Content uploaded by Marianna Ambrico
Author content
All content in this area was uploaded by Marianna Ambrico on Sep 04, 2014
Content may be subject to copyright.
NEW ROUTES TOWARDS MELANIN BIO-POLYMER BASED DEVICES
M. Ambrico*1, P. F. Ambrico1, A. Cardone2, T. Ligonzo3, S. R. Cicco2, A. Lavizzera3, V.
Augelli3 and G. M. Farinola4
1 CNR-Istituto di Metodologie Inorganiche e dei Plasmi, Sezione Territoriale di Bari
Via Orabona 4, 70125 Bari (Italy)
2 CNR-Istituto di Chimica dei Composti OrganoMetallici-UOS di Bari
Via Orabona 4, 70125 Bari (Italy)
3 Dipartimento Interateneo di Fisica, Università degli Studi di Bari “Aldo Moro”
Via Orabona 4, 70125 Bari (Italy)
4 Dipartimento di Chimica, Università degli Studi di Bari “Aldo Moro”
via Orabona 4, 70125 Bari (Italy)
ABSTRACT
The integration of biopolymers into hybrid electronic devices is one of the up to date issues in
view of the achievement of bio-compatible devices. Among ‘hot topics’ in bio-polymer research,
synthetic melanin or, briefly, “melanin”, has been recently recognized as a quite intriguing
macromolecule thanks to its multifunctional optoelectronic properties. Compared with other
polymers, up to now melanin transport properties have been mainly enlightened on pellets,
while optical absorption and conductivity properties have been investigated on melanin layers
deposited on quartz and ITO/glass substrates. Further optoelectronic features, could not be
investigated up to now, due to the unavailability of suitable procedures for melanin layer
deposition onto silicon substrates. The reason stems basically on the difference between the
hydrophilic nature of the melanin and the hydrophobic one of the silicon, that prevents adequate
melanin layers self assembling. However, we recently solve this issue and we were able to tailor
a melanin based metal/insulator/silicon structure, where synthetic melanin was embedded as the
insulating part. This allowed to disclose interesting features related to data storage capabilities
of melanin layers deposited both on ITO/glass and silicon, never investigated so far.
In this work we give an overview on the above mentioned results, and particular attention will be
paid on structure on silicon substrates. The use of pSi and nSi substrates and measurements
under different environment conditions has enabled to gain insight into ambipolar electrical
transport mechanisms, still unexplored. These results constitute a first important basic insight
into melanin/Si interface and represent a significant step towards the integration of melanin-
based bio-polymers in several kinds of hybrid organic polymer-based devices.
INTRODUCTION
The integration of polymeric organic semiconductors into Silicon-based electronic devices is one
of the up to date items in view of the achievement of effective multifunctional systems. This
mainly because of the possibility to combine the electrical and optical properties of
semiconductors with the structural versatility and processing/mechanical features typical of
organic polymeric materials. Among then, one of the hot topics concern a new ‘polymer’, the
synthetic melanin (SM) or, briefly, “melanin” recently recognized as an intriguing
macromolecule with multifunctional optoelectronic properties.[1-10] However, melanin
electrical transport properties have been enlightened so far mainly on pellets.[1-4,6] This because
of the unavailability of an effective procedure for depositing homogeneous layer of melanin on
typical supports used to fabricate electrical devices. The term melanin is used to indicate a class
of ubiquitous polymeric pigments primarily formed by various combinations of 5,6-
dihydroxyindole (HQ) and its corresponding redox forms, such as indolequinone (IQ), quinone-
methide (MQ), quinone-imine (NQ). A pioneristic work attributed to melanin electrical
properties similar to those of amorphous semiconductors, recognizing its possible
implementation in switching devices [1]. A great deal of work is nowadays in progress on
optical and optoelectronic properties of melanin aiming to possible technological application [2-
10]. As stated above, one of major drawbacks in the application of melanin as active material in
electronic devices is represented by its scarce solubility in common solvents, preventing easy
deposition of homogeneous layers. Only recently, the successful melanin layer deposition on
substrates like quartz, glass or ITO/glass, allowed to better study the optical properties like the
characteristic broadband absorption coefficient. [8] The evidence of such an absorption, not
strictly following a Tauc’s law, is one of the most recent research items under debate
underlining that the picture of melanin as an amorphous semiconductor should be considered
with care.[5,6,8-10] Meredith et al., invoked the chemical disorder model to explain spectrum
and simulated it as a linear overlapping of Gaussian shaped absorption due to different
chromophoric units.[13] The most critical point is that the model accounts for independent and
similarly featured chromophores. A more dynamic chemical disorder model has been elaborated
by d’Ischia et al., by considering eumelanin’s components as a system of delocalized -electron
interrupted at irregular interval and forming discrete chromophoric units that may equilibrate,
thus resulting in a chromophore mixing [13]. Raman measurements and X-ray absorption and
photoemission spectroscopy by synchrotron radiation enlightened on the electronic states of
condensed phase eumelanin layers due to a combination of HQ and IQ based tetramers
contribution.[14] While several groups focused their attention mainly on the absorption
coefficient, in recent years interesting studies on carrier transport properties have been
performed, thus opening a new route for melanin-based device production.[7-9,15]
In melanin, two kinds of bonded water molecules are present: one is a weak water bonded
molecule, that can be adsorbed on the surface and/or into the bulk polymer and easily removable
under vacuum, another is strongly bound to the bulk polymer structure and its removal requires
thermal treatments.[3] The main melanin peculiarity, descending from the above mentioned two
type of water bonded molecules, is the dependence of melanin electrical properties and,
specifically, of electrical conductivity, on the hydration state. It can be inferred that the
modifications of the hydration degree can strongly affect the electrical transport across any type
of melanin-based devices.
This work aims to give a comprehensive overview of the results achieved on the carrier transport
characteristics disclosed for synthetic melanin layers deposited on both ITO/glass and silicon
substrates. Here, the correlation between two extreme hydration conditions starting from ambient
(wet) to high vacuum one (dry) are enlightened and discussed. The paper is developed in two
different sections, the first one describe the working of a metal-insulator-metal (MIM) like
device build up on ITO/glass, the second one is focused on devices tailored on silicon resulting
in a melanin based metal-insultator-semiconductor (MIS) like structure. For details on chemical
and structural features of SM layers we refer the reader to our recent literature.[8,15]
EXPERIMENT
Synthetic melanin (SM) layers, around 50 nm thick, were spin coated (2000 RPM) on both
plasma treated ITO/glass and p-type (pSi) and n-type silicon (nSi). Details on plasma treatment
procedures can be found in our previous works.[8,15] The layers were deposited from solution,
prepared by dissolving synthetic melanin powder (70 mg,) in water (2mL) and 26% aq. ammonia
solution (1mL). Prior to use, the solution was stirred and sonicated for one hour, and
subsequently centrifuged at 3500 rpm for 15 minutes. Optical characterizations in
melanin/ITO/glass films were made by collecting reflectance (R) and transmittance (T) spectra in
air, in the wavelength range = 400 ÷ 800 nm, by using a Perkin Elmer Lambda 9
spectrophotometer . The absorption coefficient values were calculated following the Ritter-
Weiser expression.[16] Gold (Au) contacts were evaporated (Ø=500 m) on melanin based
structure on ITO/glass and pSi (Nd= 1.0 1015 cm-3, 1-10 cm) and nSi (Na=1.8 1015 cm-3 , 5-10
cm). Current–Voltage (IvsV) hysteresis loops on the Au/SM/ITO/glass structure in air and
vacuum were measured by using an electrometer and a power supply, and collected at different
voltage sweep speeds and loop amplitudes. The IvsV hysteresis loops under white light pulses
were performed in vacuum by irradiating the Au/SM/ITO/glass structure by a halogen lamp
white light (AM 1.5, 100 mW/cm2). The capacitance vs frequency (C vs f) read out on melanin
on ITO/glass structure were collected in the frequency range 100 ÷ 1 MHz, both in air and under
vacuum, by applying VAC = 50 mV and zero continuous bias. The capacitance-voltage (C-V)
hysteresis measurements on Au/SM/p(n)Si were performed both in air and under vacuum (p=10-5
mbar) at room temperature, by using an HP 4194A impedance analyzer at a frequency of 1 MHz
of the sine wave voltage signal with a signal amplitude VAC = 50 mV superimposed to the
continuous voltage, VDC.
DISCUSSION
Optical Absorption and Capacitance vs Frequency of Melanin Stack Layer
Figure 1a displays the characteristic broadband absorption spectrum of a 50nm thick melanin
layer as determined from reflectance and transmission results. Interestingly, such a features is
present notwithstanding the layer thickness one order of magnitude lower than those published
elsewhere, therefore evidencing a good layer homogeneity degree.[15]
The specific capacitance Ci of the melanin stack layer (Figure 2b) exhibited a pronounced
frequency dispersion in air rather than under vacuum. The Ci values calculated starting from C vs
f (Fig. 1) and were found in air Ci (1.2MHz)=180 nF cm-2 and Ci(12kHz)=590 nF cm-2 , under
vacuum Ci(1.2MHz) = 91 nF cm-2 and Ci(12kHz) = 127 nF cm-2. Interestingly, when considering
the relative dielectric permittivity in air r, where r = Ci d/0, (= 8.85 10-12 F m-1: vacuum
dielectric permittivity), the value ranged from r =100 at low frequency to r =10 at high
frequency. Under vacuum, the frequency dispersion of the dielectric permittivity was low and its
value was around r =10 in the overall examined frequency range. These results evidenced a
water-induced increase of the melanin dielectric permittivity, furthermore accompanied with the
conductivity increase at high hydration level. This behavior, observed in some highly conjugated
polymers when increasing the humidity level, was ascribed to the redistribution and reordering of
the adsorbed water molecules. [17,18]
Figure 1. (a) Broadband absorption coefficient of a 50 nm thick melanin layer on ITO/glass. (b) Specific
Capacitance (b) vs frequency dispersion for a d=50 nm thick layer of synthetic melanin spin coated on
ITO/glass substrate in air (stars) and under vacuum (triangles).
Current-Voltage Hysteresis Loops of Melanin-Based Devices on ITO/glass
The Current-Voltage (IvsV) hysteresis loops (see Figure 2(a-b)), collected in air and under dark
on Au/SM/ITO/glass MIM structure at different voltage loop amplitudes VL (2a), display an
increasing of the loop area with the increasing of the loop amplitude VL. Moreover, the higher
the voltage sweep speed dV/dt, the higher the loop area. Both results evidenced that a
displacement current
dt
dV
C
dt
dC
V
dt
CVd
Id )(
(1)
adds to the resistive part of the current. [19-21] This contribution is the sum of two terms. The
first one is depending on capacitance variation and due, for example, to charge
trapping/detrapping mechanism (VdC/dt) that add/subtract a contribution (dC/dt>0 or dC/dt<0
respectively) to the displacement current, the second one increases with the voltage variation in
time and, therefore, with voltage sweep speed. The results shown in figure 2(a-b) evidence that
in melanin based MIM structures both contributions are present.
Figure 2. Current-Voltage hysteresis loops of melanin based MIM structures voltage loop amplitude (2a)
and sweep speed (2b, from ref.8) evidencing the contribution of the displacement current to the resistive
part.
Also, when changing the voltage loop amplitude, the open circuit voltage, i.e. the voltage where
the displacement current is zero, increases too, both in the forward and reverse bias regions (see
inset in Figure 2a). Finally, while the loop area increases with the voltage sweep speed (see inset
in Figure 2b), the open circuit voltage keeps almost constant. In Figure 3, the comparison
between the room temperature (RT) hysteretic behaviour (VL = 3V, voltage sweep rate of 9
mV/s) occurring in air and vacuum (3a) evidence a relevant loop shrinking and current lowering
due to the reduction of the hydration state. In the same figure the comparison between the loops
collected in vacuum under dark and white light illumination are also shown (3b). Interestingly, a
detectable enlargement of the loop was observed if the hysteresis is collected in vacuum under
white light illumination. It has to be underline that such a behaviour was observed only in
vacuum.
Figure 3. Current-Voltage hysteresis loops collected in Au/SM/ITO/glass MIM structure in air and
vacuum evidencing the shrinkage due to water percolative paths removal. (from ref.8)
In fact, the electron/hole percolation and trapping mechanisms in melanin layer are due to the
water molecules of the hydrated melanin and therefore relates to the ambient humidity degree.
Under vacuum, the weakly bonded water that induced percolative path and trapping centers are
removed, therefore explaining the lowering of the conductivity and the observed loop shrinking.
The loop enhancement observed in vacuum and under white light evidence an interesting photo-
induced space charge generation and storage. This cannot be assigned solely to a photo-
resistance effect adding to the dark resistance one, since it should be evidenced only by the
increase of the I-V slope. More specifically, when considering the results in Figure 3b, this
corresponds to the observed rotation of the under light hysteresis loop axis respect to that
observed under dark. Therefore, the large enhancement of the loop area (about a factor of four)
indicates that photo-carriers space charge storage other than a photoresistance effect is the
predominant mechanism.
The above shown hysteretic curves, under dark and light exposure, let to hypothesize an
intriguing memory-like behavior of melanin. In fact, memory (write once read many times,
WORM) capabilities have been recently verified by preliminary tests on melanin/ITO/Glass
device[8]. We showed that a writing stage occurs at a voltage of – 3.7 V and -2.8 V in air and
vacuum respectively; with no erasing voltage. [8] The current on/off ratio was found of around
100 in both cases and consistent with similar ones observed in others organic polymers[22].
Electrical Transport in Melanin-based MIS Device on Silicon
The comparison between representative high frequency Capacitance-Voltage (1 MHz) curves
collected in air for Au/p(n)Si reference diodes and those obtained in melanin-based MIS
structures evidenced that the hysteretic behaviour is due to the melanin layer [Errore. Il
segnalibro non è definito.]. Such a characteristic is a consequence of the flat band voltage shift,
Vfb = Vr - Vf , where Vf and Vr are the flat band voltages when the MIS device is biased from
depletion to accumulation and from accumulation to depletion, respectively. The theory behind
such a flat band shift can relates both to trapped charges or to mobile ions densities Q± by the
general expression[23]:
Q±= -q ±= -q (N+-N-) = CiVfb (1)
where Q± is the net positive/negative charge density (holes/electrons or positive/negative mobile
ions ), q = 1.6x10-19 Coulomb is the elementary charge, qN+ and qN- represent the positive and
negative charge densitiesrespectively and Ci is the dielectric stack layer specific capacitance.
Therefore, a negative (positive) flat band voltage shift when reverting the voltage sweep from
accumulation to depletion, implies N+>N- (N+<N-) i.e. a net positive (net negative) charge
density and corresponds to a counter-clockwise (clockwise) hysteresis loop direction.
1.3.1 C-V loops at fixed sine wave voltage frequency and variable loop amplitude
Figure 4 shows the hysteresis loops collected at 1.0 MHz in air (8a,c) and vacuum 8(b,d) at
different loop amplitudes, VL , on melanin-based MIS devices on pSi (Au/SM/pSi-DBD) and nSi
(Au/SM/nSi-DBD) and the corresponding trapped charge vs VL and net mobile densities
calculated from the estimated flat band shifts (8e,f
In air, the C-V hysteresis windows were found always increasing with the voltage loop
amplitude (see Figure 4). Specifically, values going from -0.15V up to -1.28 V in diodes made
on pSi and from +0.15 V to +1.53 V in those made on nSi were observed for loop voltage
amplitude VL from ± 1 V up to ± 3 V. By looking at the sign of flat band voltages (i.e. hysteresis
directions), it can be observed that a hole trapping process occurs in melanin devices on pSi,
while an electron trapping process in those on nSi. Figures 8(e, f) show the calculated trapped
charge densities in air, as derived from the flat band shift values estimated from loops in figure
8a and 8c and from specific capacitance of the melanin stack layer at 1 MHz derived from C vs f
measurements in air (Ci = 180 nF cm-2). In all devices, the order of magnitude of the
hole/electron trapped densities were found around 1011cm-2 and increasing with the loop
amplitude. Furthermore, these results suggest that an almost balance between hole/electron
trapping sites is present in the melanin matrix. It has to be underline that these values are around
one order of magnitude higher, in the same voltage range, than those observed in polymer-based
structure but including metallic nanoclusters [24].
In vacuum, the hysteresis loops have a completely different shape. Relevant to note, the direction
is reversed both in devices on p and n-type substrates. The origin of such a behavior has been
recently discussed [Errore. Il segnalibro non è definito.]. Briefly, an ambipolar ion drift mechanism was
hypothesized, mainly due to the presence of both positive (H+) and negative (OH-) charged ions
deriving from dissociation of residual inner water. The ion drift is enhanced in vacuum and
minimized in air, since in the last case the higher water ions (H+, OH-) concentration produces a
uniform charge
distribution. Conversely, this allows a much stronger ion-ion interaction overwhelming the ions
separation induced by the external bias [25,26].
Figure 4. Hysteresis loop collected in air (a-c) and vacuum on melanin –based MIS devices on pSi (a-b)
and nSi(c-d). (e) Electron and hole charge densities as a function of the voltage loop amplitude calculated
from hysteresis loop collected in air on melanin based device on pSi (circle) and nSi (stars) (see expr.1)
(f) Positive (left y axis) and negative (right y axis) mobile ion density obtained from the estimated values
of flat band voltage shift of the hysteresis loops collected under vacuum on pSi (circle) and nSi (stars)
Moreover, under accumulation, positive (negative) ions are pinned at the negatively (positively)
biased Au electrode, while negative (positive) ions drift to melanin/nSi interface. Under voltage
sweep reversal, negative ions in melanin on pSi and positive ions in melanin on nSi drift towards
the Au electrode, respectively, while the corresponding opposite ions are repelled, resulting in
the observed flat band voltage shift,Vfb. The observed net mobile charge densities (negative in
devices on pSi and positive in those on nSi) were ascribed to the ion pinning effect at Au
electrode, stronger than that at the melanin/Si interface. The finale effect is the observed excess
of negative mobile ions in melanin-based devices on pSi and of positive mobile ions in those on
nSi. The net mobile ion densities were found increasing with the loop amplitude in devices on
pSi while keeps almost constant in those on the nSi.
.
CONCLUSIONS
This work constitute a summary of our recently achievements on optoelectronic transport across
melanin based device on ITO/glass (melanin-based MIM) and silicon (melanin-based MIS). The
possibility of depositing melanin layers on substrates typically used in solid state physics
(Silicon) and organic electronic (ITO/glass) allowed to evidence interesting features of melanin
transport like ambipolar behavior, modification of charge transport mechanisms and charge
storage effect under different external environment, as such as in dark or under white light
exposure. The last effect disclosed furthermore the possible memory-like melanin capabilities
against voltage or light stimuli. Respect to this item, further studies are still in progress by
devoted experiments in order to test melanin based memory devices capabilities and preliminary
results are at present under examination.
ACKNOWLEDGMENTS
Acknowledgements: Work supported by MIUR, ‘‘Progetto PRIN 2009 PRAM8L’’and by
Università degli Studi di Bari “Aldo Moro“. We acknowledge Mr. A. Sacchetti for technical
assistance.
REFERENCES
[1] J. Mc Ginnes, P.Corry and P. Proctor, Science 183 (1974), 853
[2] T. Ligonzo, M. Ambrico, V. Augelli, G. Perna, L. Schiavulli, M.A. Tamma, P.F. Biagi, A.
Minafra and V. Capozzi, J. Non-Cryst. Solids, 355 (2009) 1221
[3] M. Jatstrzebska, A. Kocot and L. Tajber, J. Photochem. and Photobiol. B: Biology 66 (2002)
201
[4] M. Jatstrzebska, A. Kocot, J.K. Vij, J. Zalewska-Rejdak and T. Witecki, J. Mol. Struct., 606
(2002) 205
[5] J.P. Bothma, J. deBoor, U.Divakar, P.E.Schwenn and P. Meredith, Adv.Mat., 20 (2008)
3539
[6] P. Meredith, B.J. Powell, J. Riesz, S.P. Nighswander-Rempel, M.R.Pederson and E.G.
Moore, Soft Matt., 2 (2006) 37
[7] M.A. Rosei, L. Mosca and F. Galluzzi, Synth. Metals. 76 (1996) 331 - 335
[8] M.Ambrico, A.Cardone, T.Ligonzo, V.Augelli, P.F.Ambrico, S.Cicco, G.M.Farinola,
M.Filannino, G.Perna and V.Capozzi, Org.Electron.
[9] J.E.de Albuquerque,C. Giacomantonio, A.G. White and P. Meredith, Appl.Phys.Lett. 87
(2005) 061920
[10] V.P. Grishchuk, S.A, Davidenko, I.D. Zholner, A.B. Verbitsckii, M.V. Kurik and Y.P.
Piryatinskii, Techn. Phys. Lett. , 28 (2002) 896
[11] K.B.Stark,J.M.Gallas,G.W.Zajac, M.Eisner,J.T.Golab, J.Phys.Chem.B Vol.107(2003),
p.3061
[12] M.L.Tran, B.J.Powell and P.Meredith, Biophys.Journ.Vol.90(2006), p.743 and refs. therein
[13] A. Pezzella, A. Iadonisi, S. Valerio, L. Panzella, A. Napolitano, M. Adinolfi, and M.
d’Ischia, J. Am.Chem.Soc.Vol.131(2009), p.15270
[14] L. Sangaletti, P.Borghetti, P.Ghosh, S. Pagliara, P. Vilmercati, C.Castellarin-Cudia, L.
Floreano, A. Cossaro, A. Verdini, R. Gebauer and A. Goldoni, Phys. Rev.B 80(2009)174203
[15] M.Ambrico, P.F.Ambrico, A.Cardone, T.Ligonzo, S.R.Cicco R.di Mundo, V.Augelli,
G.M.Farinola Adv. Mater. 2011, 23, 3332–3336
[16] D. Ritter, K. Weiser, Opt. Comm. 57 (1986) 336 – 338
[17] Wim Groenewoud in Characterization of polymers by thermal analysis, Elsevier Science,
2001, The Netherland
[18] H.A. Pohl (1986), Journal of Electronic Materials 15 (1986)201.
[19] H.S.Majumdar, A.Bandyopadhyay, A. Bolognesi and A.J.Pal, J.Appl.Phys.2002,91,2433
[20] W.Brutting, H.Riel, T.Beierlein and W. Reiss J.Appl.Phys. 2001,89,1704
[21] Y-J.Lin, J.Appl.Phys.2008,103,063702
[22] L. Li, Q.D. Ling, S.L. Lim, Y. P. Tan, C. Zhu, D. S. H. Chan, E.T. Kang and K.G. Neoh,
Org. Electron. 2007, 8, 401
[23] E. H. Nicollian, J. R. Brews, in MOS physics and Technology, Wiley, New York, NY,1981.
[24] W. L.Leong, P. S. Lee, A. Lohani, Y. M. Lam, T. Chen, S. Zhang, A. Dodabalapur and S.
G. Mhaisalkar Adv.Mat. 2008,20,2325
[25] B. Martin, H. Kliem, J. Appl. Phys. 2010, 107, 076103.
[26] B. Martin, H. Kliem, Appl. Phys. Lett. 2009, 95, 032901.
