University of Groningen

Organic field-effect transistors for sensing applicationsMaddalena, Francesco

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2011

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Maddalena, F. (2011). Organic field-effect transistors for sensing applications Groningen: s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 14-04-2018

CHAPTER 5

63

Chapter 5

Doping Kinetics of Organic Semiconductors

Investigated by Field-Effect Transistors

Abstract

The kinetics of acid doping of the semiconductor regio-regular poly-3-hexylthiophene with vaporized chlorosilane has been investigated using field-effect transistors. The dopant density has been derived as a function of temperature and exposure time from the shift of the pinch-off voltage, being the gate bias where current starts to flow. The doping kinetics are perfectly described by empirical stretched exponential time dependence with a saturation dopant density of 1 ± 0.5 ×1026 m–3 and a thermally activated relaxation time. We show that a similar relationship holds for previously reported kinetics of poly-thienylene-vinylene doped with molecular oxygen

* Part of this work was published as F. Maddalena, E. J. Meijer, K. Asadi, D. M. de Leeuw

and P. W. M. Blom, Applied Physics Letters 97, 043302 (2010).

DOPING KINETICS OF ORGANIC SEMICONDUCTORS INVESTIGATED BY FIELD-EFFECT TRANSISTORS

64

5.1 Introduction

Organic electronics is well-established in the fields of light-emitting diodes,

displays and solar cells [1, 2, 3, 4]. Presently, organic field-effect transistors

(OFETs) are being developed for low-end high-volume applications such as smart

labels and electronic barcodes, [5] and non-volatile memories [6, 7, 8].

Furthermore, OFETs are being investigated as possible gas sensors [9, 10, 11].

Upon exposure to gasses changes have been observed in key transistor parameters

such as threshold voltage and field-effect mobility [12, 13]. The most dominant

change however is a change in the conductance due to chemical doping of the

semiconductor. A gas sensor could be envisaged by monitoring the increase in

conductance upon exposure to the target analyte [14].

Figure 1. Transfer characteristics of a field effect transistor (L/W= 10/10000, VDRAIN= -2 V) in

vacuum and after 1 minute exposure to HCl vapour.

To fabricate a reliable gas sensor it is also important to understand the kinetics

of the doping process. However, the number of studies on the doping kinetics is

limited. Most studies focus on the conductivity of organic semiconductors as a

function of dopant density [15, 16]. The increase of the conductivity with time has

only been investigated for regio-regular poly-3-hexylthiophene (rr-P3HT) and

CHAPTER 5

65

poly-thienylenevinylene (PTV) exposed to molecular oxygen [17]. From analysis

of the transfer curves the dopant density and bulk mobility could be derived as a

function of exposure time. The study was performed to analyze the shelf-life of

organic transistors.

In order to understand the kinetics of the doping of organic semiconductors, we

doped regio-regular poly-3-heylthiophene (rr-P3HT) in field-effect transistors with

gaseous HCl. At the partial vapor pressures used however, the doping process was

much too fast to reliably determine the kinetics. Figure 1 shows that the polymer

was fully doped in a matter of a minutes.

To expand the time scale and have a more controlled doping of the

semiconductor, instead of HCl we used vaporized trichloro-(1H, 1H, 2H, 2H)-

perfluorooctylsilane (TCFOS). This molecule is reported to yield acid doping of

conjugated molecules [14]. The doping process of rr-P3HT with TCFOS is

sufficiently slow to measure the doping kinetics. The dopant density is derived

from the transfer curves as a function of temperature and exposure time. We show

that the kinetics can phenomenologically be described with stretched exponential

time dependence and that the same relationship holds for previously reported

doping studies [17].

5.1.1 Doping in organic semiconductors

Applying a gate bias more positive than the threshold voltage to a p-type

transistor will deplete any mobile holes present in the semiconductor layer, turning

the transistor in the off–state. This can be best observed from the transfer

characteristics of the transistor. Through the logarithm of the current against the

gate voltage one can determine the exact point when the channel current starts to

flow rising from the leakage current. Such point it called the ‘pinch–off voltage’

VPINCH. In an undoped or negligibly doped semiconductor VPINCH is equal to the

switch–on voltage, VSO, the bias point at which the transistor intrinsically switches

on if no significant dopants or traps are present in the material.

Introducing chemical dopants into the semiconducting layers will shift the

pinch–off voltage higher values, since a higher bias is needed to deplete the

semiconductor. The cartoons in Figure 2 show schematically the process of

depletion in a doped p-type material. A doped semiconductor will have two forms

of conduction. First the common field-effect conduction through the channel

formed by the gate through accumulation of charges at the semiconductor-insulator

DOPING KINETICS OF ORGANIC SEMICONDUCTORS INVESTIGATED BY FIELD-EFFECT TRANSISTORS

66

interface. Second the bulk conduction through the mobile charges in the bulk of the

semiconductor. The latter are formed not by field-effect but by doping of the

material. If the transistor is in accumulation (Figure 2 a) both conduction processes

will occur. However, since the concentration of the charges is usually far higher in

the channel than in the bulk, hence also the mobility, the contribution of the bulk

conduction is minimal. To really observe doping and bulk conduction one must go

into the depletion region of the transistor, as seen in Figure 2 b, where the channel

conduction is switched off and only bulk conduction is present. The effect of the

dopant will appear as an extra ‘shoulder’ in the transfer characteristics of the OFET

(see Figure 3). Increasing the bias to sufficiently high positive biases will

eventually also deplete the charge carriers in the bulk, turning the device

completely off (Figure 2 c).

Figure 2. Schematic view of the charge depletion in a doped p-type OFET. In this device holes are

present in the bulk, caused by the dopands. a.) When the gate bias is negative (VGATE < VSO), then the

gate will accumulate positive charges at the insulator-semiconductor interface forming a conductive

channel. The current, between source and drain, flowing through the channel will add to the current

flowing through the doped bulk. b.) By applying a positive gate bias (VGATE > VSO) no accumulation

will occur at the semiconductor-insulator interface. Only conduction through the doped bulk will

occur. c.) If the gate bias is positive enough (VGATE > VPO) then also the mobile charges caused by

doping will be depleted and no current will flow between the source and the drain.

CHAPTER 5

67

5.2 Results and Discussion

To determine the doping kinetics of our semiconductor, the transfer curves of

OFETs were measured as a function of temperature and exposure time. The

measurement chamber was evacuated to less than 10–4 mbar and vaporized TCFOS

was injected into the measurement chamber (p ~2.5·10–2

mbar) at t= 0 s as

described in the experimental section of Chapter 2. Without TCFOS the transfer

curve does not change with time; stress can be disregarded on the time scale of the

experiment.

Figure 3. a.) Transfer curves in the linear regime of a rr-P3HT organic field-effect transistor in

vacuum and after exposure to TCFOS vapor at room temperature for different exposure times. The

channel length and width are 10 µm and 10000 µm respectively. B.) The transfer curves corrected for

the threshold voltage shift with respect to the pristine undoped transistor at t=0.

The transfer curves at different times of exposure to TCFOS at room

temperature are presented in Figure 3 a. The curves shift towards positive gate bias

with exposure time. Furthermore a shoulder appears and the on-current in

accumulation slightly increases. We note that the off-current in depletion increases

as well, which is due to increased parasitic leakage currents outside of the defined

transistor area caused by the use of unshielded drain electrodes [17]. The analysis

of the shift of the transfer curves towards positive gate bias with exposure time is

complicated by the fact that two effects have to be disentangled. First, the doping

process can lead to the formation of charged states at the interface between the

insulator and the semiconductor. This will lead to a shift of the threshold voltage.

Furthermore, upon exposure to TCFOS the rr-P3HT gets chemically doped and a

higher positive bias is needed to deplete the bulk semiconductor. To get a

DOPING KINETICS OF ORGANIC SEMICONDUCTORS INVESTIGATED BY FIELD-EFFECT TRANSISTORS

68

quantitative result for the threshold voltage shift between the measurements we

linearly extrapolated the current as a function of the gate bias for each

measurement on a linear scale and determined from the intercept the threshold

voltage. The transfer curves corrected for the threshold voltage shift are presented

in Figure 3 b and show a cross-over from an accumulation mode into a bulk

depletion mode transistor. The current in accumulation, at negative bias, is

dominated by the channel current while in depletion, at positive bias, the current is

mainly flowing through the bulk semiconductor. The remaining shift of the transfer

curves now arises from the doping and is characterized by a shift in the pinch-off

voltage. At the pinch-off voltage, defined as the gate bias where current starts to

flow, the bulk semiconductor is fully depleted. When the gate bias is sufficient to

enable charge accumulation in the channel the current increases superlinearly with

gate bias, which is manifested by the shoulder in the transfer curve. This

superlinear increase is due to the fact that the charge carrier mobility in the channel

is much larger than in the bulk. This difference in charge carrier mobility originates

from the fact that the mobility is charge density dependent [18] and the charge

density in the channel is much larger than in the bulk of the semiconductor.

The dopant density can be determined from the pinch-off voltage. We assume

that the dopants are homogeneously distributed in the bulk semiconductor.

Furthermore for a small source drain bias we assume that the depletion of the

semiconductor takes place uniformly over the entire channel length. The width of

the depletion layer is then given by [17, 19]:

−

−+= 1

)(21

0

2

0

SA

ThGatei

i

S

DeplqN

VVC

CW

εε

εε (5.1)

where ε0 is the permittivity in vacuum, εS the relative dielectric constant of the

semiconductor, q the elementary charge, Ci the capacitance of the gate dielectric

per unit area, and where NA is the dopant density. Exactly at the pinch-off voltage

the whole film is just depleted from charge carriers. The width of the depletion

layer then is equal to the thickness of the semiconductor layer dS [18]. The

CHAPTER 5

69

Figure 4. a.) Dopant density versus exposure time at different temperatures for rr-P3HT doped with

silane (TCFOS) vapor. The solid lines are a fit to the data with a stretched exponential time and

temperature dependence. b.) Relaxation τ as a function of reciprocal temperature, extracted from the

fit. As expected for the relaxation time of a stretched exponential relationship, τ has an exponential

dependence with the inverse absolute temperature.

DOPING KINETICS OF ORGANIC SEMICONDUCTORS INVESTIGATED BY FIELD-EFFECT TRANSISTORS

70

insulator and semiconductor capacitances per unit area are given by Ci = ε0εi/di,

and Cs= ε0εS/ds where di and ds are the thicknesses of the gate dielectric and the

semiconductor respectively, and where εi is the relative dielectric constant of the

gate dielectric. The dopant density can now be recalculated from Eq. 5.1 as [17]:

+

=

I

IS

S

S

PO

Addd

q

VN

εε

ε

2

22

0 (5.2)

The dopant densities calculated with Eq. 5.2 using the pinch-off voltages

extracted from Figure 3 are presented in Figure 4 a on a double logarithmic scale as

a function of time. Values for the dopant densities derived at other temperatures are

included as well. Figure 4 a shows that the dopant density increases with time as a

power law. The exponent however, increases with temperature. This dependence

suggests that the time and temperature dependence can phenomenologically be

described by a single stretched exponential:

−−∞→=

β

τ

ttNtN AA exp1)()( (5.3)

where NA(t→∞) is the saturation dopant density, β is a dispersion parameter

typically equal to T/T0, with T being the absolute temperature and T0 a

characteristic temperature, and τ being the average relaxation time. The stretched-

exponential dependence is observed in a wide class of disordered systems, such as

relaxation of glasses toward equilibrium, dielectric relaxation in a charge-density-

wave system and dispersive transport in photoconductors [20]. It occurs when

there is an anomalously large distribution of characteristic times involved, and it

holds for all dispersive transport processes in an exponential distribution of trap

states [21]. The solid lines in Figure 4 a represent a simultaneous fit to all the data

points. A good agreement is obtained with NA(t→∞) = 1 ± 0.5 ×1026 m–3 and T0 =

690 ± 40 K. The value for the saturation dopant density derived indicates that about

10% of the thiophene units can be chemically oxidized. This number is in good

agreement with the maximum doping level derived from electrochemical studies;

in polythiophene about one out of every four monomeric units can be oxidised

[22]. Figure 4 b shows that the relaxation time, τ, is thermally activated by [23]:

= −

Tk

EExp

B

ACT1

0υτ (5.4)

CHAPTER 5

71

where ν0 is a frequency prefactor, EACT is the activation energy of the doping

process (or the slowest step in the process) and kB is the Boltzmann constant. The

activation energy of 0.6 eV, or about 58 kJ·mol–1, agrees with reported values for

protonation reactions of organic molecules [24, 25, 26, 27], suggesting acid doping

of rr-P3HT by TCFOS. Reported values for ν span many orders of magnitude [28].

Here we derived a value for the prefactor of about 1 Hz, which cannot yet be

explained.

The doping of rr-P3HT was investigated at different partial pressures of

TCOFS, viz. 2.5·10–2 and 2.5·10–1 mbar. The main effect is a decrease of the

relaxation time with increase of TCOFS partial vapor pressure, indicating that

increasing the pressure of the dopant increases the speed of the doping of the

semiconductor.

Figure 5. Dopant density versus time at different temperatures for PTV doped with molecular

oxygen at different pressures.17 The solid lines are the fit with a stretched exponential function.

Doping of poly-thienylene-vinylene (PTV) exposed in field-effect transistors to

molecular oxygen has previously been reported [17]. Similar to this study the

dopant density has been derived as a function of the partial oxygen pressure and

exposure time, but the doping kinetics were not further analyzed. Here, we replot

the reported values in Figure 5; the fully drawn curves are a stretched exponential

fit to the data using Eq. 5.3. Also for the oxygen doping of PTV a good agreement

is obtained. For the fit we used the same maximum dopant density, NA(t→∞), as

derived above for rr-P3HT. The characteristic temperature T0 follows directly from

DOPING KINETICS OF ORGANIC SEMICONDUCTORS INVESTIGATED BY FIELD-EFFECT TRANSISTORS

72

the slope of dopant density versus time and amounts to T0 = 1700 ± 30 K. The

dopant density has not been reported as a function of temperature in the PTV study.

Hence we can determine the average relaxation time τ but not the activation

energy. We find that τ is inversely proportional to the oxygen pressure, 4.1

2~ −

Opτ .

When we assume that the activation energy does not depend on oxygen pressure,

the relation between τ and pressure can then be rationalized by a prefactor that

increases with oxygen pressure.

The electrical transport in organic semiconductors is dominated by thermally

activated hopping between localized states at the Fermi level [29, 30]. When the

effect of doping on the charge transport is only governed by filling up of transport

states with the charge carriers that are released from the dopants, it can be expected

that the characteristic temperature derived from the doping study would be similar

to the isokinetic temperature characterizing the electrical transport states. However,

the differences obtained, viz. 690 K and 425 K for rr-P3HT and 1700 K and 380 K

[31] for PTV, indicates that chemical doping creates new states in the band gap, as

confirmed by numerous optical absorption investigations.

5.3 Conclusions

In summary, we have investigated acid doping of rr-P3HT with silane TCFOS

vapor in field-effect transistors. The transfer curves shift towards positive gate bias,

into depletion, upon exposure time. The dopant density has been derived from the

shift in pinch-off voltage as a function of temperature and exposure time. The

dopant density kinetics can phenomenologically be described by stretched

exponential time dependence. The doping is thermally activated with activation

energy of about 0.6 eV, which agrees with reported values for protonation reactions

of organic molecules. Reinterpretation of reported doping densities as a function of

time for PTV doped with molecular oxygen showed the same relationship. The

good agreements obtained indicate that the doping kinetics of organic

semiconductors follows stretched exponential time dependence.

CHAPTER 5

73

References 1 J. Lee, J. Lee and H. Yong Chu, Synth. Metals 159, 991 (2009).

2 B. P. Rand, C. Girotto, A. Mityashin, A. Hadipour, J. Genoe, and P. Heremans,

Appl. Phys. Lett. 95, 173304 (2009).

3 M. Suzuki, H. Fukagawa, Y. Nakajima, T. Tsuzuki, T. Takei, T. Yamamoto and

S. Tokito, J. Soc. Inf. Display 17, 1037 (2009).

4 M. Pagliaro, R. Ciriminna and G. Palmisano, ChemSusChem 1, 880 (2008).

5 E. Cantatore, T.C. T. Geuns, G. H. Gelinck, E. van Veenendal, A. F. A.

Gruijthuijsen, L. Schrijnemakers, S. Drews and D. M. de Leeuw , J . Solid-St. Circ.

42, 84 (2007).

6 K. Asadi, D. M. de Leeuw, B. de Boer and P. W. M. Blom, Nat. Mater. 7, 547

(2008).

7 M. Debucquoy, M. Rockele, J. Genoe, G. H. Gelinck and P. Heremans, Org.

Electron. 10, 1252 (2009).

8 D. M. de Leeuw and E. Cantatore, Mat. Sci. Semicon. Proc. 11, 199 (2008).

9 A. Caboni, E. Orgiu, M. Barbaro and A. Bonfiglio, IEEE Sens. J. 9, 1963 (2009).

10 A. K. Diallo, J. Tardy, Z. Q. Zhang, F. Bessueille, N. Jaffrezic-Renault and M.

Lemiti, Appl. Phys. Lett. 94, 263302 (2009).

11 S. M. Goetz, C. M. Erlen, H. Grothe, B. Wolf, P. Lugli and G. Scarpa, Org.

Electron. 10, 573 (2009).

12 A. Das, R. Dost, T. H. Richardson, M. Grell, D. C. Wedge, D. B. Kell, J.J.

Morrison and M. L. Turner, Sensor Actuat. B-Chem. 137, 586 (2009).

13 T. Toccoli, A. Pallaoro, M. Tonezzer, N. Coppede and S. Iannotta, Solid State

Electron. 52, 417 (2008).

14 C. Y. Kao, B. Lee, L. S. Wielunski, M. Heeney, I. McCulloch, E. Garfunkel, L.

C. Feldman, and V. Podzorov, Adv. Funct. Mater. 19, 1906 (2009).

15 C. K. Chan, W. Zhao, S. Barlow, S. Marder and A. Kahn, Org. Electron. 9, 575

(2008).

16 L. Li, G. Meller and H. Kosina, J. Appl. Phys. 101, 033716 (2007).

DOPING KINETICS OF ORGANIC SEMICONDUCTORS INVESTIGATED BY FIELD-EFFECT TRANSISTORS

74

17 E. J. Meijer, C. Detcheverry, P. J. Baesjou, E. van Veenendaal, D. M. de Leeuw

and T. M. Klapwijk, J. Appl. Phys. 93, 4831 (2003).

18 W. F. Pasveer, J. Cottaar, C. Tanase, R. Coehoorn, P. A. Bobbert, P. W. M.

Blom, D. M. de Leeuw and M. A. J. Michels, Phys. Rew. Lett. 94, 206601 (2005).

19 S.M. Sze, Physics of Semiconductor Devices, Wiley, New York (1981).

20 J. Kakalios, R. A. Street and W. B. Jackson, Phys. Rev. Lett. 59, 1037 (1987).

21 Y. F. Chen and S. F. Huang, Phys. Rev. B 44, 13775 (1991).

22 G. Tourillon, Polyhtiophene and its derivatives, page 293, in Handbook of

Conducting Polymers, Volume 1, edited by T.A. Skotheim, Marcel Dekker

(1986).

23 R. S. Crandall, Phys. Rev. B 43, 4057 (1991).

24 R. D. Bach, C. Canepa, J. E. Winter and P. E. Blanchette, J. Org. Chem. 62,

5191 (1997).

25 P. Nama, R. Flammang, H. T. Lea, P. Gerbaux and M. T. Nguyena, Int. J. Mass

Spectrom. 228, 151 (2003).

26 M. C. Burland, T. Y. Meyer and M. Baik, J. Org. Chem. 69, 6173 (2004).

27 J. Catalán and J. L. G. de Paz, J. Chem. Phys. 123, 114302 (2005).

28 W. B. Jackson, J. M. Marshall and M. D. Moyer, Phys Rev. B 39, 1164 (1989).

29 A. Miller and E. Abrahams, Phys. Rev. 120, 745 (1960).

30 H. Bässler. Phys. Stat. Sol. B 175, 15 (1993).

31 E. J. Meijer, D. B. A. Rep, D. M. de Leeuw, M. Matters, P. T. Herwig and T. M.

Klapwijk, Synth. Metals 121, 1351 (2001).