quasicrystalline behaviour of
ꢀ
TTT3I4+δ phases where the iodine counter ions are known to be a continuous combination of I3 ,
I2 and Iꢀ).3 The S 2p region of iodine doped 1e shows a binding energy for sulphur (~164 eV)
consistent with the sulfur(II) disulphide unit in 1e still being intact. The presence of two
different phases in 1a-h whose relative population is controlled by the tetrathiotetracene
substitution pattern may account for the ease with which the thermoelectric properties of the
films can be modified by substituent change. However, as no simple correlation of the
crystalline phase type and either σ(inꢀplane), Sb, or PF is found the value of the high throughput
Page 7 of 11

Journal of Materials Chemistry C
Manuscript for J. Matter. Chem. C., Garrett, et al.
Page 8 of 11
View Article Online
DOI: 10.1039/C8TC00073E
screening approach in identifying new cores for OTE application and further study is again
demonstrated.
As most stateꢀofꢀtheꢀart OTE materials are polymers,11 hybrid materials12 or multilayer
systems13 direct comparison of our own results with them is very complicated due film
morphology and thickness effects on electrical conductivity. Additionally, these literature11ꢀ13
materials can require inert atmosphere handling, while our results are attained in ambient air.
High electrical conductivity and Seebeck values should be a goal to achieve OTEs with high
operation efficiencies (ZT >1) values as ZT =
σ κ where κ = thermal conductivity. Our
(Sb)2/
reported Seebeck values for pure and doped thin films of 1f exceeds most of the values
reported in literature for stateꢀofꢀtheꢀart polymeric OTE materials (Table 4).1b Moderate to very
high Seebeck values have, however been reported for alternative organic systems including
15
pentacene14 and C60 and these are also included in Table 4 for comparison. The electrical
conductivity of our TTT derivative thin (designed only for high throughput screening) films are
much lower than those employed in many OTE devices. Further study and improvement of our
TTT films morphologies will be necessary to increase electrical conductivity without changing
Seebeck coefficient and to produce a complete understanding of the electronic and crystal
packing effects leading to improved
σ and Sb. Evidence that this will be possible is the direct
coincidence of Seebeck coefficient values of tetrathiotetracene iodide thin films and
monocrystals even though their electrical conductivity differs by a factor of >104. Additionally,
better understanding of the TE transport phenomena is attained in the latter.2
Table 4. Comparison of our optimal new TTT 1f vs. stateꢀofꢀtheꢀart OTE materials.
Sb (ꢀV Kꢀ1)
OTE [Ref.]
σ(inꢀplane) (S cmꢀ1)
1.45 × 10ꢀ4
7.00 × 10ꢀ2
604a/
1.1 × 10ꢀ1
3.51a/
5.2 × 10ꢀ1
PF (ꢀW mꢀ1 Kꢀ1)
1f (undoped)
1f (I2 doped)
900
216
19a
0.01
0.33
21.8a
PBTTT-C14/FTS [11a]
60a
60a
1.3a
PBTTT-C14/
F4TCNQ (25 mol%) [11a]
27.7a/
1.8 × 10ꢀ1
~89b
10.0a
P3HT/FTS [11a]
P3HT/TFSI [11b]
PEDOT/PSS [11f]
Pentacene [14a]
C60 [15]
~46b
~73b
20
~880b
~473b
1.7a
4.3 × 10ꢀ1a
<3 × 10ꢀ9a
200a
150000a
<0.007a
a
Measured under inert nitrogen atmosphere. b Determined from the data of the publication.
Conclusions
Because the effects of 2 and 2,8ꢀsubstituents on the thermoelectric properties of TTT
molecules (1) were unknown and the synthesis and purification of TTT derivatives can be very
time consuming, we set out to attain a method adaptable to high throughput screening of such
TTT derivatives for future improved device performance. The recent literature6 availability of 2ꢀ
and 2,8ꢀdisubstituted tetracenes (2) allow a ‘direct precipitation’ approach to 1a-h from the S8
containing DMF reaction mixtures to be developed. The substituted TTT library (1a-h) is of
appropriate purity to allow the deposition of reproducibly pure thin films directly. Both the inꢀ
Page 8 of 11

Page 9 of 11
Journal of Materials Chemistry C
Manuscript for J. Matter. Chem. C., Garrett, et al.
View Article Online
DOI: 10.1039/C8TC00073E
plane electrical conductivity and the Seebeck behaviour and I2 doping studies can be therefore
carried out quickly and with only small amounts of sample (10ꢀ20 mg). The most important
conclusion of our work is that 2,8ꢀdisubstituted TTT molecules (1e-h) show significantly
improved performance in key thermoelectric properties compared to the parent TTT 1a. This is
unexpected. Traditionally, it has been assumed that placing substituents in these positions
would upset the formation of the 1D stacking that is responsible for the very high electrical
conductivity of TTTꢀI2ꢀbased materials.3ꢀ4 The behaviour of the doped films in 1e-h in Table 3
is highly encouraging. While considerable work will be required to disentangle substituent,
doping and morphology effects 2,8ꢀdisubstituted TTT derivatives have been clearly identified
as a new privileged core for further OTE optimisation by our screening procedure. If these
results are transferable to single crystal studies2a (and thermal conductivity were not adversely
affected) efficient near ambient TTT thermoelectric materials would be realised, as has been
long predicted theoretically.16
Experimental Section
Full details of the preparation of 1a-h and measurement of their properties are given in the
ESI.
Conflicts of Interest
There are no conflicts of interest to declare.
Acknowledgments
MRG is grateful for the award of an Appalachian College Association Fellowship for sabbatical
leave support from Berea College, Kentucky, USA. MJDꢀP thanks Ayuda Plan Propio supported
by University of Cádiz. This work was supported, in part, by National Research Program of
Latvia “Multifunctional Materials and Composites, Photonics and Nanotechnology” (IMIS2)
project “Nanomaterials and Nanotechnology”. We are grateful to Jens Pflaum and his group
(Universtität Würzburg) for providing us with a pristine sample of TTT (1a). SW is grateful to
the University of Nottingham for support. An EPSRC grant for the Kratos LiPPS XPS instrument
EP/K005138/1 is acknowledged together with Dr Emily F. Smith at the Nanoscale and
microscale research centre (nmRC), University of Nottingham for acquiring the XPS spectra
and data interpretation.
References
1.
Key overviews of recent OTE developments see: (a) P. J. Taroni, I. Hoces, N. Stingelin,
M. Heeney and E. Bilotti, Isrl. J. Chem., 2014, 54, 534ꢀ552. (b) B. Russ, A. Glaudell, J. J.
Urban, M. L. Chabinyc and R. A. Segalman, Nature Rev. Mater., 2016, 1, 16050. (c) G.
Chen, W. Xu and D. Zhu, J. Mater. Chem. C, 2017, 5, 4350ꢀ4360.
2.
(a) F. Huewe, A. Steeger, K. Kostova, L. Burroughs, I. Bauer, P.Strohriegl, V. Dimitrov,
S. Woodward and J. Pflaum, Adv. Mater., 2017, 29, 1605682. (b) K. Pudzs, A. Vembris,
M. Rutkis and S. Woodward, Adv. Electron. Mater., 2017, 3, 1600429. (c) K. Pudzs, A.
Vembris, J. Busenbergs, M. Rutkis and S. Woodward, Thin Solid Films, 2016, 598, 214ꢀ
218. (d) H. Yoshino, A. Hasegawa, N. Kuroda, R. Tanaka and M. Kozaki in Abstracts of
12th International Symposium on Crystalline Organic Metals Superconductors and
Magnets, 24ꢀ29th September, 2017, Miyagi, Japan.
3.
4.
For example no examples of substituent effects on TTT electrical conductivity (σ) is given
in the key overview of this area: Extended Linear Chain Compounds, J. S. Miller (Ed.)
Plenum Press, New York 1982, Vol. 2, pp. 517. We are also aware of no study on the
effect of TTT structure on Seebeck coefficients (Sb).
Overviews of TTT materials: (a) T. Otsubo and K. Takimiya, Rev. Heteroatom Chem.,
1997, 16, 69ꢀ84. (b) T. Otsubo, K. Takimiya and Y. Aso, Phosphorus, Sulfur Silicon
Relat. Elem., 1998, 136, 447ꢀ462. (c) T. Otsubo, Synlett, 1997, 544ꢀ550. (d) M.
Bendikov, F. Wudl and D. F. Perepichka, Chem. Rev., 2004, 104, 4891ꢀ4945. (e) L.
Page 9 of 11

Journal of Materials Chemistry C
Manuscript for J. Matter. Chem. C., Garrett, et al.
Page 10 of 11
View Article Online
DOI: 10.1039/C8TC00073E
Zhang, S. M. Fakhouri, F. Liu, J. C. Timmons, N. A. Ram and A. L. Briseno, J. Mater.
Chem., 2011, 21, 1329ꢀ1337.
5.
Literature for all known and claimed TTT 1 derivatives we are aware of (substituents
present given in parentheses): (a) M. Wakaki, S. Ishihara, A. Aratani and M. Ando, Jpn.
Kokai Tokkyo Koho, 2000, JP 2000066233 A 20000303, pp. 7 [Chem. Abs., 2000, 132,
187715] (2ꢀR1,3ꢀR2,8ꢀR3,9ꢀR4ꢀTTTs where R1ꢀ4 = H, alkyl, halo). (b) B. Hilti, E. Minder
and J. Pfeiffer, PCT Int. Appl., 1996, WO 9609629 A1 19960328, pp. 52 [Chem. Abs.,
1996, 125, 102251] (2,3,8,9ꢀ(SC5H4N)4ꢀTTT). (c) E. Minder and W. Hofherr, Can. Pat.
Appl., 1994, CA 2100407 A1 19940116, pp. 15 [Chem. Abs., 1994, 121, 71617] (TTT
and 2ꢀFꢀTTT). (d) W. Hofherr, E. Minder, B. Hilti and J. P. Ansermet, Eur. Pat. Appl.,
1993, EP 521826 A1 19930107, pp. 13 [Chem. Abs., 1993, 119, 260554] (2ꢀR1,3ꢀR2,8ꢀ
R3,9ꢀR4ꢀTTT, where R1 = H, F, Cl, or Me; R2ꢀ4 = H, R1ꢀ4 = F or CH3, R1,2 = F, Cl, or Me and
R3,4 = H, or R1,3 or R1,4 = F and R2,4 or R2,3 = H). (e) H. Miyamoto, K. Takimiya, Y. Aso, T.
Otsubo and F. Ogura, Synth. Met., 1991, 42, 2389ꢀ2392 (2,3ꢀMe2ꢀTTT and 2,3,8,9ꢀMe4ꢀ
TTT). (f) T. Maruo, M. T. Jones, M. Singh and N. P. Rath, Chem. Mater., 1991, 3, 630ꢀ
634 (2,3ꢀdiMeꢀTTT). (g) T. Maruo, M. Singh, M. Jones, R. Thomas, P. Nigam and D. Min,
Mater. Res. Soc. Symp. Proc. (Adv. Org. Solid State Mater.), 1990, 173, 149ꢀ154 (2,3ꢀ
diMeꢀTTT and related compounds). (h) T. Maruo, M. Singh and M. T. Jones, NATO ASI
Series, Series B: Physics (LowerꢀDimens. Syst. Mol. Electron.), 1990, 248, 191ꢀ196
(2,3ꢀdiMeꢀTTT). (i) No Inventor data available, Jpn. Kokai Tokkyo Koho, 1990,
JP 02119976 A 19900508, pp. 14. [Chem. Abs., 1991, 114, 155216] (2ꢀR1,3ꢀR2ꢀ
disusbtitutedꢀTTTs where R1 = C1ꢀ30 alkyl, CF3, OR’, SR’, or CO2R’; and R2 = H, C1ꢀ30 alkyl,
CF3, OR’, SR’, or CO2R’ (R’ = C1ꢀ30 alkyl or fluoroalkyl). (j) B. Tieke, A. Wegmann, W.
Fischer, B. Hilti, C. W. Mayer and J. Pfeiffer, Thin Solid Films, 1989, 179, 233ꢀ238 (2ꢀnꢀ
octyloxyꢀTTT). (k) B. Tieke and A. Wegmann, Thin Solid Films, 1989, 179, 109ꢀ120 (2ꢀ
CF3ꢀTTT and 2ꢀnꢀoctyloxyꢀTTT). (l) M. Baumann, C. Mayer, W. Wernet and W. Fischer,
Eur. Pat. Appl., 1989, EP 344112 A2 19891129, pp. 12 [Chem. Abs., 1990, 112,
216483] (2ꢀCF3ꢀTTT). (m) M. Baumann, W. Fischer, V. Kvita, C. W. Mayer and W.
Wernet, Eur. Pat. Appl., 1989, EP 344111 A2 19891129, pp. 19 [Chem. Abs., 1990, 113,
115288] (2ꢀCF3ꢀTTT). (n) J. Finter, B. Hilti, C. W. Mayer, E. Minder and J. Pfeiffer, Eur.
Pat. Appl., 1988, EP 285564 A1 19881005, pp. 22 [Chem. Abs., 1989, 110, 155481] (2ꢀ
R1,3ꢀR2,8ꢀR3,9ꢀR4ꢀTTT, where R1ꢀR2=H, F, Me, alkyl; R3ꢀR4 = COR5 (R5 = halo, OH, NH2,
etc.); R3 and R4 = CO2CO, CONR6CO (R6 = H, residue of primary amine). (o) C. W. Mayer,
V. Kvita and J. Pfeifer, Ger. Offen. 1985, DE 3510092 A1 19850926, pp. 23. [Chem.
Abs., 1986, 105, 235738] (2ꢀCOR, 3ꢀR1ꢀTTT where R = C1ꢀ12 alkoxy, C5ꢀ16 cycloalkoxy,
C2ꢀ12 ωꢀhydroxyalkoxy, C6ꢀ16 aryloxy, C7ꢀ12 aralkoxy, OH, NH2, Cl, Br, monosubstituted
amino; R1 = H, COR, or RR1 together may be CO2 or Nꢀsubstituted CON). (p) V. Kampars
and O. Neilands, Zh. Obsh. Khim., 1979, 49, 2558ꢀ2560 (2ꢀtBuꢀTTT and 2ꢀMeOꢀTTT).
(q) O. N. Eremenko, A. Livdane, R. Medne, O. Neilands, M. L. Khidekel and E. B.
Yagubskii, Izv. Akad. Nauk SSSR, Ser. Khim. 1977, 696ꢀ697 [Chem. Abs., 1977, 87,
39160] (2ꢀtBuꢀTTT). (r) J. Freimanis and A. R. Voron'ko, Latvijas PSR Zinatnu
Akademijas Vestis, Kimijas Serija, 1977, 90ꢀ93 [Chem. Abs., 1977, 86, 189773] (2ꢀiPrꢀ
TTT).
6.
7.
S. Woodward, M. Ackermann, S. Ahirwar, L. Burroughs, M. R. Garrett, J. Ritchie, J.
Shine, B. Tyril, K. Simpson and P. Woodward, Chem. Eur. J., 2017, 23, 7819ꢀ7824.
E. A. PerezꢀAlberne, U.S. Patent 3723417 A 19730327, 1973, pp. 4 [Chem. Abs., 1973,
78, 159696].
8.
9.
O. Dideberg and J. Toussaint, Acta Cryst., Sec. B, 1974, 30B, 2481ꢀ2485.
R. Grzibovskis, A. Vembris and K. Pudzs, J. Phys. Chem. Solids, 2016, 95, 12ꢀ18.
10. J.ꢀL. Bredas, Mater. Horiz., 2014, 1, 17ꢀ19.
11. (a) A. M. Glaudell, J. E. Cochran, S. N. Patel and M. L. Chabinyc, Adv. Energy Mater.,
2015, 5, 1401072. (b) Q. Zhang, Y. Sun, W. Xu and D. Zhu, Energy Environ. Sci., 2012,
5, 9639ꢀ9644. (c) Q. Zhang, Y. Sun, W. Xu and D. Zhu, Adv. Mater., 2014, 26, 6829ꢀ
6851. (d) O. Bubnova, Z. U. Khan, A. Malti, S. Braun, M. Fahlman, M. Berggren and X.
Crispin, Nature Mater., 2011, 10, 429ꢀ433. (e) Q. Wei, M. Mukaida, K. Kirihara, Y. Naitoh
and T. Ishida, Mater., 2015, 8, 732ꢀ750. (f) G.ꢀH. Kim, L. Shao, K. Zhang and K. P. Pipe,
Nature Mater., 2013, 12, 719ꢀ23.
12. E. Jin Bae, Y. Hun Kang, K.ꢀS. Jang and S. Yun Cho, Nature Sci. Rep., 2016, 6, 18805.
Page 10 of 11

Page 11 of 11
Journal of Materials Chemistry C
Manuscript for J. Matter. Chem. C., Garrett, et al.
View Article Online
DOI: 10.1039/C8TC00073E
13. (a) C. Cho, K. L. Wallace, P. Tzeng, J. H. Hsu, C. Yu and J. C. Grunlan, Adv. Energy
Mater., 2016, 6, 1–8. (b) C. Cho, B. Stevens, J.ꢀH. Hsu, R. Bureau, D A. Hagen, O.
Regev, C. Yu and J. C. Grunlan, Adv. Mater., 2015, 27, 2996ꢀ3001.
14. (a) K. Harada, M. Sumino, C. Adachi, S. Tanaka and K. Miyazaki, App. Phys. Lett. 2010,
96, 253304. Higher Seebeck values (ca. 250ꢀ500 ꢀV Kꢀ1) have been reported but at low
current, see: (b) A. von Mühlenen, N. Errien, M. Schaer, M.ꢀN. Bussac, and L. Zuppiroli,
Phys. Rev. B, 2007, 75, 115338. (c) A. A. Günther, J. Widmer, D. Kasemann, and K. Leo
App. Phys. Lett., 2015, 106, 233301.
15. H. Kojima, R. Abe, M. Ito, Y. Tomatsu, F. Fujiwara, R. Matsubara, N. Yoshimoto, and M.
Nakamura, App. Phys. Expr., 2015, 8, 121301.
14. S. Andronic and A. Casian, Adv. Mater. Phys. Chem., 2017, 7, 212ꢀ222 and references
therein.
Graphical Abstract and Table of Contents text
Thermoelectric properties (conductivity and Seebeck coefficient) are in screening for new
tetrathiotetracene lead materials. Iodine doping reveals iPr,MeꢀTTT as optimal.
Page 11 of 11