Crystal growth of new charge-transfer salts based on π-conjugated donor molecules

Article abstract of DOI:10.1016/j.physb.2016.05.023

New charge transfer crystals of π-conjugated, aromatic molecules (phenanthrene and picene) as donors were obtained by physical vapor transport. The melting behavior, optimization of crystal growth and the crystal structure are reported for charge transfer

Full text of DOI:10.1016/j.physb.2016.05.023

Author’s Accepted Manuscript
Crystal Growth of new charge-transfer salts based
on π-conjugated donor molecules
Antonia Morherr, Sebastian Witt, Alisa
Chernenkaya, Jan-Peter Bäcker, Gerd Schönhense,
Michael Bolte, Cornelius Krellner
www.elsevier.com/locate/physb
PII:
S0921-4526(16)30209-5
http://dx.doi.org/10.1016/j.physb.2016.05.023
PHYSB309489
DOI:
Reference:
To appear in:
Physica B: Physics of Condensed Matter
Received date: 21 April 2016
Revised date: 19 May 2016
Accepted date: 20 May 2016
Cite this article as: Antonia Morherr, Sebastian Witt, Alisa Chernenkaya, Jan-
Peter Bäcker, Gerd Schönhense, Michael Bolte and Cornelius Krellner, Crysta
Growth of new charge-transfer salts based on π-conjugated donor molecules
Physica
B:
Physics
of
Condensed
Matter
http://dx.doi.org/10.1016/j.physb.2016.05.023
This is a PDF file of an unedited manuscript that has been accepted fo
publication. As a service to our customers we are providing this early version o
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain

Crystal Growth of new charge-transfer salts based on
π-conjugated donor molecules
Antonia Morherr*1,, Sebastian Witt1,, Alisa Chernenkaya2,3,, Jan-Peter
Ba¨cker1,, Gerd Scho¨nhense3,, Michael Bolte4,, Cornelius Krellner1,
aMax-von-Laue-Strasse 1, 60438 Frankfurt am Main, Germany
Abstract
New charge transfer crystals of π-conjugated, aromatic molecules (phenanthrene
and picene) as donors were obtained by physical vapor transport. The melting
behavior, optimization of crystal growth and the crystal structure is reported for
charge transfer salts with (fluorinated) tetracyanoquinodimethane (TCNQ-Fx,
x=0, 2, 4), which was used as acceptor material. The crystal structures were
determined by single-crystal X-ray diffraction. Growth conditions for different
vapor pressures in closed ampules were applied and the effect of these starting
conditions for crystal size and quality is reported. The process of charge transfer
was investigated by geometrical analysis of the crystal structure and by infrared
spectroscopy on single crystals. With these three different acceptor strengths
and the two sets of donor materials, it is possible to investigate the distribu-
tion of the charge transfer systematically. This helps to understand the charge
transfer process in this class of materials with π-conjugated donor molecules.
Keywords: charge transfer, TCNQ, physical vapor transport
Email address: morherr@stud.uni-frankfurt.de (Antonia Morherr*)
1Physikalisches Institut, Goethe-Universit¨at Frankfurt am Main, 60438 Frankfurt am Main
2Graduate School Materials Science in Mainz, 55128 Mainz
3Institut fu¨r Physik, Johannes Gutenberg-Universita¨t, 55099 Mainz
4Institut fu¨r anorganische Chemie, Goethe-Universit¨at Frankfurt am Main, 60438 Frank-
furt am Main
Preprint submitted to Physica B
May 21, 2016

1. Introduction
Designing new organic materials and tuning their physical properties by
adding or removing charges has a long history[1], but the detailed understanding
of mechanisms is still a matter of current research. Organic charge-transfer
(CT) salts offer a huge variability in crystal structures and physical properties
ranging from metallicity over superconductivity to Mott insulators [2, 3] on one
side. Organic semiconductors on the other side find application in various diodes
and organic electronics due to their flexibility and relatively high mobility [4].
Also, a new class of multiferroicity was found in organic charge transfer salts
5
10 and introduced a new research field [5].
In the last years, the class of phenacenes attracted a lot of attention. For
potassium doped phenanthrene (n=3 benzene rings) [6], picene (n=5) [7] and
dibenzopentacene (n=7) [8] superconducting phases were discovered with a TC
of 5 K (n=3), 18 K (n=5) and 33 K (n=7), respectively. The charge trans-
15 fer process in crystals based on these three compounds is of interest, as the
reproducibility of these samples is still under debate [9] and only few exper-
imental studies have been reported. In this paper, we expand the sets of
charge transfer (CT) salts based on aromatic donor molecules as reported for
phenanthrene/TCNQ [10] and picene/TCNQ-F4 [11] and present the crystal
20 structures of new charge transfer salts phenanthrene/TCNQ-Fx (x=2, 4) and
picene/TCNQ-Fx (x=0, 2). With respect to the weak donor ability of phenan-
threne and picene (HOMO: 6.1 eV, 5.5 eV [12]) and the strong acceptors TCNQ,
TCNQ-F2 and TCNQ-F4 (LUMO: 4.2 eV, 4.55 eV and 5.24 eV[13]) neutral CT
are expected. With the two known structures of PHN/TCNQ and PIC/TCNQ-
25 F4, we yield a set of two donors with three different acceptors. By changing the
fluorine content in the acceptor, the overall charge transfer is increased. These
novel materials expand the materials basis for the recently suggested HOMO-
LUMO engineering [13] utilizing charge-transfer compounds.
2

2. Experimental details
30 2.1. Crystal growth
Phenanthrene (PHN) and picene (PCN) as donor molecules (D) were pur-
chased by TCI with a purity of >97% and >99%, respectively. TCNQ, TCNQ-
F2 (2,5-configuration) and TCNQ-F4 as acceptors (A) were also purchased by
TCI with purities of >99%, >98% and >98%, respectively. These starting
35 molecules are shown in Figure 1. The acceptor strength increases for higher
fluorinated acceptors with an electron affinity of 2.85 eV for TCNQ; 3.02 eV for
TCNQ-F2 and 3.20 eV for TCNQ-F4 [14].
a)
b)
C6
C9
C13
C25'
C14'
c)
d)
e)
N
N
N
N
N
N
d
d
d
C33
c
c
c
b
b
b
F
H
a
F
F
H
H
H
a
a
C27
H
H
F
F
F
C30
N
C15
N
C31
N
N
C16
N
N
Figure 1: a) Phenanthrene and b) picene used as donor molecules and c) TCNQ, d) 2,5-TCNQ-
F2 and e) TCNQ-F4 as acceptors are shown. The bond lengths a-d are marked in the acceptors
to describe structural effects of charge-transfer. The colors representing different atoms in the
molecules are as follows: grey (C), white (H), blue (N) and green (F). The numbers of C
atoms refer to selected distances between acceptor and donor molecules, discussed within this
work.
Single crystals were grown by physical vapor transport (PVT)[15, 16] in
closed glass ampules. For this purpose, the starting materials (as powders) of
40 D and A were ground together and transferred into a glass ampule, which was
cleaned by ethanol and acetone before and baked in a box furnace for at least
3

Table 1: Growth parameters for phenanthrene and picene based charge-transfer salts.
mdonor [mg] macc. [mg] Tmax [◦C] TCT [◦C] stoichiometry
PHN/TCNQ
26.8
13.2
2.6
200
200
200
240
245
240
160 - 130 1:1
180 -160 1:1
186 - 147 1:1
ca. 200 1:1
220 - 185 1:1
ca. 200 1:1
PHN/TCNQ-F2 17.2
PHN/TCNQ-F4 21.8
12.1
15.9
18.2
20.1
PCN/TCNQ
10.8
PCN/TCNQ-F2 26.4
PCN/TCNQ-F4 17.1
24 hours to remove water (as described in Ref. 11). The glass ampule was
evacuated to p = 10−3 mbar before closing. The experimental oven setup and a
typical temperature gradient used in the experiments is shown in Figure 2 with
45 Tsource as temperature at the hottest position with the source material and
Txcrystal with x = 1, 2, CT, where x denotes the crystal growth temperature of
material 1, material 2 and the CT complex, respectively. Typical crystal growth
conditions are listed in table 1.
The crystal growth zone lies in a region with a steep temperature gradient,
50 which results in the best crystal growth as mentioned and suggested in Ref. 17.
The growth procedure lasted ca. 50 hours until all starting material is sublimed
and subsequently the ampule was removed from the hot oven. The section with
the crystal growth zone of an ampule after the growth procedure of phenan-
threne/TCNQ is also shown in Figure 2. Here, TCNQ (orange) resublimes
55 at higher temperatures (T1) compared to the (black) CT complex (TCT ) and
(transparent) phenanthrene (T2). All charge transfer crystals grow in a 1:1 stoi-
chiometry with the applied starting conditions. There is no other stoichiometry
detected as for example 2:1 in Coronene2/TCNQ-F4 [18]. Besides the growth
of CT salts in a closed system, the oven can be used also for crystal growth
60 or purification of organic semiconductors (molecular crystals with one type of
molecule). Here, the crystal growth is performed under a stream of argon and
a flat temperature gradient in the oven is necessary.
For phenanthrene/TCNQ small needle-like crystals were obtained with a
4

200
160
120
80
T1
TCT
T2
Tsource
40
0
4
8
12
16
Distance [cm]
Figure 2: The PVT oven for crystal growth is shown on the top designed after the working
principle used in Ref. 19. A typical temperature profile for phenanthrene based CT crystals
is shown at the bottom with the hottest zone of around 200◦C. TCNQ resublimes first at
T1=170◦C as orange crystals, the CT complex (black) at TCT =140◦C and the phenanthrene
at T2=110◦C as transparent crystals.
source temperature Tsource=200◦C. For phenanthrene/TCNQ-F4, this method
65 with a pressure of p = 10−3 mbar results only in micro-crystallites (smaller than
50 μm), where a separation from the glass walls after the growth procedure is
not possible. For optimization of the crystal growth conditions, we varied the
argon gas pressure inside the ampule. For this purpose, the ampule is first
evacuated to p = 10−3 mbar and than refilled with argon as inert gas. The
70 pressure amounts to the value at room temperature. In Figure 3 typical results
are shown for two different argon pressures for phenanthrene/TCNQ. The single
crystals in the first case for a pressure of p = 10−3 mbar are smaller than 1 mm
and have a width and thickness of 250 μm in average. For the second case with
5

an inert gas pressure of 0.5 bar, larger crystals up to a length of 2 mm in average
75 are detected.
a)
b)
Figure 3: Ampules with phenanthrene/TCNQ after growth in the charge transfer region for
different pressures of argon a) p = 10−3 mbar, b) p = 0.5 bar and c) 1 bar.
The improved crystal growth conditions are reflected in more block-like crys-
tals by higher inert gas pressure. This can be due to increased sublimation tem-
peratures and modified growth rates for different pressures mentioned in Ref. 17.
Higher pressures might lead to a suppressed nucleation rate due to a smaller
80 oversaturation at the ampule walls in the crystal growth regime. Therefore, the
crystals grow slowly and a block-like habitus appears.
2.2. DTA
For a better understanding of the crystal growth conditions, differential ther-
mal analysis (DTA) measurements were performed with phenanthrene, picene,
85 TCNQ, TCNQ-F4 and a mixture of donor and acceptor materials shown in Fig-
ure 4. The starting materials were filled into DTA glass ampules. The ampules
were evacuated to p = 10−3 mbar, refilled with 0.5 bar argon and closed as de-
scribed for the growth ampules above. The measurements were obtained by a
STA 409 from NETZSCH and a DTA sample holder (type E). For heating and
90 cooling, rates of 5 K/min were applied. The four ampules with pure phenan-
threne (0.16 mmol), picene (0.089 mmol), TCNQ (0.047 mmol) and TCNQ-F4
(0.041 mmol) were heated to temperatures above the melting points of the dif-
ferent materials and then were cooled down again. Both processes were recorded
with DTA and are presented in Figure 4. Melting points in DTA experiments
6

95 are defined as the onset of the peak. The melting point for phenanthrene was
reported to be at 100◦C [20] and TCNQ melts around 293◦C [21]. These temper-
atures are reproduced in DTA measurements shown in Figure 4 for a) phenan-
threne and b) TCNQ with onsets at T=100◦C and at T=290◦C, respectively.
An additional peak appears for phenanthrene at 69◦C. This peak shows the
100 phase transition (PT) to the high temperature crystal structure for phenan-
threne as reported in Ref. 20. By cooling down phenanthrene, the two peaks
were also detected with Tcryst. = 87◦C and TP T =73◦C for the low temperature
crystal structure. For TCNQ (see Figure 4 b)) an exothermal peak appears
in the heating curve at Tcrack=375◦C after the melting point at TM =290◦C.
105 The recrystallization peak is missing in the cooling part, which means, that the
exothermal peak in the heating curve signalized a decomposing of the TCNQ
molecules. The ampule with both starting materials with a surplus of phenan-
threne (0.037 mmol of PHN and 0.024 mmol of TCNQ) is first heated to a tem-
perature of T=305◦C to be above the melting point of TCNQ, but far below
110 the cracking temperature. Phenanthrene melts and starts to solve TCNQ above
100◦C. After this first run, the powder was dark red, confirming a charge trans-
fer process. The ampule was heated again to 300◦C and cooled down with
measuring the DTA signal. In the heating part of the curve, a peak around
100◦C appears again, which is due to the intact phenanthrene, not incorporated
115 into the CT complex. The melting point of the CT appears at TM =242◦C (see
Figure 4c)). The peak also is observed in the cooling part at Tcryst. = 228◦C
meaning both compounds were intact and TCNQ was not decomposed in this
temperature regime.
The melting point for the complex phenanthrene/TCNQ lies in between the
120 melting points for the single compounds, which looks reasonable and is in good
agreement to the different crystal growth areas for acceptor, CT complex and
donor[11, 22, 18] and is also apparent in Figure 2. Sublimation points could not
be detected by DTA but are stronger depending on pressure than melting points.
By using additional inert gas inside the ampule, the sublimation temperature is
125 shifted to higher temperatures and sublimation rates are slowed down achieving
7

a)
b)
c)
d)
e)
f)
Figure 4: DTA measurements for a) phenanthrene, b) TCNQ and c) both compounds in a
closed DTA glass ampule are shown. In d)+e) the same measurements for picene and TCNQ-
F4 are shown. The peaks for melting points show up at TMP HN =100◦C, TMT CNQ=290◦C,
TMP HN/T CNQ=242◦C, TMP CN =364◦C and TMT CNQ−F4 =290◦C. Phenanthrene shows a phase
P CN
P T
transition at TPPTHN =69◦C and a similar peak is measured for picene at T
=234◦C. The
mixed compound of picene and TCNQ-F4 in f) shows no CT peak as the cracking temperature
of TCNQ-F4 lies below the melting point of picene.
larger crystals.
For picene (0.023 mmol) and TCNQ-F4 (0.023 mmol), DTA measurements were
also performed and presented in Figure 4d) together with the TCNQ data in
Figure 4e). Also for picene a phase transition probably due to a structural
130 transition is visible around 234◦C, similar to the one detected for phenanthrene.
For TCNQ-F4 we detect an exothermal peak after the melting point, similar to
TCNQ but at a lower temperature of 315◦C. The melting point of picene lies at
362◦C as shown in Figure 4 and is therefore similar to the cracking temperature
of TCNQ and above the cracking temperature of TCNQ-F4. No detection of
135 the melting point is possible for both complexes as the acceptor molecules will
decompose before melting.
8

The shown DTA measurements imply two important points for the improve-
ment of crystal growth by PVT: To avoid cracking of the acceptor molecules,
temperatures below 315◦C are necessary for crystal growth of pure TCNQ but
140 also for CT salts based on these acceptors from the vapor phase. To avoid
a much faster sublimation of phenanthrene at p = 10−3 mbar resulting in a
separation in crystal growth area and a small yield of charge transfer salts, ad-
ditional argon pressure is essential. This increased pressure results in a much
better crystal growth for all acceptor compounds with phenanthrene as donor
145 molecule.
2.3. Single Crystal Structure
Single crystal X-ray diffraction was performed on selected crystals to deter-
mine the crystal structure. The detailed information of crystal structures are
summarized in the SI.
150
Data for phenanthrene/TCNQ-F2, phenanthrene/TCNQ-F4 and picene/TCNQ-
F2 were collected on a STOE IPDS II two-circle diffractometer with a Genix
˚
Microfocus tube with mirror optics using MoKα radiation (λ = 0.71073 A) and
were scaled using the frame scaling procedure in the X-AREA program system
[23]. The structures were solved by direct methods using the program SHELXS
155 [24] and refined against F2 with full-matrix least-squares techniques using the
program SHELXL [24]. Data for picene/TCNQ were collected on a Nonius
KappaCCD four-circle diffractometer with graphite monochromated MoKα ra-
˚
diation (λ = 0.71073 A) and were processed with the DENZO/SCALEPACK
software [25]. The structure was solved by direct methods using the program
160 SHELXS [24] and refined against F2 with full-matrix least-squares techniques
using the program CRYSTALS [26].
The phenanthrene molecules in phenanthrene/TCNQ-F2 and phenanthrene/TCNQ-
F4 are disordered about a centre of inversion over two equally occupied positions.
In picene/TCNQ-F2, the F and H atoms bonded to the six-membered ring are
165 disordered over two positions with a site occupation factor of 0.835(3) for the
major occupied sites.
9

2.4. Spectroscopic analysis
Infrared spectroscopy was measured on a small amount of needle-like single
crystals with no preferred orientation of PHN/TCNQ-Fx and picene/TCNQ-Fx
170 and these were compared to spectra of powder samples of the pure acceptors
TCNQ, TCNQ-F2 and TCNQ-F4. The spectra were obtained by using a Nicolet
730 FT IR spectrometer between 4000 and 500 cm−1 at room temperature with
a resolution of 1 cm−1
.
3. Results and discussion
175 3.1. Crystal structure of charge transfer salts
X-ray single crystal diffraction studies detect a ratio of 1:1 for donor and ac-
ceptor molecules and formation of mixed stack geometry in all compounds. Typ-
ical pictures of single crystals and the corresponding crystal structures resulting
from single crystal X-ray diffraction are presented in Figure 5 for phenanthrene
180 and in Figure 6 for picene based CT salts.
b)
c)
a)
100µm
b
200µm
200µm
a
c
c
b
c
b
a
a
Figure 5: Single crystals and crystal structures of a) PHN/TCNQ; b) PHN/TCNQ-F2 and c)
PHN/TCNQ-F4. The phenanthrene molecules are disordered in all structures.
The crystal structure of phenanthrene/TCNQ-F2 is isostructural to the re-
ported for phenanthrene/TCNQ in Ref. 10, whereas the space group (SG)
¯
changes from P1 to P21/n for phenanthrene/TCNQ-F4. Both, donor and
acceptor molecules are planar and have similar dimensions. Especially, the
˚
185 longest axes of phenanthrene and TCNQ(-F2; -F4) are roughly equal with 9 A
˚
for phenanthrene and 8.4 A for TCNQ (8.6 A for TCNQ-F2; 8.75 A for TCNQ-
˚
˚
F4), respectively. The central part of TCNQ-Fx and the benzene ring in the
10

a)
b
250µm
b)
c
a
250µm
c)
250µm
Figure 6: Single crystals and crystal structures of a) PCN/TCNQ; b) PCN/TCNQ-F2 and c)
PCN/TCNQ-F4. TCNQ-F2 is disordered in the structure. The structures for these CT salts
are isostructural.
middle of PHN overlap with a little offset which is compensated by disorder,
due to a mirroring of PHN on the long axis of the molecule shown in Figure 7
190 reported also for PHN/TCNQ[10].
a)
b)
b
c
Figure 7: Disordering for PHN/TCNQ-F2 in a) left orientation and b) right orientation of the
phenanthrene molecule.
With a closer look to the stacking directions, the spacing becomes smaller for
˚
stronger acceptors as lattice parameter a decreases from 6.719 A in PHN/TCNQ
˚
˚
over 6.682 A in PHN/TCNQ-F2 to 6.632 A in PHN/TCNQ-F4. By increas-
ing the acceptor strength and geometry, the angle between donor and acceptor
¯
195 molecules and axis c in P1 (axis b in P21/n) becomes flatter. For picene/TCNQ
11

and picene/TCNQ-F2 the same space groups P21/n are revealed as for picene/TCNQ-
F4 reported in Ref. 11 and no disorder of the aromatic donor occurs in this
set. Instead of the donor, the acceptor in the picene/TCNQ-F2 structure is
disordered. Here, no differentiation between one ordering of 2,5-configuration
200 or mirroring at the bc-plane is possible. In picene/TCNQ-Fx no space group
changes are observed which can be due to different sizes of donor and acceptor
˚
in general (the longest axis is 13 A for picene). This is compensated by packing
in pairs of D and A in P21/n geometry as described in Ref. 11.
a)
b)
3.425 Å
3.490 Å
3.259 Å
3.737 Å
3.327 Å
3.475 Å
3.478 Å
3.947 Å
3.425 Å
3.259 Å
3.737 Å
3.327 Å
Figure 8: Distances between D and A in a) phenanthrene/TCNQ-F2 and b) picene/TCNQ
charge transfer salts and bending of donor and acceptor molecules.
A structural indication for charge transfer in theses materials is, that short-
205 est distances between e.g. atoms C16 and C14’ in phenanthrene/TCNQ-F2 or
atoms C6 and C33 in picene/TCNQ are not equal as shown in Figure 8 along
the stacking axis. This effect occurs in all presented charge transfer salts but is
the strongest for phenanthrene/TCNQ-F4 with the largest bending of the cyano
end groups. In general, the smallest C-C distance between donor and acceptor
˚
210 is in all cases below 3.5 A resembling a good π − π overlap in all structures
[27]. An overview of the crystal structure data is shown in Table 2 and in the
Supporting Information in more detail for each new CT complex.
3.2. Charge transfer by structural data analysis and infrared spectroscopy
Another more pronounced property of CT salts is the amount of charge
215 transfer going from donor to acceptor by formation of the CT compound. This
12

Table 2: Structural parameters for the new charge transfer salts phenanthrene/TCNQ-
F2, phenanthrene/TCNQ-F4, picene/TCNQ and picene/TCNQ-F2 together with the known
phenanthrene/TCNQ [10] and picene/TCNQ-F4 [11]. The value of charge transfer obtained
from structural changes in the acceptor molecule is shown in the last column.
CT
α [◦]
β [◦]
γ [◦]
C-C
ρ
[e−]
˚
˚
˚
c [A]
˚
[A]
shortest
SG
a [A] b [A]
6.719 8.086
6.682 8.078
struct.
¯
P1
PHN/TCNQ
9.118
9.277
101.21 97.8
101.66 96.31
99.46
3.325
-0.04[10]
0.05
¯
PHN/TCNQ-F2 P1
100.27 3.327
3.235
PHN/TCNQ-F4 P21/n 6.632 18.296 8.367
90
102.36 90
0.17
PCN/TCNQ
P21/n 7.975 7.104
41.187 90
42.177 90
42.941 90
92.88
91.64
91.43
90
90
90
3.259
3.284
3.296
0.0
PCN/TCNQ-F2 P21/n 7.905 7.081
PCN/TCNQ-F4 P21/n 7.920 7.000
0.09
0.19 [11]
value is relevant to classify CT salts and to understand their physical properties.
One way to extract the CT of a system is to compare the bond lengths of
the acceptor molecules. These changes of geometry inside the acceptor are a
first indication to detect the amount of CT and were investigated for a wide
220 range of TCNQ compounds [28]. The same approach was also used for TCNQ-
F2 [29] and TCNQ-F4 [11]. In all cases, the pure acceptor compound has a
quinoid structure. While charge transfer takes place, this quinoid structure in
the acceptor changes to a more aromatic structure, which means that bond c is
extended and bond b and d are shortened to become a more aromatic character.
225 This is shown in the literature for a variety of TCNQ salts. After Ref. 28, we
estimate the charge transfer by
αCT − α0
CT
struct.
ρ
=
(1)
α
−1 − α0
where the parameter α is defined by the bond length ratio c/(b+d) (see Fig.
1) and the indexes CT, 0 and -1 denote values for the CT compound, the pure
acceptor and the acceptor anion, respectively. With this approach, the charge
230 transfer was calculated for the presented compounds and shown in Table 2.
Both sets of different acceptors result in a larger transfer for stronger ac-
ceptors, which is in good agreement with a smaller spacing between donor and
13

acceptor in the structure and a larger overlap between orbitals of donor (D) and
acceptor (A).
235
Another method to estimate the amount of charge transfer is the investi-
gation with infrared spectroscopy [30, 31]. This method was used widely for
TCNQ and also in a few cases for TCNQ-F2 [29] and TCNQ-F4 [11]. For CT
salts based on TCNQ, the C≡N stretching modes are dominant in the spectra
and the shifting from pure acceptor to CT complex is prominent. Besides the
240 shift in C≡N vibration mode, also the C=C stretching mode of the acceptor is
a good indication for the amount of charge transfer [31]. The C≡N stretching
mode is easier to analyze, but as the cyano groups are situated at the end of the
acceptor molecule and can be influenced by surrounding molecules, this stretch-
ing reflects the CT not as accurate as the C=C vibration in the center of the
245 molecule. In Figure 9, we present IR spectra of CT crystals and compare these
with spectra on powder of the pure acceptor compounds in the C≡N stretching
mode region.
For the neutral acceptors, we observed a strong C≡N stretching mode at
2222.7 cm−1 for TCNQ, at 2229.5 cm−1 for TCNQ-F2 and at 2227.7 cm−1 for
250 TCNQ-F4 (orange arrows in Figure 9). According to Ref. 32 these bands can
be assigned to b3gν42 for TCNQ and to b1uν18 in TCNQ-F4.
Compared to the pure acceptors, the charge transfer compounds show no
additional peaks, besides the picene/TCNQ-F4 complex. The relative intensities
of the second peak around 2214 cm−1 increases compared to the large peak
255 around 2227.7 cm−1 as described in Mahns et al. [11]. All other compounds
show no changes in relative peak intensities. The shift of the C≡N stretching
mode is prominent in all measurements, despite for phenanthrene/TCNQ-F4
and are listed in Table 3. The degree of charge transfer of the shift in the C≡N
stretching mode in infrared absorption spectroscopy can be calculated from the
260 empirical relation [33]:
2Δν
IR
CN
ρ
=
(2)
ν12
ν02
ν0(1 −
)
14

a)
x4
PHN/TCNQ-F4
TCNQ-F4
PHN/TCNQ-F2
TCNQ-F2
PHN/TCNQ
TCNQ
b)
PCN/TCNQ-F4
TCNQ-F4
PCN/TCNQ
TCNQ
Figure 9: IR spectra of polycrystalline samples in the C≡N stretching mode regions for a)
phenanthrene/TCNQ-Fx (x=0, 2, 4) and b) picene/TCNQ-Fx (x=0, 4). The C≡N stretching
is always red-shifted in the case of charge transfer compounds, compared to pure acceptors.
Where Δν = ν0 − νCT and the subscripts 0, 1 and CT denote the C≡N
stretching modes of the pristine acceptor, the acceptor anion and the charge
transfer salt. A second method to detect the charge transfer is a linear ap-
proach between neutral acceptor and fully ionized acceptor with charge of 1 e−
265 comparing the wave numbers of the respective maximum. The wave num-
ber of the C≡N vibration for the fully ionized acceptor in alkaline salts is
−1
2186 cm−1 for TCNQ−1, 2188 cm−1 for TCNQ-F2 and 2190 cm−1 for TCNQ-
−1
F4 [32, 34, 31].
With linear approach, we receive a charge transfer of 0.22 e− for PHN/TCNQ,
15

270 0.28 e− for PHN/TCNQ-F2 and 0.05 e− for PHN/TCNQ-F4 when analyzing the
C≡N vibration. This is also in good agreement to results from eq. 2. For the
charge transfer salts based on picene, we obtain 0.20 e− for PCN/TCNQ and
0.16 e− for PCN/TCNQ-F4, which matches nicely the value of 0.14 e− in Ref. 11.
Especially, the results for phenanthrene/TCNQ-Fx are of interest as the charge
275 transfer first increases in the stronger acceptor TCNQ-F2 compared to TCNQ,
but decreases again for the strongest acceptor TCNQ-F4. The conclusion from
both sets of materials is, that the value of CT is lowest for the strongest ac-
ceptor, which is an unexpected result. This can be due to the shift in crystal
structure and the slightly different surrounding of the C≡N groups in the struc-
280 ture. As the C≡N groups are located at the very end of the acceptor molecules,
the shift of the IR mode is also strongly depending to neighboring molecules.
On the other side, this can be also an indication of a different localization of
the charge on the acceptor molecule depending on its fluorination. Therefore,
the C=C stretching mode outside the carbon ring in the acceptor needs to be
285 investigated. This shift is more independent to changes in the surrounding and
sensitive to another part of the acceptor molecule. These spectra are shown in
Figure 10. To estimate the CT, we used the linear approach, as discussed be-
fore. Therefore, the b1uν20 mode, identified as the C=C stretching mode, was
observed at 1542 cm−1 in neutral TCNQ, at 1549 cm−1 in neutral TCNQ-F2
290 and the b1uν19 mode at 1550 cm−1 in neutral TCNQ-F4 and at 1504 cm−1 in
the TCNQ anion [32] with total charge transfer. For the TCNQ-F4 anion the
b1uν19 mode is assigned to C=C stretching and is located at 1500 cm−1 [31]. As
TCNQ-F2 is in between TCNQ and TCNQ-F4, we expect the shift of the C=C
mode in the anion to be around 1502 cm−1
.
295
With this assumption, we yield a charge transfer of 0.01 e− for PHN/TCNQ,
0.04 e− for PHN/TCNQ-F2 and for PHN/TCNQ-F4. For the two picene com-
pounds, we determined a charge transfer of 0.04 e− for PCN/TCNQ and 0.08 e−
for PCN/TCNQ-F4. The results for CT strongly differ for C≡N and C=C
stretching. For the C≡N vibration modes, the CT decreases for stronger accep-
300 tors (especially for PHN/TCNQ-F4), whereas it increases for the C=C vibration
16

a)
PHN/TCNQ-F4
TCNQ-F4
TCNQ-F2
PHN/TCNQ-F2
PHN/TCNQ
TCNQ
1450
1500
1550
1600
1650
Wavenumber [cm-1]
b)
PCN/TCNQ-F4
TCNQ-F4
PCN/TCNQ
TCNQ
1450
1500
1550
1600
1650
Wavenumber [cm-1]
Figure 10: IR spectra in the region of different carbon-carbon stretching modes. The C=C
stretching mode outside the ring of the acceptor (orange arrows) is located around 1550 cm−1
(b1uν20 for TCNQ and b1uν19 for TCNQ-F4). The shift of C=C stretching mode is for a)
phenanthrene/TCNQ-Fx but also for b) picene/TCNQ-Fx rather weak, which means a small
amount of charge transfer.
modes with increasing acceptor strength. As described before, the C≡N vibra-
tion modes are located at the end of the acceptor and are probably influenced
CN
CT
CC
CT
by their surrounding. In addition, the difference between ρ
and ρ
can be
due to an inhomogeneous distribution of the CT over the acceptor molecule.
305 For increasing acceptor strength the contribution of CT might be shifted to the
center ring. Therefore, the tendency for the C=C stretching seems more rea-
sonable with a larger amount of CT for increasing acceptor strength although
the observed values are probably underestimated. The value of charge transfer
lies in all new compounds in the neutral region after definition in [35]. Addi-
17

Table 3: IR-frequencies of C≡N and C=C stretching modes of different charge transfer (CT)
salts and pristine acceptors (A) presented in the first two columns. In the third column the
difference between these two values Δν is shown. In the last two columns the charge transfer
resulting from linear approach is shown for both the C≡N and C=C stretchings.
CN
CC
νCT [cm−1
]
νA[cm−1
2222.7
1542
]
Δν
8.1
0.5
ρ
[e−]
ρ
[e−]
CT
CT
PHN/TCNQ
2214.6
0.22
1541.5
0.01
0.04
0.04
0.04
0.08
PHN/TCNQ-F2 2218
1547
2229.5
1549
11.5 0.28
2.0
PHN/TCNQ-F4 2225.8
1547.9
2227.7
1550
1.9
2.1
7.3
1.5
5.9
3.8
0.05
0.20
0.16
PCN/TCNQ
2215.4
1540.5
2222.7
1542
PCN/TCNQ-F4 2221.8
1546.2
2227.7
1550
310 tional investigation of the electronic structure of these novel complexes using
high sensitive spectroscopic measurements are now in progress to investigate the
distribution of CT on the acceptor molecule in more detail.
4. Conclusion
Single crystals of four new charge transfer complexes with phenanthrene
315 and picene as π-conjugated donors were obtained by physical vapor transport
method. Differential thermal analysis measurements detect the melting temper-
atures of the respective starting molecules, as well as for PHN/TCNQ. Pressure
dependent growth studies improved the crystal growth of these complexes re-
sulting in thicker crystals sufficient for single-crystal x-ray diffraction. The new
320 crystal structures expand the existing data of different acceptor strengths in
the phenanthrene and picene based salts and can be used as model systems to
investigate the influence of different acceptors strengths to the amount of charge
transfer. Based on these examples and the improvement of the physical vapor
transport technique, further novel materials can be designed.
18

325 5. Acknowledgement
The authors acknowledge financial support from of the DFG within the
SFB/TR49, the MainCampus program of the Stiftung Polytechnische Gesellschaft
Frankfurt am Main (A.M.), the Graduate School of Excellence Materials Science
in Mainz and by Centre for Complex Materials COMATT in Mainz (A.C.). We
330 thank F. Ritter, C. Klein and M. Baumgarten for fruitful discussion and K.-D.
Luther for technical support.
6. Supporting Information
Detailed crystal structure data for each new CT compound are shown in
the supporting information. CCDC files 1448360 (phenanthrene/TCNQ-F2),
335 CCDC 1448361 (phenanthrene/TCNQ-F4), CCDC 1448362 (picene/TCNQ),
and CCDC 1448363 (picene/TCNQ-F2) contain the supplementary crystallo-
graphic data for this paper and can be obtained free of charge from the Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data request/cif.
References
340
345
350
[1] N.
Thorup,
G.
Rindorf,
H.
Soling,
K.
Bechgaard,
The structure of di(2,3,6,7-tetramethyl-1,4,5,8-tetraselenafulvalenium) hexafluorophosphate, (tmtsf)2pf6,
Acta Cryst. B 37 (6) (1981) 1236–1240.
URL http://dx.doi.org/10.1107/S0567740881005566
[2] T. Kistenmacher, T. Philips, D. Cowan, Acta Cryst. B 30 (1974) 763.
[3] N. Toyota, M. Lang, J. Mu¨ller, Low-Dimensional Molecular Metals,
Springer-Verlag, Berlin, 2007.
[4] V. Podzorov, V. Pudalov, M. Gershenson, Appl. Phys. Lett. 82 (2003) 1739.
[5] P. Lunkenheimer, J. Mu¨ller, S. Krohns, F. Schrettle, A. Loidl, B. Hart-
mann, R. Rommel, M. de Souza, C. Hotta, J. A. Schlueter, M. Lang,
Multiferroicity in an organic charge-transfer saltaˆthat is suggestive of electric-dipole-driven magnetism,
19

Nat. Mater. 11 (9) (2012) 755–758.
URL http://dx.doi.org/10.1038/nmat3400
[6] X. F. Wang, R. H. Liu, Z. Gui, Y. L. Xie, Y. J. Yan, J. J. Ying, X. G. Luo,
X. H. Chen, Superconductivity at 5 k in alkali-metal-doped phenanthrene,
Nat. Commun. 2 (2011) 507.
355
360
365
URL http://dx.doi.org/10.1038/ncomms1513
[7] R. Mitsuhashi, Y. Suzuki, Y. Yamanari, H. Mitamura, T. Kambe, N. Ikeda,
H. Okamoto, A. Fujiwara, M. Yamaji, N. Kawasaki, Y. Maniwa, Y. Kubo-
zono, Superconductivity in alkali-metal-doped picene, Nature 464 (7285)
(2010) 76–79.
URL http://dx.doi.org/10.1038/nature08859
[8] M. Xue, T. Cao, D. Wang, Y. Wu, H. Yang, X. Dong, J. He, F. Li, G. F.
Chen, Superconductivity above 30 k in alkali-metal-doped hydrocarbon,
Scientific Reports 2 (2012) 389.
URL http://dx.doi.org/10.1038/srep00389
[9] S.
Heguri,
M.
Kobayashi,
K.
Tanigaki,
Questioning the existence of superconducting potassium doped phases for aromatic hydrocarbons,
Phys. Rev. B 92 (1) (2015) 014502.
URL http://link.aps.org/doi/10.1103/PhysRevB.92.014502
370 [10] M. A. Dobrowolski, G. Garbarino, M. Mezouar, A. Ciesielski, M. K. Cyran-
ski, Structural diversities of charge transfer organic complexes. focus on benzenoid hydrocarbons and 7,7,
CrystEngComm 16 (2014) 415–429. doi:10.1039/C3CE41703D.
URL http://dx.doi.org/10.1039/C3CE41703D
[11] B. Mahns, O. Kataeva, D. Islamov, S. Hampel, F. Steckel, C. Hess,
375
M. Knupfer, B. Bu¨chner, C. Himcinschi, T. Hahn, R. Renger, J. Kortus,
Crystal growth, structure, and transport properties of the charge-transfer salt picene/2,3,5,6-tetrafluoro-7
Cryst.
1346.
Growth
Des.
14
(3)
(2014)
1338–
arXiv:http://dx.doi.org/10.1021/cg401841n,
20

doi:10.1021/cg401841n.
380
385
390
395
400
405
URL http://dx.doi.org/10.1021/cg401841n
[12] Y. Kaji, R. Mitsuhashi, X. Lee, H. Okamoto, T. Kambe,
N. Ikeda, A. Fujiwara, M. Yamaji, K. Omote, Y. Kubozono,
High-performance {C60} and picene thin film field-effect transistors with conducting polymer electrodes
Organic
doi:http://dx.doi.org/10.1016/j.orgel.2009.01.006.
URL http://www.sciencedirect.com/science/article/pii/S1566119909000081
Electronics
10
(3)
(2009)
432
–
436.
[13] P.
Crystal growth, homo-lumo engineering and charge transfer degree in perylene-fxtcnq (x=1, 2, 4) organic
Cryst. Growth Des. (ja) (0) null.
Hu,
K.
Du,
F.
Wei,
H.
Jiang,
C.
Kloc,
0
arXiv:http://dx.doi.org/10.1021/acs.cgd.5b01675,
doi:10.1021/acs.cgd.5b01675.
URL http://dx.doi.org/10.1021/acs.cgd.5b01675
[14] T. J. Emge, M. Maxfield, D. O. Cowan, T. J. Kistenmacher,
Solution and solid state studies of tetrafluoro-7,7,8,8-tetracyano-p-quinodimethane, tcnqf4. evidence for lo
Mol.
Cryst.
Liq.
Cryst.
65
(3-4)
(1981)
161–178.
arXiv:http://dx.doi.org/10.1080/00268948108082132,
doi:10.1080/00268948108082132.
URL http://dx.doi.org/10.1080/00268948108082132
[15] C.
Physical vapor growth of centimeter-sized crystals of -hexathiophene,
J. Cryst. Growth 182 (3-4) (1997) 416
doi:http://dx.doi.org/10.1016/S0022-0248(97)00370-9.
URL http://www.sciencedirect.com/science/article/pii/S0022024897003709
Kloc,
P.
Simpkins,
T.
Siegrist,
R.
Laudise,
–
427.
[16] R.
Laudise,
C.
Kloc,
P.
Simpkins,
T.
Siegrist,
454.
Physical vapor growth of organic semiconductors,
J.
Cryst.
Growth
187
21
(1998)
449
–

doi:http://dx.doi.org/10.1016/S0022-0248(98)00034-7.
URL http://www.sciencedirect.com/science/article/pii/S0022024898000347
[17] F. Rosenberger, J. Ouazzani, I. Viohl, N. Buchan, Physical vapor transport
revisited, J. Cryst. Growth 171 (1997) 270–287.
410
415
420
[18] Y.
mura,
Yoshida,
Y.
Y.
Kumagai,
M.
Mizuno,
Kishida,
K.
Iso-
Nakamura,
H.
G.
Saito,
Improved dynamic properties of charge-transfer-type supramolecular rotor composed of coronene and f4t
Cryst. Growth Des. 15 (11) (2015) 5513–5518.
arXiv:http://dx.doi.org/10.1021/acs.cgd.5b01138,
doi:10.1021/acs.cgd.5b01138.
URL http://dx.doi.org/10.1021/acs.cgd.5b01138
[19] C. Goldmann, S. Haas, C. Krellner, K. P. Pernstich, D. J. Gundlach,
B. Batlogg, Hole mobility in organic single crystals measured by a “flip-
crystal” field-effect technique, J. Appl. Phys. 96 (4) (2004) 2080–2086.
[20] K. Ueberreiter, H.-J. Orthmann, Z. Naturforsch. 5 (2) (1950) 101–108.
[21] D. S. Acker, W. R. Hertler, Substituted quinodimethans. i. preparation and chemistry of 7,7,8,8-tetracyano
J.
Am.
Chem.
Soc.
84
(17)
(1962)
3370–3374.
arXiv:http://dx.doi.org/10.1021/ja00876a028,
doi:10.1021/ja00876a028.
425
URL http://dx.doi.org/10.1021/ja00876a028
[22] O. Kataeva, M. Khrizanforov, Y. Budnikova, D. Islamov, T. Burganov,
A. Vandyukov, K. Lyssenko, B. Mahns, M. Nohr, S. Hampel, M. Knupfer,
Crystal growth, dynamic and charge transfer properties of new coronene charge transfer complexes,
430
Cryst.
Growth
Des.
16
(1)
(2016)
331
–
338.
arXiv:http://dx.doi.org/10.1021/acs.cgd.5b01301,
doi:10.1021/acs.cgd.5b01301.
URL http://dx.doi.org/10.1021/acs.cgd.5b01301
22

[23] S. . Cie, X-area (version 1.18) and x-red32 (version 1.04), Stoe & Cie,
Darmstadt, Germany.
435
[24] G. Sheldrick, A short history of shelx, Acta Cryst. A 64 (1) (2008) 112–
122.
URL http://dx.doi.org/10.1107/S0108767307043930
[25] Z. Otwinowski, W. Minor, C. C. W Jr, Processing of x-ray diffraction data
collected in oscillation mode, Method. Enzymol.
440
[26] P. W. Betteridge, J. R. Carruthers, R. I. Cooper, K. Prout, D. J. Watkin,
Crystals version 12: software for guided crystal structure analysis,
Appl. Crystallogr. 36 (6) (2003) 1487.
J.
URL http://dx.doi.org/10.1107/S0021889803021800
445 [27] M. Schwoerer, H. C. Wolf, Organische Molekulare Festk¨orperphysik, Wiley-
VCH, 2005.
[28] T. Kistenmacher, T. Emge, A. Bloch, D. Cowan, Acta Cryst. B38 (1982)
1193.
[29] T. J. Emge,
F. M. Wiygul,
J. S. Chappell,
A. N.
450
Bloch, J. P. Ferraris, D. O. Cowan, T. J. Kistenmacher,
Crystal structures for the electron donor dibenzotetrathiafulavalene, dbttf, and its mixed-stack charge-tra
Mol.
Cryst.
Liq.
Cryst.
87
(1-2)
(1982)
137–161.
arXiv:http://dx.doi.org/10.1080/00268948208083778,
doi:10.1080/00268948208083778.
455
URL http://dx.doi.org/10.1080/00268948208083778
[30] J.
den,
S.
Chappell,
Maxfield,
A.
T.
N.
O.
Bloch,
W.
D.
A.
O.
Bry-
M.
Poehler,
Cowan,
Degree of charge transfer in organic conductors by infrared absorption spectroscopy,
J. Am. Chem. Soc. 103 (9) (1981) 2442–2443.
460
arXiv:http://dx.doi.org/10.1021/ja00399a066,
23

doi:10.1021/ja00399a066.
URL http://dx.doi.org/10.1021/ja00399a066
[31] M. Meneghetti, C. Pecile, Charge-transfer organic crystals: Molecular vi-
brations and spectroscopic effects of electron-molecular vibration coupling
of the strong electron acceptor tcnqf4, J. Chem. Phys. 84 (8) (1986) 4149–
4162.
465
470
475
480
485
[32] R.
Influence of the intermolecular charge transfer interaction on the solution and solid state infrared spectra
J. Chem. Soc., Faraday Trans. 74 (1978) 235–248.
doi:10.1039/F29787400235.
URL http://dx.doi.org/10.1039/F29787400235
Bozio,
I.
Zanon,
A.
Girlando,
C.
Pecile,
2
[33] J. G. Robles-Mart´ınez, A. Salmer´on-Valverde, J. Argu¨elles-Ram´ırez,
A. Zehe, Variation of charge transfer in zn-naphtholimines with tcnq,
Molecular
doi:10.1023/A:1008336505281.
URL http://dx.doi.org/10.1023/A%3A1008336505281
Engineering
8
(4)
(1999)
411–417.
[34] Y.
S.
Takahashi,
Yokokura,
K.
Hayakawa,
H.
K.
Takayama,
J.
Harada,
Hasegawa,
T.
Inabe,
Charge conduction properties at the contact interface between (phthalocyaninato)nickel(ii) and electron a
Chem. Mater. 26 (2) (2014) 993–998.
arXiv:http://dx.doi.org/10.1021/cm403033b,
doi:10.1021/cm403033b.
URL http://dx.doi.org/10.1021/cm403033b
[35] J. B. Torrance”, Discovery of a neutral-to-ionic phase transition in
organic materials, Physical Review Letters 46 (4) (1981) 253–257.
doi:10.1103/PhysRevLett.46.253.
24