Research progress on polymer–inorganic thermoelectric nanocomposite materials
Post on 11-Sep-2016
Progress in Polymer Science 37 (2012) 820 841
Contents lists available at SciVerse ScienceDirect
Progress in Polymer Science
j ourna l ho me p ag e: www.elsev ier .com/ locate /ppolysc i
Research progress on polymerinorganic thermoelectricnanocomposite materials
Yong Dua,b, Shirley Z. Shenb,, Kefeng Caia,, Philip S. Caseyb
a Functional Materials Research Laboratory, Tongji University, 1239 Siping Road, Shanghai 200092, Chinab Materials Science and Engineering Division, CSIRO, Private Bag 33, Clayton South, 3169, Victoria, Australia
a r t i c l e i n f o
Article history:Received 13 JuReceived in reAccepted 10 NAvailable onlin
Keywords:Conductive poNanocompositSeebeck coefElectrical condThermal cond
a b s t r a c t
A thermoelectric (TE) material is a material where a potential difference is generated as a
Abbreviationsconductivity; poly(3,4-ethylD), poly(1,12-poly(3-methypoly(vinyl acebutoxy; CSA, dimethyl sulftetrathiofulval
0079-6700/$ doi:10.1016/j.vised form 9 November 2011ovember 2011e 18 November 2011
result of a temperature difference or the corollary of this where a temperature differenceis generated when a voltage is applied. These phenomena can be used to generate elec-tricity and/or control temperature. Traditionally, thermoelectric materials are inorganicsemiconductors which have been limited in their application by low efciency and highcost. Since the 1990s, both theoretical and experimental studies have shown that low-dimensional TE materials, such as superlattices and nanowires, can enhance the value ofthe TE gure of merit (ZT) which is an indicator of TE thermodynamic efciency. To date ithas not been feasible to apply these materials in large-scale energy-conversion processesbecause of limitations in both their heat transfer efciency and cost. When compared toinorganic materials, organic conducting polymers possess some unique features, such aslow density, low cost, low thermal conductivity, easy synthesis and versatile processabilityand their use in preparing polymer-inorganic TE nanocomposites appears to have greatpotential for producing relatively low cost and high-performance TE materials. Recently,an increasing number of studies have reported on polymeric and polymer-inorganic TEnanocomposite materials. The purpose of this paper is to review the research progress onthe conducting polymers and their corresponding TE nanocomposites. Its main focus is theTE nanocomposites based on conducting polymers such as polyaniline (PANI), polythio-phene (PTH), poly (3, 4-ethylenedioxythiophene): poly (styrenesulfonate) (PEDOT:PSS),as well as other polymers such as polyacetylene (PA), polypyrrole (PPY), polycarbazoles(PC) and polyphenylenevinylene (PPV). Typically, polymer-inorganic TE nanocompositesare produced by physical mixing, solution mixing and in situ polymerization. The keyfactors that limit the use of these polymers and their polymer-inorganic TE nanocomposites
: TE, thermoelectric; ZT, gure of merit; PF, power factor; , Seebeck coefcient; , electrical conductivity; , thermalT, absolute temperature; RT, room temperature; DOS, density of states; PANI, polyaniline; PTH, polythiophene; PEDOT, PSSenedioxythiophene): poly(styrenesulfonate); PA, polyacetylene; PPY, polypyrrole; PC, polycarbazole; PIC, polyindolocarbazole; P(2Cz-bis(carbazolyl)dodecane; PPV, polyphenylenevinylene; PMMA, poly(methyl methacrylate); PTT, polythieno[3,2-b]thiophene; PMet,lthiophene); P3HT, poly(3-hexylthiophene); P3HTT, poly(3-hexylthiothiophene; P3OT, poly(3-octylthiophene); PP, polypropylene; PVAc,tate); PCVH, poly(N-octyl-3,6-dihexyl-2,7-carbazolenevinylene); PMeOPV, poly(2,5-dimethoxyphenylenevinylene); EtO, ethoxy; BuO,()-10-camphorsulfonic acid; CNT, carbon nanotube; SWNT, single-walled nanotube; BFEE, boron triuoride diethyl etherate; DMSO,oxide; EG, ethylene glycol; DMF, N,N-dimethylformamide; THF, tetrahydrofuran; Tos, tosylate; PF6, hexauorophosphate; TTF-TCNQ,enetetracyanoquinodimethane; VRH, variable range hopping; NDH, nearest-neighbour distance hopping.ding author. Tel.: +61 3 92526227.ding author. Tel.: +86 21 65980255; fax: +86 21 65980255.resses: Shirley.Shen@csiro.au (S.Z. Shen), email@example.com (K. Cai).
see front matter 2011 Elsevier Ltd. All rights reserved.progpolymsci.2011.11.003ne 2011
Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841 821
as TE materials are their low ZT values. More recent developments designed to overcomethe limitation including, for example, the use of carbon nanotubes and graphenes and theuse of computational modelling to accelerate the selection of suitable pairs of conductivepolymer and inorganic TE materials to achieve best possible nanocomposites are reviewed.
2011 Elsevier Ltd. All rights reserved.
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. PANI-inorganic TE nanocomposites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1. PANI as a TE material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.2. PANI-inorganic TE nanocomposites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.1. Physical mixing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.2.2. Solution mixing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3. In situ oxidative polymerization/intercalation. . . . . . . . . . . . . . . . .
3. PTH-inorganic TE nanocomposites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. PTH as a TE material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. PTH-inorganic TE nanocomposites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4. PEDOT:PSS-inorganic TE nanocomposites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. PEDOT:PSS as a TE material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.
5. Other5.1. 5.2. 5.3.
TE enerused in heaFig. 1 illustformance inconverted effect whertricity. TE and p-typedescribed ator) respectsuch as loneasy mainteuse is curreof limited erials used [
Fig. 1. TE heaheats or cools current ow. (is generated thPEDOT:PSS-inorganic TE nanocomposites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 829 polymers and polymer-inorganic TE nanocomposites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 829PA as a TE material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 829PPY as a TE material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 830PC as a TE material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 831ary and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 838wledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 838
ences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839
gy-conversion is an all-solid-state technologyt pumps and electrical power generators [1,2].rates the principal effects governing TE per-cluding: (A) the Peltier effect where power is
to a heating and cooling and (B) the Seebecke temperature differences are converted to elec-devices are typically an assembly of both n-
semi-conductors which may simplistically bes electron rich (donor) and electron poor (accep-ively. TE devices have many attractive featuresg operating lifetime, no moving parts, no noise,nance and high reliability [1,3]. However, theirntly limited to only niche applications becausefciency which is determined by the TE mate-4,5]. Thermoelectric efciency is measured by
t engines. (A) When current is run across a TE junction, itthrough the Peltier effect, depending on the direction of the
the dimensionless gure of merit, ZT = 2T/, where isthe Seebeck coefcient, is the electrical conductivity, isthe thermal conductivity and T is the absolute temperature. In polymer and polymer-inorganic TE nanocompositesthe value of the Seebeck coefcient typically ranges from4088 to 1283 V/K, electrical conductivity from 107 to104 S/cm and thermal conductivity from 0.02 to 1.2 W/mK.The best materials available today for devices that oper-ate near room temperature have a ZT of about 1, and TEcoolers with a ZT of 1 operate at only 10% of Carnot ef-ciency. A domestic refrigerator operates at about 30% ofCarnot efciency and this efciency could be achieved bya TE device having a ZT of about 4 . To improve TEperformance, a combination of high electrical conductiv-ity and low thermal conductivity is necessary and whichare interdependent and determined by the electronic struc-ture (band gap, band shape, and band degeneracy near theFermi level) and the scattering of charge carriers (electronsor holes) . For example, the effect of a local increase indensity of states (DOS) on is given by the Mott expressionas Eq. (1) :
Here, depends on the energy derivative of the energy-dependent taken at the Fermi energy EF as in Eq.(2):
(E) = n(E)q(E), n(E) = g(E)f (E), (2)B) When heat ows across the junction, electrical currentrough the Seebeck effect.
where g(E)tion, q is t. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 821. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 823
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 823. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 826. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 828. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 828 is the density of states, f(E) is the Fermi func-he carrier charge, and (E) is the mobility.
822 Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841
Therefore, Eq. (3).
The = el + phand ph corrfor a given mrier concenSo, changinoptimizatiodifcult [6progress hiwith enhanoutlines thiin TE mater
In 1993,reduced phconnemenmance becanear the babeck coefsame time,at the quancan reduce ttion was su[7,1127]. Rwill give risinstance, p-
m temtum-dcauseon scamatercompoe bulkFig. 2. Timeline of TE and related research
Eq. (1) may be expressed in the form of
Bephonbulk nanoBiSbTof a semiconductor can be expressed as, where el corresponds to carrier contribution,esponds to the phonon contribution. Therefore,aterial, increases with increasing charge car-
tration and so does , while decreases [4,8].g one parameter alters the others making then of these contradictory properties extremely]. This is why there has been only modeststorically in nding or synthesizing materialsced ZT values at room temperature (RT). Fig. 2s by presenting a timeline of signicant eventsials research over the last 30 years.
Hicks and Dresselhaus [9,10] predicted that theysical dimensionality of TE materials (quantumt) may lead to greatly enhanced TE perfor-use the enhanced DOS for electrons (or holes)nd edge can increase the magnitude of the See-cient at a given carrier concentration. At the
the increased boundary scattering of phononstum well-barrier interfaces in the superlatticehe lattice thermal conductivityph. This predic-bsequently conrmed in a number of studieseductions in the lattice thermal conductivity
e to a high ZT in corresponding superlattices. Fortype Bi2Te3/Sb2Te3 superlattices have a ZT = 2.4
bulk materporous thinpower factothe other.
Althougmaterials hhigh cost owell as heapoor procesConsequencompetitivepotential be
Organic polyanilineethylenedio(PEDOT:PSSpolycarbazotheir derivaThey possesing low deninto versatiimportant c
Althougrials, they acoefcient ae last 30 years.
perature , and n-type PbSe0.98Te0.02/PbTeot superlattices have a ZT = 3 at 550 K .
of the decreased thermal conductivity by thettering effect, ZT = 1.7 at 700 K for a AgPb18SbTe2ial , 1.47 at 448 K for a Bi2Te3/Sb2Te3 bulksite , 1.4 at 373 K for a nanostructured
sample , 1.56 at 300 K for a Bi0.52Sb1.48Te3
ial , and 1.8 at 300 K for a Bi0.4Sb1.6Te3
lm (with the measurement error 13% forr and 28% for ZT)  were reported one after
h some recently prepared nano inorganic TEave exhibited relatively high ZT values, thef raw materials and production facilities asvy metal pollution considerations  and thesability limit their wider application [4,2931].tly, developing high performance using cost
TE materials is of the key importance for thenet of their properties to be realized [30,32].
conducting polymers, including (PANI), polythiophene (PTH), poly(3,4-xythiophene): poly(styrenesulfonate)), polyacetylene (PA), polypyrrole (PPY),les (PC), polyphenylenevinylene (PPV) andtives, have great potential for use in TE devices.s a number of advantageous properties includ-sity, low cost and easy synthesis and processingle forms. The chemical structures of the mostonducting polymers are presented in Table 1.h they are electrically conducting, as TE mate-re limited by electrical conductivity, Seebecknd power factor (PF). The PF of most polymer TE
Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841 823
Table 1Chemical structures of a few typical conductive polymers.
Materials Chemical structure Materials Chemical structure
Polyaniline:leucoeme(y = 1),emeraldin(y = 0.5), apernigran(y = 0)
materials isis 3 orders oart inorganto overcomposite wheTE nanostruthe TE proppure polympaid their ature compo
The mosPANI, PTH Some othenanocompo
2.1. PANI a
PANI, likpresence o(localizatiolic domainCharge tranhigh electrinect them and Seebecon preparatraldine
n in the range of 1061010 W m1 K2, whichf magnitude lower than that of the state-of-the-ic TE materials . An emerging approache these limitations is to use a polymer nanocom-re the composite material consists of inorganicctures embedded in a polymer matrix such thaterties of the composite are greater than theer. Recently, more and more researchers havettentions to polymer-inorganic TE nanostruc-sites.t promising nanocomposites are those based onand PEDOT: PSS and these are reviewed rst.r conducting polymer and polymer-inorganicsites are also reviewed.
organic TE nanocomposites
s a TE material
e most conducting polymers, is typied by thef metallic domains and microscopic disordersn) with the electrical conductivity in the metal-s higher than that in the disordered regions.sport between metallic domains occurs viacally conductive PANI chains which intercon-. Consequently, the electrical conductivityk coefcient of PANI are strongly dependention conditions including temperature  with
electrical tring the mat
In ()-1or blends (PANI-CSA/812 V/K dent upon low as 1%) ithe diffusiostates withat metal-inpathways oA CSA-dopeof electricalcally conduwas reportexhibiting athat of bulkdoped PANfurther.
For thethe electricconductionture depenfrom tempular metaland Seebecwith HCl inPolythiopheneS
Poly(para-phenylenevinylene) nansport properties greatly improved by reduc-erials degree of disorder .0-camphorsulfonic acid (CSA) doped PANI of PANI-CSA with poly(methyl methacrylate)PMMA) , a Seebeck coefcient (300 K) ofwas reported and found to be linearly depen-temperature even at low volume fractions (asndicating that microscopic transport is throughn of charge carriers in extended electronic
relatively narrow bandwidth. Charge transportsulator boundaries and along interconnectedf PANI-CSA occurred via hopping mechanisms.d PANI multilayered lm structure, composedly insulating emeraldine base layers and electri-cting CSA-doped emeraldine salt layers [31,41],ed to have a ZT value of 1.1 102 at 423 K,
six fold increase in the ZT value at 300 K over CAS-doped PANI. Axial stretching of the CSA-I lm was found to increase the ZT value even
PANI samples doped with HCl or H2SO4,al conductivity was dominated by hopping
[39,42,43] but showed a U-shaped tempera-dence of Seebeck coefcient which may ariseerature-dependent tunnelling between gran-lic islands . Both electrical conductivityk coefcient of organo-soluble PANI dopedcreased with increasing temperature (electrical
824 Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841
conductivitbeck coefdoped PANization conductivitdeclined wwhich displtivity, howeand ZTmax =concentrati
The relaSeebeck coe. Here,ity varied that increaopposite ebeck coefcdecreased various proinvestigatedconductivitcorrelated wof dopant. cient anddecreasing acid-dopedtures werenanostructuular orderinmobility shcoefcient ously increaPANI lms would be rtures into thmorphologing the bouheat transfe
PANI canresultant pbres to n
uctivitcular we of cping [5d Seebeparatuctivithe use. Whil
is insrical alectric
couldods, ZTE mat
PANvity v-inorgple, Pits allo7] and-inorg. 3. The effect of main chain structure of PTH lms on the relationship be
y from 0.47 107 to 105 S/cm and See-cient from 6 to 93 V/K) . In HClI synthesized by chemical oxidative polymer-, as doping concentration increased, electricaly and ZT value initially increased but thenhich was in contrast to Seebeck coefcient,ayed the opposite behaviour. Thermal conduc-ver, was insensitive to the doping conditions
2.67 104 at 423 K was obtained when HClon was 1.0 M.tionship between electrical conductivity andfcient of doped PANI and PPY was investigated
the logarithm of the electrical conductiv-linearly with the Seebeck coefcient suchsing carrier concentration had diametricallyffects on electrical conductivity and See-ient, so that increasing electrical conductivitythe Seebeck coefcient. The TE properties oftonic acid-doped PANI lms have also been
 and exhibited extremely low thermaly for an electrically conducting material, andith neither electrical conductivity nor the kindIn an attempt to increase the Seebeck coef-
electrical conductivity while simultaneouslythermal conductivity, -naphthalene sulfonic
condmoledegreof doity anon prcondods, tlmsity, itelectthe ePANImethas a T
AsductiPANIexamand [30,5PANI PANI nanotubes and PANI without nanostruc- prepared so as to examine the inuence ofred material on TE properties . If the molec-g of polymer chains is improved, charge carrierould be enhanced such that both the Seebeckand electrical conductivity should simultane-se consistent with results observed in stretched. Additionally, the free path of phononseduced effectively by introducing nanostruc-e system due to the strongly bent and entwinedy of PANI nanotubes, which resulted in increas-ndary phonon scattering at the interfaces in ther process .
be prepared using different techniques withroducts varying from coarse to ne powders,anobers or as thin lms, etc. The electrical
posites areoxidative pcial polyme
2.2.1. PhysiDry pow
were physi1 GPa [54cient beingelectrical coof its origidecrease inTE propertof electricareagent andmaterials welectrical conductivity and gure-of-merit .
y of PANI depends on variables includingeight, oxidation level, molecular arrangement,
rystallinity, inter-chain separation and degree0]. Besides this, both the electrical conductiv-eck coefcient of PANI are strongly dependention conditions and temperature with electricaly being enhanced by various synthesis meth-
of multilayered lms as well as stretched thee, PANI has extremely low thermal conductiv-ensitive to the sample preparation conditions,s well as the kind of dopant . Therefore, ifal conductivity and the Seebeck coefcient of
be improved by dopants and new preparation would be further improved and its applicationerial would have a bright future.
organic TE nanocomposites
I is stable and has a high electrical con-alue, there have been many nanocompositeanic TE materials studies undertaken using forANI with metal oxides , Bi [52,53], Bi2Te3ys , NaFe4P12, carbon nanotubes
PbTe . The main methods for preparinganic and other polymer-inorganic TE nanocom- physical mixing, solution mixing, in situolymerization/intercalation and in situ interfa-rization.
cal mixingders of PANI and Bi0.5Sb1.5Te3 (17 wt.% PANI)cally mixed and cold pressed at a pressure of]. The resultant material had a Seebeck coef-
10% lower than that of Bi0.5Sb1.5Te3 but annductivity decreasing signicantly to 3070%nal value. So the material had a remarkable
PF. Thus the principal way to improve theies of these composites may be to enhancel conductivity through optimizing the doping
its level in the polymer, and using inorganicith high electrical conductivity forming the
Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841 825
composites by pressing under vacuum to promote inti-mate contact. Most recently, HClO4-doped PANI/graphitecomposites were prepared (graphite has a high electri-cal conductivity (5.74 102 S/cm)) using mechanical ballmilling follonanolayer ithe graphitconsequentand Seebecto 120 S/cmwhile the increasing and 50 wt.%50 wt.% grananocompofound that toxidation oSeebeck coposite lmslm alone conductivitprepared bcient 50while the ealmost the of the compPANI and reture .
2.2.2. SolutIn gener
vide hybrida more effenation. Howby physicalthe oppositbeck coefby physicalmixing. Thibe better dmixing. Whprepared inites were neslight increSeebeck cocomposite w
2.2.3. In sitWhere P
by in situ oxor aniliniumboth durinlation prepared sathe degree ably increafrom 30 toand the deg
Both PAnanowires ing method
temperatures of the nanowires being greater than eitherPANI or whiskers. It is likely due to the greater quantumconnement in nanowires and consequent increase in thedensity of states/unit volume in the nanocomposite.
PbTeures structuo 3738 V/K
to 0.0here er co
uctivitthreshed, wht rementratierties ias focms [30
and P singleintera
at a honon o a ZT =compostep 1olymerical cos likelyt causeNTs. Ttill rel
ordercompoay beve thiined wrical co
e elecdo-potsoft Vent schas a v, whil
d PTHry lowwed by cold pressing . A large numbers ofnterfaces between the HClO4-doped PANI ande phases were produced during ball milling andly, when pressed, both electrical conductivityk coefcient increased remarkably (from 1.23
and from 0.82 to 18.66 V/K, respectively),thermal conductivity increased slightly withgraphite content (from 0.29 to 1.20 W/mK at 0, respectively). A ZT of 1.37 103 at 393 K atphite was obtained. When CSA-doped PANI/Bisites were prepared by ball milling , it washe PANI reduced the degree of aggregation andf the Bi particles during the milling step. Theefcient of the CSA-doped PANi/Bi nanocom-
was larger than that of the CSA-doped PANiwhile the reverse was found for the electricaly. Most recently, the PANI/Bi2Te3 compositesy mechanical blending had a Seebeck coef-
V/K at 300 K being similar to that of Bi2Te3,lectrical conductivity of the composites wassame as PANI (2 S/cm). Consequently, the PFosites was lower than that of either Bi2Te3 ormained essentially unchanged with tempera-
ion mixingal, it is thought that solution mixing should pro-
lms with higher electrical conductivity due toctive polymer to inorganic nanoparticle coordi-ever, when PANI/Bi2Te3 nanohybrids prepared
and solution mixing methods and compared,e result was obtained : a higher PF and See-cient value were found in the lms prepared
mixing rather in those prepared by solutions suggests that the Bi2Te3 nanoparticles mayispersed by physical mixing than by solutionere a series of PANI-Bi nano-composites were
solution , the properties of the compos-arly identical to those of pure PANI with only aase of electrical conductivity and the observedefcient, thermal conductivity and ZT of theere almost the same as those of pure PANI.
u oxidative polymerization/intercalationANI was incorporated into a V2O5nH2O xerogelidative polymerization/intercalation of aniline
in air, oxygen acted as an electron acceptorg the in situ reaction and long after interca-. The electrical conductivity at RT of freshlymples was 104101 S/cm and depended onof polymerization inside the layers and invari-sed upon aging. The Seebeck coefcient varied
200 V/K, depending on the polymer contentree of polymerization .NI/NaFe4P12 whiskers and PANI/NaFe4P12were prepared using an in situ compound-
 with the Seebeck coefcient at high
PAinterusingtructnano293 tto 570.019
Wpolymcondtion formcienconcproption hsysteCNTsusing moreallowcoefwereing i(evento phrise tnano (the pelectIt waeffecthe Cwas s
ThpseuultragradiPTH 0.9 eVdopeto veTe composite powders were prepared by in situolymerization at RT by the current researchers
nanoparticles, PANi/PbTe coreshell nanos-and PbTe/PANi/PbTe three-layer spherelikeres. When the temperature was increased from
K, the Seebeck coefcient decreased from 626 while electrical conductivity increased from22 S/cm .segregated-network carbon nanotube (CNT)-mposites were prepared , the electricaly dramatically increased when the percola-old was reached and a network of CNTs wasile the thermal conductivity and Seebeck coef-ained relatively insensitive to changes in lleron, making it potentially feasible to tune then favour of a high ZT. Subsequently, much atten-ussed on the use of CNTs in conjugated polymer,57]. For instance, nanocomposites containingANI were prepared by in situ polymerization-walled nanotubes (SWNTs) . Here, strongctions between the PANI and CNTs generate aed composite structure than that of pure PANI,creased carrier mobility so that both Seebeckand electrical conductivity of the compositesr than that of the pure PANI. Even more promis-the thermal conductivity remained constantigh SWNT loading of 41.4 wt.%) presumably duescattering at the SWNTPANI interface giving
0.004 at RT. In another study where PANI/CNTsites were prepared using a two-step method: the formation of a thick CNT network; step 2:rization of PANI in situ), the Seebeck coefcient,nductivity and PF were dramatically improved.
attributed to a size-dependent energy-lteringd by the nanostructured PANI coating layer onhe thermal conductivity for these compositesatively low (0.390.5 W/mK).
to improve the TE properties of PANI-inorganicsites, the enhancement of electrical conductiv-
the principal path, and possible methods tos are to optimize the doping reagent and levelith the use of inorganic materials having highnductivity.
rganic TE nanocomposites
a TE material
tronic band structure of PTH was obtained fromential plane-wave calculations employing ananderbilt pseudo-potential and a generalizedheme by Gao et al. . They found thatery simple band structure with a band gap ofe the very high Seebeck coefcient for p- and n-
(100 V/K and 140 V/K, which correspond doping concentrations of 0.04 hole/monomer
826 Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841
unit and 0.0also predict
Shinohaan electrolyditions andas electricaof 1.5 1ductivity weffect of caTE propertismooth surhigher electconductivittion and inchains incrhigh electriable range hin log()(1tivity lm sT curve anearest-neian activatiotemperaturwith whichchain moleand directlyshows the erelationshipmerit , and NDH [6
Besides chain in PTTE propertpolythieno[than that conjugatedconductivitof because it of main chdecrease inthe smallererties for thare compara BiTe sysfor PTH seridecreases a
performance (this phenomenon is related to presence (orabsence) and size of side chains).
Besides the above, the stereo-regularity of the extendedconjugated chain structure and/or the chain packing
affect 3-metrochemoride e carrlatile ar electck co02 evperforhievinarity r morthougonmen
engini levenstratcient d polytes. In(8 wt.%e tetnt conthe Secrease.7 1
of P3H2% ancient cantl
packtures i levelerties o
date, anic Tuble, iig. 4. Schematic views of VRH and NDH .
2 e/monomer unit, respectively) at 300 K wased .ra et al. [63,64] have prepared PTH lms usingtic polymerization method under various con-
found that the Seebeck coefcient decreasedl conductivity increased and a gure of merit04 K1 was achieved when the thermal con-as assumed as 0.1 W/mK. They have studied therrier conduction between main chains on thees of PTH lms and found that the lms withfaces and a higher degree of crystallinity hadrical conductivity, indicating that the electricaly was dominated by between-chain conduc-creased as the molecular ordering of the maineased. They found that the behaviour for thecal conductivity lm was well tted by vari-opping (VRH), which presents a linear relation/T)1/4 plot, while that of low electrical conduc-hows a non-linear log()(1/T)1/4 plot, and itst low and high temperatures can be tted byghbour distance hopping (NDH, which obeysn law, and the nearest-neighbour distance ise-dependent) and VRH, respectively. The ease
carrier conduction can occur between maincules determines the conduction mechanism
affects the ZT value of the PTH polymers. Fig. 3ffect of main chain structure of PTH lms on the
between electrical conductivity and gure-of-while Fig. 4 shows the schematic views of VRH4].main chain effects, the size of the sideH derivatives also has a signicant effect onies. For instance, the Seebeck coefcient of3,2-b]thiophene (PTT) lm was much higherof PTH and attributed to the extended -
structure . When comparing the electricaly at the same Seebeck coefcient (in the range/cm) f poly(3-alkylthiophenes), synthesizede polymerization , those with shorter sidelted in a higher Seebeck coefcient (in the
1 S/cm, PTH has the highest Seebeck coefcienthas no side chain). An increase in the densityains per unit volume will be promoted by a
also poly(electtriuchargof voloweSeebeat >1high-by acreguland/o
Alenvircientlimitto beFermdemocoefdopeof stalm as thdopaboth ity into 3tion i.e, coefsigniGivenmay mogestate
Toinorginsol the volume of side chains. It follows then that the side chain, the better should be the TE prop-e PTH series. The TE properties of a PTH seriesed with some other conductive polymers andtem in Fig. 5 . This comparison shows thates as conductivity increases, Seebeck coefcientnd the smaller the side chain, the higher the TE
ites were ppolymer poelectrical cthan 5 ordenated counTE properties. High tensile strength PTH andhylthiophene) (PMeT) lms were prepared byical polymerization in freshly distilled boron
diethyl etherate (BFEE) . The decrease inier concentration resulting from the presencend water-sensitive dopants from BFEE led to arical conductivity but a substantial increase inefcient, allowing ZT values to be maintaineden after two months storage. This suggests thatmance TE conducting polymers can be obtainedg a high charge mobility (due to greater stereo-of the extended conjugated chain structure)e effective chain packing.h both doped and undoped PTH shows goodtal stability, they have a low Seebeck coef-w electrical conductivity. To overcome these
the electronic structure of the polymers needseered for a favourable density of states andl . Interestingly, Sun et al.  recentlyed that a simultaneous increase in Seebeckand electrical conductivity was possible in a(alkylthiophene) blend with a dened density
composites of Poly(3-hexylthiophene) (P3HT) poly(3-hexylthiothiophene (P3HTT) in P3HT)rauorotetracyanoquinodimethane (F4TCNQ)centration was increased from 0.2 to 0.7 wt.%,ebeck coefcient and the electrical conductiv-d from 460 to 530 V/K and from 2 10505 S/cm, respectively. Where the concentra-TT and F4TCNQ in the P3HT lm were lower,d 0.25 wt.%, respectively, the value of Seebeckwas very high (700 V/K). The PF increasedy from 4.62 104 to 7.58 103Wm1 K2.
it indicates that organic-based compositionsigned with intentional density-of-state inho-s guided by theoretically preferred density ofutions for TE materials.ion of PTH lms with smooth surfaces, highss, high crystallinity, greater stereo-regularitynded conjugated chain structure and/or theing as well as engineered polymer electronicto produce a favourable density of states and
are among the possible methods to improve TEf PTH.
organic TE nanocomposites
only a few studies have been reported on PTH-E nanocomposites due primarily to PTH beingnfusible and having a relatively low electricaly.ctylthiophene)/silver (P3OT/Ag) nanocompos-repared by Pinter et al.  by impregnatingwder using silver perchlorate salt solutions. Theonductivity of the nanocomposites was morers of magnitude greater than its non impreg-terpart. A P3OT tablet, compressed from the
Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841 827
Fig. 5. Relatio hiophene (), polydodecylthiophene (), polyoctylthiophene (),polyhexylthio
When pris typically oxidant whin situ oxidily oxidizedprepared bthe rst stemal syntheIn the secon(80 MPa) atconductivit623 K. Thisgreater thannitude greais, howeverBi2Te3 hot Seebeck coeindicating ncient waswhile the hposite prespressed at bulk Bi2Te3Bi2Te3 phain the compof the lattefrom the thture exceedradicals thahigher the pPTH decomFig. 6a and and PTH showhile Fig. 7tron micronship between Seebeck coefcient and electrical conductivity for polytphene (+), polyacetylene(), polyaniline () and Bi-Te (*) .
site powder, exhibited an extremely large See-ient of 1283 V/K.eparing PTH by oxidative polymerization, FeCl3chosen as the oxidant. However, if used as theen preparing PTH-Bi2Te3 nanocomposites byative polymerization, the Bi2Te3 will be eas-. To avoid this, PTH/Bi2Te3 nanocomposite wasy our group using a two-step method . Inp, Bi2Te3 and PTH were prepared by hydrother-sis and oxidative polymerization, respectively.d step, a mixture of Bi2Te3 and PTH was pressed
different temperatures. The highest electricaly was 8 S/cm for a composite hot pressed at
is approximately seven orders of magnitude that of pure PTH and about two orders of mag-ter than that of a composite pressed at 473 K. It, still much lower than the conductivity of bulk
pressed at 80 MPa and 623 K (384 S/cm). Thefcient values for all composites were negative,-type conduction. The highest Seebeck coef-
98 V/K for a composite pressed at 473 K.ighest PF was 2.54 W m1 K2 for a com-sed at 623 K (20 times greater than that of473 K) but still much lower than that of thepressed at the same temperature. Besides the
se, a rhombohedral Bi2Te2S phase was foundosites pressed at 473 and 623 K. The presencer was presumably a result of reactions arisingermal decomposition of PTH as the tempera-s 473 K , which produces free S and SHt react with Bi2Te3 to form Bi2Te2S . Theressing temperature, the greater the degree ofposition and the more the Bi2Te2S formation.b presents TEM images of prepared Bi2Te3 wing the differences in particle size and shape,a, b and cpresents eld emission scanning elec-scope (FESEM) images of the fracture surface
Fig. 6. TEM images of the Bi2Te3 nanopowders prepared by the hydrother-mal synthesis method (a)  and PTH prepared by the chemical oxidativepolymerization (b).
828 Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841
Fig. 7. FESEM images of the fracture surface along the direction parallelto the hot pressing direction of the PTH-Bi2Te3 composite hot pressed at(a) 298, (b) 473 and (c) 623 K and 80 MPa.
along the direction parallel to the hot pressing direction.These images clearly show that as pressing temperatureincrease, the morphology of the composite changes.Fromthis work it was determined that in order to achieve opti-mal propertof PTH in thsintering tethe decomp
The conattention betransparencexibility, lbility and esolubility awater some
The elebe raised dielectric sN,N-dimethbecause theenhance caDMSO (01PEDOT:PSSthe Seebeckto 41 V/Kdeterminedof PEDOT:Prated and uby Jiang etlene glycollinear or exponent ledwhile the Sstant. Schospin-coatinelectrical cochanging thwith a ZT vtrical conduDMSO dopenot increasmobility whstanding PEof PEDOT:Pdoped PEDand Seebeously . 63.13 S/cm,20.7 V/K a300 K.
Most renanowires electrodepoSeebeck coewas greaterThe electricies in PTH-Bi2Te3 nanocomposites, the contente composite should be low (
Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841 829
also greater than that of the lms as electron mobilitiesin nanowires were increased by a factor of 3.
Because the electrical conductivity of PEDOT/PSS can begreatly enhanced by the addition of various organic sol-vents (up tpromising c
The moswas reportPEDOT-tosyEDOT montris-ptoluenspin-coatedatmospherethe electrotion level wconductivitthe SeebeckZT = 0.25 wwhich is thmers, while
To improof inorganic
CNTs powhen incorcentration composite mal conducunchangedites preparparticles wbridging tuholes) to traproviding a35 wt.% CNTthese tube-in vibrationhighest ZT 35 wt.% SWthermally dtions betweYu et al. [32
The secnanorods [der etc. whnanorods, sal. in 2010lm havingorganic/inoincreased wbeck coefcthat of the major chargoccur throuplayed a higPSS or unfuPEDOT: PSSimproved in0.1 at RT
sequence by which PEDOT:PSS passivated Te nanorods areformed into a smooth nanocomposite lm by solution cast-ing. Fig. 9 shows SEM and TEM images of the produced lms.
ew PEising cIOS PHSeebeW m
f both enhaners inan be ues as a
alterosite rsing (BaytrropylT:PSS
improdecreao4O9 cwouldE propchoost becauctivitoweveeck coeparticlu at Rncreasf the c16 10cient d. Furth
PA as a
as thef PA d3], or
coefcted anmperaeck co
transi, WCl6 104 Sh-orieuctiviteck coeo 1000 S/cm), PEDOT/PSS may be considered aandidate for use as a TE material.t striking experimental result related to PEDOTed by Bubnova et al. . They preparedlate (Tos) polymers by directly mixing theomers and an oxidative solution of iron (III)esulfonate. To enhance the TE properties,
PEDOT-Tos lms were reduced in an inert with tetrakis(dimethylamino)ethylene to alternic structure of the lms. When the oxida-as diminished from 36 to 15%, the electrical
y decreased from 300 to 6 104 S/cm, while coefcient increased from 40 to 780 V/K. Aas yielded at the oxidation level of 22% at RT,e highest one ever shown for conducting poly-
this polymer is also stable in air.
:PSS-inorganic TE nanocomposites
ve the TE properties of PEDOT: PSS, two classes materials have been used.ssess very high electrical conductivity andporated in a polymer matrix, as the ller con-is increased the electrical conductivity of thecan be dramatically increased, while the ther-tivity and Seebeck coefcient remain relatively. For example, in CNT-lled PEDOT:PSS compos-ed by Kim et al. in 2010 , the PEDOT:PSSere found to decorate the surface of the CNTsbe-tube junctions and allowing electrons (i.e.,vel more efciently through the composite and
high electrical conductivity (up to 400 S/cm ats). In contrast to this, thermal transport acrosstube junctions was impeded due to mismatchesal spectra between CNT and PEDOT:PSS. The(0.02) was obtained in composites containingCNT. This concept of simultaneously providingisconnecting and electrically connecting junc-en CNTs agrees with those results reported by].ond class of inorganic materials include Te84], Bi2Te3  and Ca3Co4O9  pow-ich have high Seebeck coefcient values. Teynthesized in situ with PEDOT:PSS by See et
, yielded a continuous, two-component a continuous electrical network of nanoscalerganic interfaces where electrical conductivityhile thermal conductivity diminished. The See-ient was positive and signicantly greater thanpure polymer, indicating that holes were thee carrier and that transport did not exclusivelygh the PEDOT:PSS. All of these hybrid lms dis-her electrical conductivity than either PEDOT:nctionalized Te nanorods suggesting that the
prevented oxidation of the Te nanorods andterparticle contact. Consequently, a high ZT ofwas obtained. Fig. 8 schematically shows the
NpromCLEVhigh 47 PFs obeenpowdites cdevic
Ancompdispetion polypPEDOlittlecant Ca3C
It the Tis to ciencond
HSeebnanoin sitwas iity oto 6.coefinglywithticle/inves
Inbe prstateties o[919beckrepor
TeSeebwithZrCl4 = 3 stretccondSeebDOT:PSS commercial products (CLEVIOS) areandidates for composite materials, particularly1000 and FE-T which show an unexpectedlyck coefcient suggesting promising TE PFs of1 K2 and 30 W m1 K2, respectively. Thep- and n-type polymer/Bi2Te3 composites haveced by incorporating both p- and n- type Bi2Te3to the PEDOT:PSS products and these compos-sed in solution-processed exible substrate TE
new fabrication option .native approach is where PEDOT:PSS/Ca3Co4O9lms have been prepared by mechanicallyCa3Co4O9 powder in a PEDOT:PSS solu-on P) and casting the resultant mixture onene (PP) lm substrates. When compared with
lm, the PF was substantially reduced withvement in Seebeck coefcient but a signi-se in electrical conductivity with increasingontent .
appear that the most preferable way to improveerties of PEDOT:PSS-inorganic nanocompositese inorganic materials with high Seebeck coef-use PEDOT:PSS has a relatively high electricaly and low thermal conductivity.r, when PEDOT has been properly doped, itsfcient can be as high as over 4000 V/K. PbTee modied PEDOT nanotubes were fabricatedT by our group . When the PbTe contented from 0 to 44 wt.%, the electrical conductiv-omposite increased by an order of magnitude3 S/cm, while the absolute value of Seebeckecreased from 4088 to 1205 V/K correspond-er work on alternative inorganic nanostructureslectrical conductivities, such as Ag/Cu nanopar-ods, CNTs, or graphene, is, therefore, undern. A synergic effect would be expected.
olymers and polymer-inorganic TEosites
doped PA was the rst conductive polymer tod changing from a non-metallic to a metallic
doping level increased . The TE proper-oped with transition metal halides , iodine
alkali-metal-halide  as well as the See-ient of stretch-oriented PA [93,95] have beend are summarized as follows.ture-dependent electrical conductivity and theefcient were measured for PA lms dopedtion metal halides such as FeCl3, TaC15, NbCl5,and MoCl5 by Park et al. . A maximum/cm at 220 K was achieved for the FeCl3-dopednted PA. For iodine doped PA, the electricaly ranged between 101 and 104 S/cm while thefcient was relatively low (122 V/K) [91,93]
830 Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841
Fig. 8. Synthesis of PEDOT:PSS passivated Te nanorods, followed by formation of smooth nanocomposite lms during solution casting . 2010 Reproduced with the permission of American Chemical Society.
and a quasi-linear temperature dependence of Seebeckcoefcient (from RT to 4.2 K) was observed by Choi et al. in 1997. This is clearly different from metal-halide(FeCl3, NbCl5, and AuCl3) doped PA where the tempera-ture dependence of Seebeck coefcient was quasi-linearat high temperature but was independent of temperaturebelow 20 K with some lms showing a negative Seebeckcoefcient with a broad minimum peak. To clarify the causeof these differences, the Seebeck coefcient was measuredunder a high magnetic eld (up to 20 T) . The similar-ity of the dependence of the Seebeck coefcient on dopantconcentratispin interaspin in medopant loca
Alkali-mined . H(Rb0.17CH)Xpower lawwhile for KSeebeck coetivity of (K3.7 103 S/ranging fromand 260 K. Texplained bstate mode
The effects of anisotropy on Seebeck coefcient instretch-oriented PA (doped with MoC15, iodine and fer-ric chloride, respectively) have also been reported [93,95].For MoC15 doped PA, the electrical conductivity parallelto the stretching direction of the lm was found to be6.4 103 S/cm, and the ratio of the parallel/perpendicularwas 25 and independent of temperature. For iodineor ferric chloride, the orientation direction associatedwith a higher Seebeck coefcient was the same as thatof higher electrical conductivity. This phenomenon wasexplained by assuming there was intrinsic inhomogene-
ithin nly ched itllic thtion.thougucted y its isearchted.
Y has varie
Fig. 9. (a) 2010 Reprodon and the magnetic eld indicates that spin-ctions exist between the conduction electrontallic PA chains and the spin localized in theted nearby the chains.etal-doped PA systems have also been exam-ere, the electrical conductivity of (K0.14CH)X andsystems followed a temperature-dependent
((T)/(260 K) = A + BTn (A and B constants),-doped PA the temperature dependence of thefcient was quasi-linear. The electrical conduc-0.14CH)X doped PA ranged from 1.7 103 tocm while the Seebeck coefcient was quite low,
0.5 to 8.5 V/K for temperatures between 20he positive value of Seebeck coefcient can bey the quasi-one dimensional soliton condensedl.
ity wnot omodimetadirec
Alcondited bno rerepor
PPited amore SEM image of a drop-cast composite nanorod lm. (b)TEM image showing the cuced with the permission of American Chemical Society.the polymer and that mechanical stretchinganged inner geometry but also substantiallys structure. The stretched material was morean the pristine material and valid for either
h a signicant amount of research has beenon PA, its practical application is severely lim-nstability in air. To the best of our knowledge,
on PA-inorganic TE nanocomposites has been
a TE material
been prepared in various ways and has exhib-ty of properties. Chemically, it is signicantly
than PA. If it could be doped to obtain a highrystalline Te nanorod passivated with PEDOT:PSS .
Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841 831
electrical conductivity, it could be considered as a reason-ably good candidate for use as a TE material.
PPY lms have been prepared by Maddison et al. usingelectrochemical polymerization . The electrical con-ductivity onormally dcoefcient reaching 5trical condconductingcoefcient with the re
Conductreported byorophosphadisplayed dtemperaturtively high samples pra rise as tmum obserPPY lms pared by ancoefcient Seebeck codency from50 K. The rdrag effectand/or electstates .
To exameffect, PPY ccoefcient pared by dicooling effemer was oelectrical d
Althouglow electricTE materialon PPY-ino
5.3. PC as a
In doped(PIC) oxidized beSeebeck cocient will adcharge-carrthe level of charge locafor conduct2,7-carbazochains on tet al. . Htivity and was a TE madolocarbazodifferent siatom of the
polymer had a good Seebeck coefcient (max 290 V/Kat RT, doped with FeCl3) but its electrical conductivityremained low.
Because the poly(1,12-bis(carbazolyl)dodecane (P(2Cz-as hiophenrical ctions s
both ting imhtly h
(0.23only 0sses grical cmateri
inspg polyin 200holog
side cnits (
the cW m
rical coeck coaring tsides ylene thioful) [106OPV) oth ylenevxy) anted. Io
W mt 39.1
the S50 ased. F, the tion oentratithermstretchly es
0.1 (streSeebevely hF-TCN
electrf these lms were 826 S/cm for lightly tooped lms, respectively, while the Seebeckwas increased near linearly with temperature
V/K at 200 K but typically decreased as elec-uctivity increased as is the case with other
polymers. However, the change in Seebeckwith temperature was insignicant and agreedsults of Yan et al. .ivities of more than 300 S/cm have been
Sato et al. for tetramethylammonium hexau-te (PF6) doped PPY lms . The samplesramatically different behaviours at very lowes. The electrical conductivity remained rela-as temperature approached 0 K (especially inepared at low temperature) and even showedemperature was very close to 0 K (a mini-ved at 1520 K), which was consistent withdoped with hexauorophosphate (PF6) pre-
electrochemical method . The temperatureof conductivity changed sign at T = 17 K. Theefcient showed a linear temperature depen-
RT to 50 K, but displayed a positive hump beloweason for this might be explained by phonon
because of a greater degree of crystallinityron-phonon interactions in disordered metallic
ine the Seebeck coefcient and the Peltieroated fabrics were prepared . The Seebeckbetween two PPY coated fabric samples (pre-fferent methods) was found to be 10 V/K. Act due to the presence of the conductive poly-bserved but it was unsteady primarily due toegradation of the conducting polymer.h PPY is much more stable than PA, its relativelyal conductivity constrains its applications as a
and to the best of our knowledge, no researchrganic TE nanocomposites has been reported.
polycarbazoles (PC) and polyindolocarbazoles, charge is very localized because nitrogen isfore the backbone and should result in a largeefcient. However, this high Seebeck coef-versely affect electrical conductivity because ofier pinning. Therefore, it is important to balanceoxidation of the polymer to maintain a degree oflization but still allow enough charge mobilityivity . To this end, poly(N-octyl-3,6-dihexyl-lenevinylene) (PCVH) containing exible sidehe carbazole cycle was prepared by Levesqueowever this system had low electrical conduc-as unstable and thus was considered unsuitableterial. To overcome these limitations, polyin-le derivatives containing alkyl side chains andde chains (alkyl or benzoyl) on the nitrogen
backbone were synthesized . The resultant
D) hb]thielectvariawithity bea sligalonewas posseelecta TE
Annatinet al. morpalkyling ualong19 electSeebappe
BephentetraTCNQ(PMeof bphen(ethorepor7.1cienwhilerelatimuchiodinthe helectP(EtOthe stime,(25increlmsvariaconcThe the roughvalue313 Khigh relati
TThavegh mechanical property, while thieno[3,2-e (TT) has rigid -conjugated structure and highonductivity. Therefore, further attempts usinguch as P(2Cz-D-co-TT) have been reported he Seebeck coefcient and electrical conductiv-proved. These copolymer lms also exhibitedigher PF (0.33 W m1 K2) compared to PTT
W m1 K2) but their electrical conductivity.10.3 S/cm. Although this family of polymersood Seebeck coefcient, an improvement inonductivity is required for them to be used asal.iring result was obtained in a series of alter-(2,7-carbazole) derivatives synthesized by Aich9 . X-ray analyses showed that a structuredy could be obtained by introducing a secondaryhain on the carbazole unit as well as pack-e.g. benzene and benzothiadiazole moieties)onjugated backbone. A maximum PF value of1 K2 (doped lm) was obtained due to a highnductivity (up to 500 S/cm) and relatively highefcient (up to70 V/K) with the advantage ofo be stable in air.the above, the TE properties of polypara-[104,105], polyparaphenylene vinylene ,valenetetracyanoquinodimethane (TTF-110], poly(2,5-dimethoxyphenylenevinylene)and a series of copolymers consisting
unsubstituted and 2,5-dialkoxy-substitutedinylenes (P(ROPV-co-PV); RO = MeO, EtOd BuO (butoxy)) lms [111,112] have beendine-doped PMeOPV lms showed a high PF
1 K2 due to relatively high Seebeck coef- V/K and electrical conductivity 46.3 S/cm,e-doped P(ROPV-co-PV) lms exhibited alsoigh Seebeck coefcient (40.849.4 V/K), butr electrical conductivity (1.02.9 S/cm). Theed P(ROPV-co-PV) lms were stretched, sooriented lms were yield. As a result, theonductivities of both P(MeOPV-co-PV) and-PV) lms increased (from 1 to 350 S/cm) asing ration increased (from 1 to 5), at the sameeebeck coefcients remained relatively stableV/K). It is mainly because the carrier mobilityor the stretched iodine-doped P(BuOPV-co-PV)Seebeck coefcient varied inversely with af the electrical conductivity due to the carrieron being changed by the stretching treatment.al conductivity along stretching direction ofed iodine-doped P(EtOPV-co-PV) lms wastimated to be 0.25 W m1 K1, so a high ZTwas obtained for the P(EtOPV-co-PV) lms attching ratio being 3.1) because of relativelyck coefcient (47.3 V/K) accompanied withigh electrical conductivity (349.2 S/cm).Q was the rst organic compound shown toical conductivity. It is a charge transfer salt
832 Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841
Fig. 10. (a) Sc red lineSchematic of t ional netSEMs of the crline in panel (b(indicated by the SEMs indic 2008 Reprod
consisting with p -etronic wavmolecules igibly small
Table 2The , , as w
PANI-inorgaPTH [60,63PTH-inorganPEDOT:PSS [PEDOT-Tos [PEDOT:PSS-
[68,8386PPY [96,97,9PA [90,91,93PC [33,35,10PMeOPV [11P(ROPV-co-P
Others polym(polymer/of CNT [32hematic of CNTs suspended in an aqueous emulsion. Gray spheres and he emulsion-based composite after drying. The CNTs form a three-dimens
oss-sections of 5 wt.% CNT composites are shown in panels c and d after the comp)). The high-magnication SEM shown in panel d is a portion of the sample in c inarrows) are wrapped around the emulsion particles (indicated by yellow dottedate 1 m .uced with the permission of American Chemical Society.
of TTF and TCNQ planar molecules bondedlectron orbitals . The overlap of the elec-e function along the chain of the stackeds signicant while that between chains is negli-
consequently yielding highly anisotropic and
quasi-one-delectrical chigher thanbeck coefthe a- and
ell as the highest ZT value of a few typical polymer and polymer-inorganic TE n
37,39,40,42,43,4548] 107320 1622
nic TE nanocomposites [30,34,5159] 0140 306265,67] 102103 10100ic TE nanocomposites [70,71] 7.18.3 561287681] 0.06945 8888 82] 6 10-4300 40780 inorganic TE inorganic nanocomposites]
9,100,116118] 0340 140 95,119] 1.53 1032.85 104 0.512,103] 4.0 1055 102 4.9600 1] 46.3 39.1 V) (RO = MeO, EtO and BuO) [111,112] 183.5354.6 21.347.
er-inorganic TE nanocompositescarbon nanotube with different contents])
048 4050 s represent emulsion particles and CNTs, respectively. (b)work along the surfaces of the spherical emulsion particles.
osites were freeze-fractured (for instance, along the dotteddicated by a yellow solid square. It clearly shows that CNTs
lines) rather than homogenously mixed. The scale bars in
imensional conduction behaviour . Theonductivity along the b-axis was signicantly
that along the a- and c-axes , the See-cient was positive and negative parallel to
b-axes, respectively , and the thermal
(W/m K), maximum PF (W/m K2) orZT
5 , 0.020.542 ZTmax, 1.1 102 at423 K
6 , 0.251.2 , 0.0280.17 ZTmax, 2.9 102 at 250 K3 PFmax, 2.5 102
, 0.34 ZTmax, 1.0 102 at 300 K, 0.37 ZTmax, 0.25 at RT
67 , 0.220.4 ZTmax, 0.10 at RT
, 0.2 ZTmax, 3 102 at 423 K077
PFmax, 19PFmax, 7.1
3 , 0.250.80 (estimated) ZTmax,9.87 102 at 313 K, 0.180.34 ZTmax: 0.006 at RT
Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841 833
Table 3The TE properties of the polymers and polymer-inorganic TE nanocomposites covered in this review.
Year and authors Materials (S/cm) (V/K) (W/m K), PF(W/m K2) and ZTvalue
Part A The TE properties of polianiline (PANI) and PANI-inorganic TE nanocompositesWang et al.  HCl-doped PANI with
764 (30300 K)
Yoon et al. [39,40] PANI lms doped withCSA
012 (5300 K)
Sixou et al.  PANI lms doped withCSA, aging at differenttimes
0.5320 (RT) 17 (RT)
Mateeva et al.  PANI lms stretched2.5 times (parallel) anddoped with oxalic acid
103100 (300 K) 2590 (300 K)
PANI lms stretched2.5 times (parallel) anddoped with citric acid
104101 (300 K) 1025 (300 K)
Yan et al.  PANI lms doped withvarious protonic acid
104188 (300 K) , 0.020.25 (300 K)
ZT, 7 105103(300 K)
Toshima et al.  PANI lms doped withCSA and been stretched
PF, 15 (345423 K)
ZT, 1031.1 102(345423 K)
Liu et al.  PANI powder dopedwith HCl (prepared at293 K)
0.88 (303 K) 6.17 (303 K) , 0.542 (303 K)
ZT, 1.87 106 (303 K)PANI powder dopedwith HCl (prepared at273 K)
10.0 (303 K) 1.28 (303 K) , 0.538 (303 K)
ZT, 9.18 107 (303 K)Emeraldine basePANI(prepared at293 K) redoped withHCl
3.7 (303 K) 177 (303 K) , 0.530 (303 K)
ZT, 6.63 103 (303 K)Emeraldine basePANI(prepared at293 K) redoped withtoluene-psulfonic acid
2.86 (303 K) 17.96 (303 K) , 0.354 (303 K)
ZT, 7.90 105 (303 K)Yakuphanoglu et al.  PANI doped with HCl 107105
(298413 K)1693 (303416 K)
Li et al.  PANI doped with HCl 06 (303423 K) 555 (303423 K) , 0.2760.34(303423 K)ZT,0.1 1042.67 104(303423 K)
Sun et al.  PANI nanotubes dopedwith -Naphthalenesulfonic acid
7.7 103 300 K 150225(180300 K)
, 0.170.22(200320 K)
ZT, 4.86 105 (300 K)Wu et al.  (PANI)XV2O5nH2O 104101 (RT) 30200 (RT)Zhao et al.  PANI powders doped
with HClO42026 (300388 K) 16 (300388 K)
PANI/Bi0.5Sb1.5Te3composites withdifferent contents ofPANI
30100 (300388 K) 160178 (300388 K) PF, 90300(300388 K)
Liu et al.  PANI/NaFe4P12 whiskercomposites
523 (313473 K)
1528 (313473 K)
PANI 0.100.20(293473 K)
1726 (313473 K)
834 Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841
Table 3 (Continued )
Year and authors Materials (S/cm) (V/K) (W/m K), PF(W/m K2) and ZTvalue
Hostler et al.  PANI/Binanocomposite lmswith different contentsof 10 nm Bi particles
0140 (RT) 914 (RT) , 0.250.7 (RT)
ZT,0.4 1031.3 103(RT)
Anno et al.  PANI lms doped withCSA
102 (50300 K) 510 (100 300 K)
PANI/Binanocomposite lmsdoped with CSA
102101 (RT400 K) 1254 (180400 K)
Meng et al.  PANI-coated CNT sheetwith different contentsof PANI
3090 (300 K) 1228 (300 K) , 0.390.5 (300 K)
PF, 0.55 (300 K)Yao et al.  PANI/SWNT
composites withdifferent contents ofSWNT
10125 (RT) 1140 (RT) , 0.51 (RT)
PF, 020 (RT)ZT, 0.004 (RT)
Toshima et al.  PANI/Bi2Te3 compositelms prepared byphysical mixturemethod
60 (350 K) 110 (350 K) PF, 5.1 (350 K)ZT, 0.18 (350 K)(estimated)
PANI/Bi2Te3 compositelms prepared bysolution mixturemethod
2 (350 K) 130 (350 K) PF, 2.6 (350 K)ZT, 0.009 (350 K)(estimated)
Our group  PANI/PbTe compositenanopowders
626578 (293373 K) PF, 0.7130.757
(293373 K)Wang et al.  HClO4-doped
PANI/graphitecomposites withdifferent contents ofgraphite
1.23120 (303393 K) 0.8218.66(303393 K)
, 0.291.2 (273323 K)
ZTmax, 1.37 103(393 K)
Li et al.  PANI/Bi2Te3 compositetablets
2 (300473 K) 50 (300473 K)
Part B The TE properties of Polythiophene (PTH) and PTH-inorganic TE nanocompositesGao et al.  PTH 102103 Doped with
PF622 Doped with FeCl3
Shinohara et al.  PTH lms synthesizedby electrolyticpolymerization
102102 (RT) 101102 (RT)
Hiraishi et al.  PTH lms weresynthesized byelectrolyticpolymerization
201 (RT) 23 (RT) The authors assumedthe is 0.1.
PF, 10.3 (RT)Lu et al.  PTH lms prepared by
1547 (100320 K) 2844 (100320 K) , 0.0280.17(100320 K)
ZT,0.4 1022.9 102(100320 K)
Poly(3-methylthiophene)(PMeT) lms preparedby electrochemicalpolymerization
4773 (100320 K) 1732 (100320 K) , 0.020.15(100320 K)
ZT,0.8 1023.1 102(100320 K)
Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841 835
Table 3 (Continued )
Year and authors Materials (S/cm) (V/K) (W/m K), PF(W/m K2) and ZTvalue
Sun et al.  P3HT blend withF4TCNQ
1.75 1043.75 104(RT)
400580 (RT) ,0.170.48
PF,4.876 1036.053 103ZT,3.03 1061.06 105(RT)
P3HT blend withP3HTT and F4TCNQ
2.0 1051.4 103(RT)
200700 (RT) ,0.170.48
PF,4.62 1047.58 103ZT,2.87 1072.03 105(RT)
Yue et al.  Polythieno[3,2-b]thiophene (PTT)lms synthesized byelectrolyticpolymerization
0.021.5 (100306 K) 3885 (140306 K) The authors estimatedthe ZT values of PTTlms based on of PTHlms , 0.030.17;(100320 K)PF, 0.051.1(140306 K)ZT, 0.012.3 103(140306 K)
Pinter et al.  Poly(3-octylthiophene)(P3OT) tabletcompressed from itspowder
Du et al.  PTH/Bi2Te3 TE bulknanocomposites hotpressed at 623 K
7.18.3 (298473 K) 56 to 46(298473 K)
2,1.7 1022.5 102(298473 K)
Part C The TE properties of PEDOT:PSS and PEDOT:PSS-inorganic TE nanocompositesJiang et al.  PEDOT:PSS pellets 48 (150300 K) 1115 (150300 K) ZT, 4 104
(150300 K)PEDOT:PSS pellets(doped with DMSO)
1847 (150300 K) 815 (150300 K) ZT,4 1041.75 103(150300 K)
PEDOT:PSS pellets(doped with ethyleneglycol)
2855 (150300 K) 813 (150300 K) ZT,6 1041.75 103(150300 K)
Chang et al.  PEDOT:PSS lms(doped with DMSO)
0.06224 13888 PF, 0.044.78
Scholdt et al.  PEDOT:PSS(commercial product,Clevios PH750) lms(doped with DMSO)
570 (RT) 13.5 (RT) , 0.34 (RT)
PF, 10.4 (RT)ZT, 9.2 103 (RT)
Kong et al.  PEDOT:PSS lms(doped with urea)
263.13 (100300 K) 520.7 (100300 K) PF, 02.71(100300 K)****
Liu et al.  PEDOT:PSS lms(doped with EG)
150240 (100300 K) 413 (200300 K) ZT, 1 1037 103(200300 K)
PEDOT:PSS lms(doped with DMSO)
180298 (100300 K) 714 (200300 K) ZT, 5 1031 102(200300 K)
Taggart et al.  PEDOT nanowires 6.940.5 (310 K) 122 to 35 (310 K) PF, 1.212 (310 K)PEDOT lms 3.218.3 (310 K) 57 to 34 (310 K) PF, 0.634.4 (310 K)
Bubnova et al.  PEDOT-Tos lms 6 104300 (RT) 40780 (RT) , 0.37ZT, 00.25 (RT)
Kim et al.  PEDOT:PSS/SWCNT 280400 (RT) 2125 (RT) The authors estimated is 0.4 W/m KPF, 1425 (RT)ZT, 0.02 (RT) (35wt.%SWCNT and 35 wt.%PEDOT:PSS)
PEDOT:PSS/CNT 0124 (RT) 1740 V/K (RT) , 0.260.38 (RT)PF, 111 (RT)
836 Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841
Table 3 (Continued )
Year and authors Materials (S/cm) (V/K) (W/m K), PF(W/m K2) and ZTvalue
Zhang et al.  PEDOT:PSS(CleviosPH1000) doped withdifferent contents ofDMSO
0945 2050 PF, 047
PEDOT:PSS(CleviosPH1000)/Bi2Te3 withdifferent volumefractions of CleviosPH100(for P typeBi2Te3 with HClrinsing)
60150 60150 PF, 60131
PEDOT:PSS(CleviosPH1000)/Bi2Te3 withdifferent volumefractions of CleviosPH100(for N typeBi2Te3 with HClrinsing)
55250 125 to 0 PF, 080
See et al.  PEDOT:PSS/Te lms 19.3 ( 2.3) (RT) 163 ( 4) (RT) PF, 70.9 (RT), 0.220.30 (RT)ZT, 0.10 (RT)
Liu et al.  PEDOT:PSS/Ca3Co4O9(with differentcontents ofCa3Co4O9)
50135 (100300 K) 118 (100300 K) PF, 0.14 (100300 K)
Wang et al.  PEDOT/PbTe (withdifferent contents ofPbTe)
0.0640.616 (RT) 12054088 (RT) PF, 1.071.44 (RT)
Part E The TE properties of the others polymer and polymer-inorganic nanocompositesPark et al.  FeCl3-doped PA
(6.6%)28500 (RT) 18.4 (RT)
6100 (RT) 13.5 (RT)
375 (RT) 17.3 (RT)
ZrCl4-doped PA (1.2%) 4.3 (RT) 28.7 (RT)WCl6-doped PA(1.1%)
4.3 (RT) 46.2 (RT)
0.06 (RT) 84.9 (RT)
Yoon et al.  PA lms (doped withMoC15)
1.53 1039580 (RT) 11.41077 (RT)
Kaneko et al.  PA lms (doped withiodine and aging withdifferent times)
101104 (4.2300 K) 118 (4.2300 K)
Pukacki et al.  Stretched PA samples(doped with iodineand ferric chloride)
525010,200 (300 K)parallel 92223 (300 K)perpendicular
122 (4.2300 K)
Park et al.  PA (doped with K) 16743720(50300 K)
0.58.5 (20 -300 K)
Choi et al.  PA lms (doped withAuC13)
103104 (1.5300 K) 0.5 to 10 (5120 K)
Maddison et al.  PPY lms prepared byelectrochemicalpolymerization(doped withp-toluenesulfonateanion (PPpTS))
26 (300 K) 6 (300 K)
Maddison et al.  PPY lms prepared byelectrochemicalpolymerization(doped with (PPpTS))
053 (100290 K) 140 (100290 K)
Lee et al.  PPY lms prepared byelectrochemicalpolymerization(doped withhexauorophosphate(PF6))
85153 (0300 K) 0.27.14 (0300 K)
Kemp et al.  PPY samplesprepared by differentmethods
0.01340 (4.2300 K) -116 (4.2300 K)
Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841 837
Table 3 (Continued )
Year and authors Materials (S/cm) (V/K) (W/m K), PF(W/m K2) and ZTvalue
Yan et al.  PPY lms prepared byelectrochemicalpolymerization (dopedwith PPpTS)
55175 (300 K) , 0.2 (300 K)
PF, 2 (300 K)ZT,1.8 1023 102(300423 K)
Hu et al.  PPY powder preparedby oxidativepolymerization
100 (RT) 10 V/K (bres coatedwith PPY (RT))
Kemp et al.  PF6 doped PPY lmsprepared at differenttemperatures (afterexposure to ammonia)
0200 (4.2300 K) 012 (4.2300 K)
Levesque et al.  poly(N-octyl-3,6-dihexyl-2,7-carbazolenevinylene)(PCVH) lms dopedwith FeCl3
200600 (RT) PF,5 1037.5 102(RT)
poly(3-decylthiophene-2,5-diyl) lms dopedwith FeCl3
10150 (RT) PF, 5 1021.4 (RT)
Levesque et al.  Poly[(3,6-dihexyl)2,7-carbazole] derivativespellet doped with FeCl3
4.5 103 (RT) 55 (RT) PF, 1.4 103 (RT)
Poly(2,7-N-hexylbenzoyl)carbazolederivatives pelletdoped with FeCl3
1.2 1020.29 (RT) 6171 (RT) PF,4.6 1031.5 101(RT)
Polyindolocarbazolederivatives pelletdoped with FeCl3
2.7 1040.29 (RT) 4.9290 (RT) PF,1.0 1041.2 101(RT)
Aich et al.  poly[N-9-heptadecanyl-2,7-carbazole-alt-5,5-(4,7-di-2-thienyl-2,1,3-benzothiadiazole](PCDTBT) lms dopedwith FeCl3
0500 1575 PF, 419
Yue et al.  Poly(1,12-bis(carbazolyl)dodecane-co-thieno[3,2-b]thiophene) lmssynthesized with thedifferent ratios ofTT/2Cz-D
4.0 1050.26(200 K- RT)
66169 (200 K- RT) PF, 0.020.33(200 KRT)
Hiroshige et al.  PMeOPV lms dopedwith iodine
46.3 39.1 PF, 7.1 (313 K)
Hiroshige et al. [111,112] P(MeOPV-co-PV) lmsdoped with iodine(stretching ratio 4.4)
183.5 43.5 , 0.8 (estimated)
ZT, 1.36 102 (313 K)P(EtOPV-co-PV) lmsdoped with iodine(stretching ratio 3.1)
349.2 47.3 , 0.25 (estimated)
ZT, 9.87 102 (313 K)P(BuOPV-co-PV) lmsdoped with iodine(stretching ratio 4.4)
354.6 21.3 , 0.75 (estimated)
ZT, 0.67 102 (313 K)Yu et al. Polymer/Carbon
nanotube withdifferent contents ofCNT
048 (300 K) 4050 (300 K) , 0.180.34 (300 K)
ZT, 0.006 (RT)
838 Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841
conductivity was estimated to be 1 W/m K at RT byextrapolation. Because growing a large TTF-TCNQ singlecrystal is difcult and the crystal mechanically fragile,TTF-TCNQ thin lms were fabricated and TE propertiesmeasured b7.8 101may have p
Apart fnanocompoCNT-polymemulsion) The electridramaticallthe compocoefcient rtion. This bbut electricto tune proconcentratielectrical co0.34 W/mKthe CNTs sformed a ththe sphericaof the crossFig. 10(c) ahigh-magnithe sample
This wographene adates for fupolymer co
Table 2 coefcient, of a numbpolymer-incomprehenmers and pthis review
In summber of key polymer anthree decadundergone ductors andperformancused as TE constraints
To date,polymer beeasy preparto PANI becand good sreported onand is insolimental stusignicantlstable than
limits its potential use as a TE material. A promising poly-mer may be PC or its derivatives as it has good stability, ahigh electrical conductivity and Seebeck coefcient.
To improve the TE properties of polymers, it is primar-portature ang leve
simurovidr consy and nce th
decreavely lostructue ller
TE nnce thase thlectricositesis critifacial c
to mastructueen thnostrrom thars to organ
ltaneoonentthe coosite smputble polthat thtimize
coatinods fon of thcularlyt andpriate
to be dnally, when nterfaces and ing be
Key Laesis any Tamayo et al.  in 2010 (maximum PF isW m1 K2). While preliminary, this approachotential for application in TE devices.rom PANI-, PTH-, PEDOT: PSS-inorganicsites reviewed above, segregated-networker (poly(vinyl acetate) (PVAc) homopolymercomposites have also been reported .cal conductivity of these composites can bey increased by creating a network of CNTs insite while thermal conductivity and Seebeckemain relatively insensitive to ller concentra-ehaviour results from thermally disconnectedally connected junctions making it feasibleperties in favour of a higher ZT. With a CNTon of 20 wt.%, these composites exhibit anductivity of 48 S/cm, thermal conductivity of
and a ZT greater than 0.006 at RT. Fig. 10 showsuspended in an aqueous emulsion (a), andree-dimensional network along the surfaces ofl emulsion particles (b) schematically, and SEM-sections of 5 wt.% CNT composites is shown infter the composites was freeze-fractured. Thecation SEM shown in Fig. 10(d) is a portion ofin (c) indicated by a yellow solid square.rk suggests that carbon materials such asnd carbon nanotubes may be promising candi-rther development of lightweight and low-costmposites for TE applications.
ry and conclusions
presents the electrical conductivity, Seebeckthermal conductivity and the highest ZT valueer of conducting polymer and conductingorganic nanocomposites. Table 3 presents asive summary of the TE properties of the poly-olymer-inorganic nanocomposites covered in
for reference purpose.ary, the review presented here provides a num-ndings to guide and focus future research ond polymer-inorganic TE nanocomposites. Afteres of development, conductive polymers havea series changes from insulators to semicon-
conductors and have a number of excellente characteristics. However, before they can bematerials they need to overcome a number of.
PANI has been the most studied conductivecause of its high conductivity, good stability,ation and easy processing. PEDOT: PSS is secondause of its relatively high electrical conductivitytability whereas only a few studies have been
PTH which has a low electrical conductivityuble and infusible. Many theoretical and exper-dies have reported on PA but its usefulness isy limited by its oxidation instability. PPY is more
PA but its relatively low electrical conductivity
ily imstrucdopinwhilethat pOtheabilit
Sisionsrelatinanoas thganicenhaincrethe ecomp
It interturesnanobetwTE naing. Fappethe insimucompmer, comp
Cosuitasuch is opMore
Pothe mspin methzatioparticienapproneed
Fiered the iphasbond
ThFounStateSynthnt to enhance their crystallinity, control theird surface morphology as well as optimizing thel required to achieve the balance of propertiesltaneously engineering an electronic structurees a favourable density of states and Fermi level.iderations include thermal stability, process-mechanical properties.e ZT value of a material increases as its dimen-se and 1D nanostructures can form networks atw volume fractions, it is preferable to use 1D TEres such as nanowires, nanotubes or nanobelts
in TE composites. Moreover, coating the inor-anostructure with a polymeric nanolayer toe size-dependent energy-ltering effect and toe carrier mobility can simultaneously enhanceal conductivity and Seebeck coefcient of the.cal to homogenously disperse and obtain a goodonnection between polymer and TE nanostruc-ximize desired properties when using 1D TEres. In order to reduce the interface resistancee two phases, the oxidation of the inorganicuctures must be avoided during the process-is viewpoint, in situ interfacial polymerizationbe a good method to homogenously disperseic nanostructures within polymer matrix andusly prevent their oxidation.As inorganic TE
usually has higher TE performance than poly-ntent of inorganic TE nanostructures in thehould be as high as possible.ational modelling can assist in the selection ofymer matrices and inorganic TE nanostructurese interplay and volume fraction of each phased to maximize the ZT value of the composite.
on modelling would be expected.-inorganic nanostructures composite lms areitable form for application and processing by
g, drop casting, and spraying are the preferredr preparing such lms. However, characteri-e TE properties of lms can be problematic
in regard to the out-of-plane Seebeck coef- electrical and thermal conductivities and
measurement techniques and methodologieseveloped.some deeper questions need to be consid-studying these systems especially in regard toial region between the inorganic and organicunderstanding the nature of bonding or pseudotween the inorganic and organic phases.
rk was supported by National Natural Science of China (50872095) and the foundation of theboratory of Advanced Technology for Materialsd Processing, Wuhan University of Technology.
Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841 839
The authors would like to thank China Scholarship Councilfor the nancial support for Yong Dus study at CSIRO.
 Bell LE.heat wi
 Harmandot sup2002;29
 Boukai JR. Silic2008;45
 Chung M, Uhemoelec2000;28
 HochbaM, Majurough s
 Hicks LDthermo
 Hicks Ldimens
 Hicks LDthe effeof meri
 Venkatathermomerit. N
 Lin YMsemiconthermo
 Larson structutemper
 Hsu KFPolychrthermo2004;30
 Yang Rple andPhys Re
 Kim W,dar A. Tmerit inductors
 Ohta HT, Nakathermoin SrTiO
 DresselZF, Fleumoelec
 Cao YQmoeleclaminat
 Poudel Muto AZF. Higantimo
 Zhou MmateriaSoc 200
 Yang SHin high-rials. Na
 Li H, Tang XF, Zhang QJ, Uher C. High performance InxCeyCo4Sb12thermoelectric materials with in situ forming nanostructured InSbphase. Appl Phys Lett 2009;94, 102114/1-3.
 Xie WJ, Tang XF, Yan YG, Zhang QJ, Tritt TM. Unique nanostructuresand enalloys. AXie WJ,JRD, Brotructur(Bi,Sb)2KashiwAdachi antimoLi JJ, Tatric pro2010;1Yao Q,thermotubes/p2010;4ToshimmateriaYu C, ior of s2008;8LevesquRatcliffmoelecpolydiinLiu H, Wture ancomposLevesquHighly cal app2005;6Ikkala On thecally co98556Liu J, Zpropert2003;1Menon line neaRev B 1Yoon Cpower Synth MYoon COprocess1993;4Yan H, ered lmpolyaniSixou Bder on 7034.Wang and EPductor 41902Li QM, Ction in Yakuphity, thesoluble2008;3MateevseebeckpolypyrYan H, electricmateriaSun YNin thermtures. SYan H, ((+/))-thermons JP, Jovovic V, Toberer ES, Saramat A, Kurosaki K, Charoen-e A, Yamanaka S, Snyder GJ. Enhancement of thermoelectriccy in PbTe by distortion of the electronic density of states.
2008;321:5547. Cooling, heating, generating power, and recovering wasteth thermoelectric systems. Science 2008;321:145761.
TC, Taylor PJ, Walsh MP, LaForge BE. Quantumerlattice thermoelectric materials and devices. Science7:222932.
FJ. Thermoelectric cooling and power generation. Science5:7036.AI, Bunimovich Y, Tahir-Kheli J, Yu JK, Goddard WA, Heathon nanowires as efcient thermoelectric materials. Nature1:16871.
dar A. Thermoelectricity in semiconductor nanostructures. 2004;303:7778.DY, Hogan T, Brazis P, Rocci-Lane M, Kannewurf C, Bastear C, Kanatzidis MG. CsBi4Te6: a high-performance ther-tric material for low-temperature applications. Science7:10247.um AI, Chen RK, Delgado RD, Liang WJ, Garnett EC, Najarianmdar A, Yang PD. Enhanced thermoelectric performance ofilicon nanowires. Nature 2008;451:1637., Dresselhaus MS. Effect of quantum-well structures on the
electric gure of merit. Phys Rev B 1993;47:1272731.D, Dresselhaus MS. Thermoelectric gure of merit of a one-ional conductor. Phys Rev B 1993;47:166314., Harman TC, Sun X, Dresselhaus MS. Experimental study of
ct of quantum-well structures on the thermoelectric guret. Phys Rev B 1996;53:104936.subramanian R, Siivola E, Colpitts T, OQuinn B. Thin-lmelectric devices with high room-temperature gures ofature 2001;413:597602.
, Rabin O, Cronin SB, Ying JY, Dresselhaus MS. Semimetal-ductor transition in Bi1 xSbx alloy nanowires and theirelectric properties. Appl Phys Lett 2002;81:24035.P, Mahanti SD, Chung DY, Kanatzidis MG. Electronicre of CsBi4Te6: a high-performance thermoelectric at lowatures. Phys Rev B 2002;65, 045205/1-5., Loo S, Guo F, Chen W, Dyck JS, Uher C, Hogan T,oniadis EK, Kanatzidis MG. Cubic AgPbmSbTe2 + m: bulkelectric materials with high gure of merit. Science3:81821.
G, Chen G, Dresselhaus MS. Thermal conductivity of sim- tubular nanowire composites in the longitudinal direction.v B 2005;72, 125418/1-7.
TC, Walsh MP, Laforge BE, Turner GW. Nanostructuredelectric materials. J Electron Mater 2005;34:L1922.
Zide J, Gossard A, Klenov D, Stemmer S, Shakouri A, Majum-hermal conductivity reduction and thermoelectric gure ofcrease by embedding nanoparticles in crystalline semicon-. Phys Rev Lett 2006;96, 045901/1-4., Kim S, Mune Y, Mizoguchi T, Nomura K, Ohta S, Nomuranishi Y, Ikuhara Y, Hirano M, Hosono H, Koumoto K. Giantelectric Seebeck coefcient of two-dimensional electron gas3. Nat Mater 2007;6:12934.haus MS, Chen G, Tang MY, Yang RG, Lee H, Wang DZ, Renrial JP, Gogna P. New directions for low-dimensional ther-tric materials. Adv Mater 2007;19:104353., Zhao XB, Zhu TJ, Zhang XB, Tu JP. Syntheses and ther-tric properties of Bi2Te3/Sb2Te3 bulk nanocomposites withed nanostructure. Appl Phys Lett 2008;92, 143106/1-3.B, Hao Q, Ma Y, Lan YC, Minnich A, Yu B, Yan XA, Wang DZ,, Vashaee D, Chen XY, Liu JM, Dresselhaus MS, Chen G, Renh-thermoelectric performance of nanostructured bismuthny telluride bulk alloys. Science 2008;320:6348., Li JF, Kita T. Nanostructured AgPbmSbTem + 2 system bulkls with enhanced thermoelectric performance. J Am Chem8;130:452732., Zhu TJ, Sun T, He J, Zhang SN, Zhao XB. Nanostructuresperformance (GeTe)x(AgSbTe2)100 x thermoelectric mate-notechnology 2008;19, 245707/1-5.
 hanced thermoelectric performance of melt-spun BiSbTeppl Phys Lett 2009;94, 101111/1-3.
He J, Kang HJ, Tang XF, Zhu S, Laver M, Wang SY, Copleywn CM, Zhang QJ, Tritt TM. Identifying the specic nanos-
es responsible for the high thermoelectric performance ofTe3 nanocomposites. Nano Lett 2010;10:32839.agi M, Hirata S, Harada K, Zheng YQ, Miyazaki K, Yahiro M,C. Enhanced gure of merit of a porous thin lm of bismuthny telluride. Appl Phys Lett 2011;98, 023114/1-3.ng XF, Li H, Yan YG, Zhang QJ. Synthesis and thermoelec-perties of hydrochloric acid-doped polyaniline. Synth Met60:11538.
Chen LD, Zhang WQ, Liufu SC, Chen XH. Enhancedelectric performance of single-walled carbon nano-olyaniline hybrid nanocomposites. ACS Nano:244551.a N. Conductive polymers as a new type of thermoelectricl. Macromol Symp 2002;186:816.Kim YS, Kim D, Grunlan JC. Thermoelectric behav-egregated-network polymer nanocomposites. Nano Lett:442832.e I, Bertrand PO, Blouin N, Leclerc M, Zecchin S, Zotti G,
e CI, Klug DD, Gao X, Gao FM, Tse JS. Synthesis and ther-tric properties of polycarbazole, polyindolocarbazole, anddolocarbazole derivatives. Chem Mater 2007;19:212838.ang JY, Hu XB, Boughton RI, Zhao SR, Li Q, Jiang MH. Struc-d electronic transport properties of polyaniline/NaFe4P12ite. Chem Phys Lett 2002;352:18590.e I, Gao X, Klug DD, Tse JS, Ratcliffe CI, Leclerc M.soluble poly(2,7-carbazolenevinylene) for thermoelectri-lications: from theory to experiment. React Funct Polym5:2336.OT, Pietila LO, Ahjopalo L, Osterholm H, Passiniemi PJ.
molecular recognition and associations between electri-nducting polyaniline and solvents. J Chem Phys 1995;103:3.hang LM, He L, Tang XF. Synthesis and thermoelectricies of polyaniline. J Wuhan Univ Technol Mater Sci Ed8:535.R, Yoon CO, Moses D, Heeger AJ, Cao Y. Transport in polyani-r the critical regime of the metal-insulator-transition. Phys993;48:1768594.O, Reghu M, Moses D, Cao Y, Heeger AJ. Thermoelectric-of doped polyaniline near the metal-insulator-transition.et 1995;69:2734., Reghu M, Moses D, Heeger AJ, Cao Y. Counterion-inducedibility of polyanilinethermoelectric-power. Phys Rev B8:140804.Toshima N. Thermoelectric properties of alternatively lay-s of polyaniline and (+/)-10-camphorsulfonic acid-doped
line. Chem Lett 1999:12178., Travers JP, Nicolau YF. Effect of aging induced disor-transport properties of PANI-CSA. Synth Met 1997;84:
ZH, Scherr EM, Macdiarmid AG, Epstein AJ. TransportR studies of polyanilinea quasi-one-dimensional con-with 3-dimensional metallic states. Phys Rev B 1992;45:02.ruz L, Phillips P. Granular-rod model for electronic conduc-polyaniline. Phys Rev B 1993;47:18405.anoglu F, Senkal BF, Sarac A. Electrical conductiv-rmoelectric power, and optical properties of organo-
polyaniline organic semiconductor. J Electron Mater7:9304.a N, Niculescu H, Schlenoff J, Testardi LR. Correlation of
coefcient and electric conductivity in polyaniline androle. J Appl Phys 1998;83:31117.Sada N, Toshima N. Thermal transporting properties ofally conductive polyaniline lms as organic thermoelectricls. J Therm Anal Calorim 2002;69:8817., Wei ZM, Xu W, Zhu DB. A three-in-one improvementoelectric properties of polyaniline brought by nanostruc-
ynth Met 2010;160:23716.Ohta T, Toshima N. Stretched polyaniline lms doped by10-camphorsulfonic acid: Anisotropy and improvement ofelectric properties. Macromol Mater Eng 2001;286:13942.
840 Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841
 Bhadra S, Khastgir D, Singha NK, Lee JH. Progress in prepara-tion, processing and applications of polyaniline. Prog Polym Sci2009;34:783810.
 Wu CG, DeGroot DC, Marcy HO, Schindler JL, Kannewurf CR, Liu YJ,Hirpo W, Kanatzidis MG. Redox intercalative polymerization of ani-line in V2O5 xerogel. The postintercalative intralamellar polymergrowthmolecu
 Anno Hmura RPolyaniMilling.
 HostlerBurda Cfor theNanocoConf; 20
 Zhao XBi0.5Sb1Mater L
 Li Y, ZhBi2Te3/doi:10.1
 Toshimbrids aline an2010;42
 Meng Cmoelec2010;22
 Wang Ythermotructure
 Wang Lerties o2011;65
 Gao X,Theoretlow-ban7.
 Kresse transiti
 Shinohaeffect opropertTwentyYork: IE
 Yue RRand thepolythie2011;15
 Shinohalenge oChandr2006, P23293
 Lu BY, moelecpoly(3-057201
 Zhang electricBi2Te331708
 Sun J, YSimultaa dopedMacrom
 Pinter Eization by solut
 Du Y, Catric pro
Conference on Mechanical, Industrial, and Manufacturing Engi-neering (MIME 2011); 2011. pp. 4625.
 Majid K, Tabassum R, Shah AF, Ahmad S, Singla ML. Compara-tive study of synthesis, characterization and electric properties ofpolypyrrole and polythiophene composites with tellurium oxide. JMater Sci Mater Electron 2009;20:95866.PlotinskG. TelluCarpathDu Y, Cture on Bi2Te3 Kim JY, of poly(a changChang Peng YCoxythioMater 2Jiang Fmoelecpoly(styScholdtWinkleapplicaLiu CCJK, Liethylenlms. J Kong FGD, Wconducethylenwith urTaggartthermonanowiBubnovCrispin conduc2011;10Kim D,moelecwith pAcs NanSee KCRA. WaelectricLiu CC, PEDOT-materiaWang YpropertnanotubChiang EJ, Gau acetylenShirakaSynthesderivati57880HowellBethesd1994. hPark YWConducpolyaceKanekoand theiodine-Choi ESelectric1997;84Pukackipower 1994;62Park EBthermoMet 199 in polyaniline/metal oxide nanocomposites is facilitated bylar oxygen. Chem Mater 1996;8:19922004., Fukamoto M, Heta Y, Koga K, Itahara H, Asahi R, Sato-, Sannomiya M, Toshima N. Preparation of Conductingline-Bismuth Nanoparticle Composites by Planetary Ball
J Electron Mater 2009;38:14439. SR, Kaul P, Day K, Qu V, Cullen C, Abramson AR, Qiu XF,. Thermal and electrical characterization of nanoconipositesrmoelectrics. Thermal and Electrical Characterization ofmposites for Thermoelectrics ITHERM 06 10th Intersoc.06. pp. 1400-5.
B, Hu SH, Zhao MJ, Zhu TJ. Thermoelectric properties of.5Te3/polyaniline hybrids prepared by mechanical blending.ett 2002;52:1479.ao Q, Wang YG, Bi K. Synthesis and characterization ofpolyaniline composites. Mater Sci Semicond Process 2011,016/j.mssp.2011.02.019.a N, Imai M, Ichikawa S. Organic-inorganic nanohy-s novel thermoelectric materials: hybrids of polyani-d bismuth(III) telluride nanoparticles. J Electron Mater:898902.Z, Liu CH, Fan SS. A promising approach to enhanced ther-tric properties using carbon nanotube networks. Adv Mater:5359.Y, Cai KF, Yin JL, An BJ, Du Y, Yao X. In situ fabrication andelectric properties of PbTe-polyaniline composite nanos-s. J Nanopart Res 2011;13:5339., Wang DG, Zhu GM, Li JQ, Pan F. Thermoelectric prop-f conducting polyaniline/graphite composites. Mater Lett:10868.
Uehara K, Klug DD, Patchkovskii S, Tse JS, Tritt TM.ical studies on the thermopower of semiconductors andd-gap crystalline polymers. Phys Rev B 2005;72, 125202/1-
bilt D. Soft self-consistent pseudopotentials in a generalizedlue formalism. Phys Rev B 1990;41:78925.G, Hafner J. AB-initio molecular-dynamics for open-shellon-metals. Phys Rev B 1993;48:131158.i K, Masuhara A, Nakanishi H, Oikawa H, Shinohara Y.ion of thermoelectric properties of polythiophene lms syn-
by electrolytic polymerization. Jpn J Appl Phys 2009;48,/1-4.ra Y, Isoda Y, Imai Y, Hiraishi K, Oikawa H, Nakanishi H. Thef carrier conduction between main chains on thermoelectricies of polythiophene. In: Kim I, editor. Proceedings ICT 07:-Sixth International Conference on Thermoelectrics. NewEE; 2008. p. 4102., Chen S, Lu BY, Liu CC, Xu JK. Facile electrosynthesisrmoelectric performance of electroactive free-standingno[3,2-b]thiophene lms. J Solid State Electrochem:53948.ra Y, Imai Y, Isoda Y, Hiraishi K, Nakanishi H. A new chal-f polymer thermoelectric materials as ecomaterials. In:a T, Tsuzaki K, Militzer M, Ravindran C, editors. THERMECts. 15. Stafa-Zurich: Trans Tech Publications Ltd.; 2007. p.2.Liu CC, Lu S, Xu JK, Jiang FX, Li YZ, Zhang Z. Ther-
tric performances of free-standing polythiophene andmethylthiophene) nanolms. Chin Phys Lett 2010;27,/1-4.B, Sun J, Katz HE, Fang F, Opila RL. Promising thermo-
properties of commercial PEDOT:PSS materials and theirpowder composites. ACS Appl Mater Interfaces 2010;2:.eh ML, Jung BJ, Zhang B, Feser J, Majumdar A, Katz HE.neous increase in seebeck coefcient and conductivity in
poly(alkylthiophene) blend with dened density of states.olecules 2010;43:2897903., Fekete ZA, Berkesi O, Makra P, Patzko A, Visy C. Character-of Poly(3-octylthiophene)/silver nanocomposites preparedion doping. J Phys Chem C 2007;111:118728.i KF, Qin Z, Shen SZ, Casey PS. Preparation and thermoelec-perties of Bi2Te3/Polythiophene nanocomposite materials.
 aya OY, Damian F, Prokoev VY, Kovalenker VA, Damianrides occurrences in the baia mare region, Romania.ian J Earth Environ Sci 2009;4:89100.ai KF, Li H, An BJ. The inuence of sintering tempera-the microstructure and thermoelectric properties of n-typexSex nanomaterials. J Electron Mater 2011;40:51822.Jung JH, Lee DE, Joo J. Enhancement of electrical conductivity3,4-ethylenedioxythiophene)/poly(4-styrenesulfonate) bye of solvents. Synth Met 2002;126:3116.KC, Jeng MS, Yang CC, Chou YW, Wu SK, Thomas MA,. The thermoelectric performance of poly(3,4-ethylenediphene)/poly(4-styrenesulfonate) thin lms. J Electron009;38:11828.X, Xu JK, Lu BY, Xie Y, Huang RJ, Li LF. Ther-tric performance of poly(3,4-ethylenedioxythiophene):renesulfonate). Chin Phys Lett 2008;25:22025.
M, Do H, Lang J, Gall A, Colsmann A, Lemmer U, Koenig JD,r M, Boettner H. Organic semiconductors for thermoelectrictions. J Electron Mater 2010;39:158992., Jiang FX, Huang MY, Yue RR, Lu BY, Xu
u GD. Thermoelectric performance of poly(3,4-edioxythiophene)/poly(styrenesulfonate) pellets andElectron Mater 2011;40:64851.F, Liu CC, Xu JK, Jiang FX, Lu BY, Yue RR, Liuang JM. Simultaneous enhancement of electrical
tivity and Seebeck eoefcient of poly(3,4-edioxythiophene):poly(styrenesulfonate) lms treatedea. Chin Phys Lett 2011;28, 037201/1-4.
DK, Yang YA, Kung SC, McIntire TM, Penner RM. Enhancedelectric metrics in ultra-long electrodeposited PEDOTres. Nano Lett 2011;11:12531.a O, Khan ZU, Malti A, Braun S, Fahlman M, Berggren M,X. Optimization of the thermoelectric gure of merit in theting polymer poly(3,4-ethylenedioxythiophene). Nat Mater:42933.
Kim Y, Choi K, Grunlan JC, Yu CH. Improved ther-tric behavior of nanotube-lled polymer compositesoly(3,4-ethylenedioxythiophene) poly(styrenesulfonate).o 2010;4:51323., Feser JP, Chen CE, Majumdar A, Urban JJ, Segalmanter-processable polymer-nanocrystal hybrids for thermo-s. Nano Lett 2010;10:46647.Jiang FX, Huang MY, Lu BY, Yue RR, Xu JK. Free-standingPSS/Ca3Co4O9 composite lms as novel thermoelectricls. J Electron Mater 2011;40:94852.Y, Cai KF, Yao X. Facile fabrication and thermoelectricies of PbTe-modied poly(3,4-ethylenedioxythiophene)es. ACS Appl Mater Interfaces 2011;3:11636.CK, Fincher CR, Park YW, Heeger AJ, Shirakawa H, LouisSC, Macdiarmid AG. Electrical-conductivity in doped poly-e. Phys Rev Lett 1977;39:1098101.wa H, Louis EJ, Macdiarmid AG, Chiang CK, Heeger AJ.is of electrically conducting organic polymershalogenves of polyacetylene, (CH)X . Chem Commun 1977:.
B. Thermoelectric properties of conducting polymers.a MD: Carderock Div Naval Surface Warfare Center Press;ttp://handle.dtic.mil/100.2/ADA284058., Yoon CO, Lee CH, Shirakawa H, Suezaki Y, Akagi K.
tivity and thermoelectric-power of the newly processedtylene. Synth Met 1989;28:D2734.
H, Ishiguro T, Takahashi A, Tsukamoto J. Magnetoresistancermoelectric-power studies of metal-nonmetal transition indoped polyacetylene. Synth Met 1993;57:49005., Seol YH, Song YS, Park YW. Low temperature thermo-
power of the metal-halide doped polyacetylene. Synth Met:7478.
W, Plocharski J, Roth S. Anisotropy of thermoelectric-of stretch-oriented new polyacetylene. Synth Met:2536., Yoo JS, Park JY, Park YW, Akagi K, Shirakawa H. Positiveelectric-power of alkali-metal-doped polyacetylene. Synth5;69:614.
Y. Du et al. / Progress in Polymer Science 37 (2012) 820 841 841
 Yoon CO, Na BC, Park YW, Shirakawa H, Akagi K. Thermoelectric-power and conductivity of the stretch-oriented polyacetylene lmdoped with MOCl5. Synth Met 1991;41:1258.
 Maddison DS, Unsworth J, Roberts RB. Electrical-conductivity andthermoelectric-power of polypyrrole with different doping levels.Synth Met 1988;26:99108.
 Yan H, Ishida T, Toshima N. Electrical-conductivity andthermoelectric-power of polypyrrole with different dopinglevels. In: 20th International Conference Thermoelectrics. NewYork: IEEE; 1988. p. 3103.
 Sato K, Yamaura M, Hagiwara T, Murata K, Tokumoto M. Study onthe electrical-conduction mechanism of polypyrrole lms. SynthMet 1991;40:3548.
 Lee WP, Park YW, Choi YS. Metallic electrical transport of PF6-doped polypyrrole: dc conductivity and thermoelectric power.Synth Met 1997;84:8412.
 Hu E, Kaynak A, Li YC. Development of a cooling fabric fromconducting polymer coated bres: proof of concept. Synth Met2005;150:13943.
 Zotti G, Schiavon G, Zecchin S, Morin JF, Leclerc M. Electrochemical,conductive, and magnetic properties of 2,7-carbazole-based conjugated polymers. Macromolecules 2002;35:21228.
 Yue RR, Chen S, Liu, CC, Lu BY, Xu JK, Wang JM, LiuGD. Synthesis, characterization, and thermoelectric proper-ties of a conducting copolymer of 1,12-bis(carbazolyl)dodecaneand thieno[3,2-b]thiophene. J Solid State Electrochem 2011;doi:10.1007/s10008-011-1292-0.
 Aich RB, Blouin N, Bouchard A, Leclerc M. Electrical and thermo-electric properties of poly(2,7-carbazole) derivatives. Chem Mater2009;21:7517.
 Moreau C, Antony R, Moliton A, Francois B. Sensitive thermo-electric power and conductivity measurements on implantedpolyparaphenylene thin lms. Adv Mater Opt Electron 1997;7:28193.
 Moliton A, Bellati A, Lucas B, Ratier B. Thermoelectric Power Stabil-ity of the polyparaphenylene implanted with Cesium ions. SynthMet 1999;101:3512.
 Tamayo E, Hayashi K, Shinano T, Miyazaki Y, Kajitani T. Rubbingeffect on surface morphology and thermoelectric properties of TTF-TCNQ thin lms. Appl Surf Sci 2010;256:45548.
 Itahara H, Maesato M, Asahi R, Yamochi H, Saito G. Thermoelectricproperties of organic charge-transfer compounds. J Electron Mater2009;38:11715.
 Weyl C, Jerome D, Chaikin PM, Bechgaard K. Pressure-dependenceof the thermoelectric-power of TTF-TCNQ. J Phys 1982;43:116772.
 Harada K, Sumino M, Adachi C, Tanaka S, Miyazaki K. Improvedthermoelectric performance of organic thin-lm elements utilizinga bilayer structure of pentacene and 2,3,5,6-tetrauoro-7,7,8,8-tetracyanoquinodimethane (F-4-TCNQ). Appl Phys Lett 2010;96,253304/1-3.
 Jeszka JK, Tracz A, Kryszewski M. Thermoelectric-power in reticu-late doped polymers. Synth Met 1993;55:10914.
 Hiroshige Y, Ookawa M, Toshima N. High thermoelectric perfor-mance of poly(2,5-dimethoxyphenylenevinylene) and its deriva-tives. Synth Met 2006;156:13417.
 Hiroshige Y, Ookawa M, Toshima N. Thermoelectric gure-of-merit of iodine-doped copolymer of phenylenevinylene withdialkoxyphenylenevinylene. Synth Met 2007;157:46774.
 Kistenma Tj, Phillips TE, Cowan DO. Crystal-structure of 1-1 radicalcation-radical anion salt of 2,2-bis-1,3-dithiole (TTF) and 7,7,8,8-tetracyanoquinodimethane (TCNQ). Acta Crystallogr, Sect B: StructSci 1974;30:7638.
 Kwak JF, Chaikin PM, Russel AA, Garito AF, Heeger AJ.Anisotropic thermoelectric-power of TTF-TCNQ. Solid State Com-mun 1975;16:72932.
 Sakai M, Iizuka M, Nakamura M, Kudo K. Organic nano-transistorfabricated by co-evaporation method under alternating electriceld. Synth Met 2005;153:2936.
 Maddison DS, Roberts RB, Unsworth J. Thermoelectric-power ofpolypyrrole. Synth Met 1989;33:2817.
 Kemp NT, Kaiser AB, Liu CJ, Chapman B, Mercier O, Carr AM, Tro-dahl HJ, Buckley RG, Partridge AC, Lee JY, Kim CY, Bartl A, DunschL, Smith WT, Shapiro JS. Thermoelectric power and conductivityof different types of polypyrrole. J Polym Sci Part B Polym Phys1999;37:95360.
 Kemp NT, Kaiser AB, Trodahl HJ, Chapman B, Buckley RG, PartridgeAC, Foot PJS. Effect of ammonia on the temperature-dependent con-ductivity and thermopower of polypyrrole. J Polym Sci Part B PolymPhys 2006;44:13318.
 Choi ES, Suh DS, Kim GT, Kim DC, Park YW. Magneto thermoelectricpower of the doped polyacetylene. Synth Met 1999;101:3756.
Research progress on polymerinorganic thermoelectric nanocomposite materials1 Introduction2 PANI-inorganic TE nanocomposites2.1 PANI as a TE material2.2 PANI-inorganic TE nanocomposites2.2.1 Physical mixing2.2.2 Solution mixing2.2.3 In situ oxidative polymerization/intercalation
3 PTH-inorganic TE nanocomposites3.1 PTH as a TE material3.2 PTH-inorganic TE nanocomposites
4 PEDOT:PSS-inorganic TE nanocomposites4.1 PEDOT:PSS as a TE material4.2 PEDOT:PSS-inorganic TE nanocomposites
5 Other polymers and polymer-inorganic TE nanocomposites5.1 PA as a TE material5.2 PPY as a TE material5.3 PC as a TE material
6 Summary and conclusionsAcknowledgmentsReferences