Lecture 8 Organic Semiconductors

Transcription

1 Lecture 8 Organic Semiconductors 1/78 Announcements omework 1/4: Pick it up from me today. The solutions will be online later today. omework 2/4: Is online now. Due Thursday October 18 th at the start of the lecture (10:00am). I will return it on the day of the midterm (October 25 th ). omework 2 consists of content covered in Lectures 5, 6 and 7 (charge transport, silicon, oxides). 2/78 1

2 Announcements Midterm Thursday October 25 th at 10:00am in STAG 211 Exam will last 80 minutes. The exam will start exactly at 10:00am! losed book and closed notes. You can, and are expected to, use a calculator. hoose 2 out of 3 questions. If you answer 3 I will take best 2 scores. It will contribute 30% of overall grade for class. The exam will material covered in lectures 1-9 (inclusive). There will be 25 marks per question. 50 total. 3/78 Additional Information The course textbook (Brotherton) hapter 10 covers the basics of this subject. This is another very active area of research, so there are plenty of reviews on the subject. ma ma /cs/b909902f#!divAbstract 4/78 2

3 Lecture 8 Organic hemistry Background. Electronic Structure of Organic Semiconductors. Organic Thin-Film Transistors. OTFT Performance. 5/78 Organic hemistry Background 6/78 3

4 Organic ompounds When we hear organic, we normally think: When we (scientists) say organic compounds, we generally mean compounds containing at least one carbon atom. They are normally drawn like this: 7/78 Organic ompounds owever, the definition is not really that strict. 60 (fullerene) is generally referred to as an organic semiconductor: arbon nanotubes and graphene are not generally referred to as organic semiconductors: 8/78 4

5 ybrid Semiconductors Some compounds are described as being hybrid inorganic-organic compounds. Kim et. al. Sci. Rep., 2 (2012) 591. More on these in the next lecture. 9/78 Why Organic Electronics? So why would be interested in organic electronics? Plastics are organic compounds. Polypropylene Polyvinylidene hloride = = Polyethylene = Polystyrene = And plastics have interesting mechanical properties 10/78 5

6 Why Organic Electronics? Rather than just having a plastic substrate, what if the whole device could be organic (plastic)? We could then in theory produce much more versatile electronics. Leading to electronics with a wide range of physical form factors. 11/78 Why Organic Electronics? The cost has the potential to be incredibly low. If we could produce plastic electronics at a comparable cost to consumer plastics, it could change everything. We are a long way off this at the moment. 12/78 6

7 Solar Power Being able to easily-deploy very cheap solar cells is also very attractive. 13/78 Organic Molecules Before we look at the properties of organic molecules, let s have a very quick refresh on the notation used in organic chemistry. arbon is a group 14 element (like silicon), and therefore has 4 valence electrons: arbon Electronic structure: = [e]2s 2 2p 2 14/78 7

8 Alkanes The organic molecules we are concerned with are all covalent. I.e. carbon atoms will be sharing electrons. The most simple set of organic compounds are the alkanes. Starting with methane ( 4 ): arbon arbon ydrogen ydrogen 15/78 Alkanes We would normally draw methane ( 4 ): as follows: Other alkenes would be drawn as: Ethane ( 2 6 ) Propane ( 3 8 ) Butane ( 4 10 ) 16/78 8

9 affeine Using this notation, things can get very complicated very quickly. Take caffeine ( 8 10 N 4 O 2 ) for example: In organic chemistry, we use the following notation: Sometimes called skeletal formula, line-angle formula or shorthand formula. 17/78 Shorthand Notation There are a few rules for shorthand notation. For us, we will just be interested in two: 1). ydrogens are Implicit. They bond to carbons via single bonds in any available positions. I.e. they are where they should be. N O O N N N O N O N N N 18/78 9

10 Shorthand Notation There are a few rules for shorthand notation. For us, we will just be interested in two: 2). arbons exist in unlabeled positions Any unlabeled vertex is now a carbon atom. Lines to nowhere are bonds to carbons. O N O N N N 19/78 Electronic Structure of Organic Semiconductors 20/78 10

11 Traditional Polymers onsider a traditional, insulating, plastic such as polyethylene (PE). Longhand Shorthand Brackets and an n mean this is repeating n times Every carbon bonds to 4 other atoms. Two carbons and 2 hydrogens (expect at ends). All electrons are involved in single covalent bonds, so there are no free (delocalized) electrons. PE is a good insulator. 21/78 Trans-Polyacetylene Now consider the following polymer (longhand): Each carbon bonds to 3 other atoms. What happens to the other valence electron? We will now quickly discuss how such a polymer could exist.? 22/78 11

12 sp 2 ybridization The ground state of carbon is: * denotes excited state = [e]2s 2 2p 2 Or =1s 2 2s 2 2p 2 If one electron is excited from the 2s to the 2p state the configuration becomes: * = [e]2s 1 2p 3 Or * =1s 2 2s 1 2p 3 This could be provided through extra energy during synthesis for example. In this case we now have one 2s electron and three 2s electrons to participate in bonding 23/78 sp 2 ybridization As we have discussed before, these states can hybridize into a linear combination of orbitals. For a material such as trans-polyacetylene, two of the three 2p orbitals, and the one 2s orbital, hybridize into three sp 2 orbitals: Or, in Dirac Notation, each electron is described as: หsp 2 = 1/3ȁ2s + 2/3ȁ2p 24/78 12

13 sp 2 ybridization These three orbitals are highly directional. It is energetically favorable for the three sp 2 orbitals to form a planar structure with 120 between each lobe. This explains the three valence electrons, but recall we have 4 valence electrons in carbon: * =1s 2 2s 1 2p 3 We have one extra 2p orbital. 25/78 sp 2 ybridization It turns out the remaining p orbital will orientate with lobes above and below the plane of the sp 2 orbitals. The sp 2 orbitals from each carbon atom strongly overlap with the two adjacent carbon atoms and one adjacent hydrogen atom. 26/78 13

14 sp 2 ybridization This strong, in-phase, orbital overlap results in strong covalent bonds, called σ-bonds. The remaining p orbital forms another weak bond with one of the adjacent carbon atoms. These weak bonds are called π-bonds. 27/78 onjugation Because this second, weak, bond is only made with one of each carbon s neighbors, we describe the polymer as having alternating single and double bonds along the backbone: An organic molecule consisting of alternating single and double bonds is called conjugated. Trans-polyacetylene is a conjugated polymer. 28/78 14

15 onjugation shorthand polyacetylene In reality, electrons don t really behave like this. The electrons delocalize across the backbone: Because these delocalized states are only half filled (one bond per every two carbon atoms), the electrons behave as if they are in a half-filled band. I.e. there is space for the electrons to move along the polymer backbone. I.e. the plastic conducts electricity! 29/78 Doped Polyacetylene In 1977 Alan MacDiarmid, Alan eeger, and ideki Shirakawa published a paper on doped transpolyacetylene. You need to induce charges into the polymer to get them to move. They did this via doping with AsF 5. 30/78 15

16 Doped Polyacetylene The conductivity was observed to be very high (~100 Ωcm -1 ) It won them the 2000 Nobel Prize. hiang et. al. PRL, 39 (1977) /78 Benzene Before we can talk about devices, let us discuss one more example system. Benzene ( 6 2 ) is a conjugated molecule: Longhand Shorthand It is a bit like curled-up polyacetylene. 32/78 16

17 Benzene As with polyacetylene, three sp 2 hybridized electrons participate in strong σ-bonding with adjacent carbons and hydrogens. Leaving the lobes of porbitals above and below the ring. And as before, these orbitals delocalize to form an electron cloud. 33/78 Benzene Orbitals At this point it is worth reminding ourselves what orbitals are. They are surfaces, in real space, of equal wavefunction amplitude. They are possible states that exist (as defined by quantum mechanics). They can be occupied with electrons. But they can also be empty, and ready to accept electrons, which then have the described wavefunction. 34/78 17

18 Benzene Orbitals For each molecule, a range of different orbitals exist. Each with a different energy. These states are degenerate The Pauli-Exclusion principle states that only 2 electrons can occupy each state. The states are filled according und s Rules. From low energy to high. These states are degenerate layden, et. al.,. Organic hemistry, Oxford University Press, (1999). 35/78 Benzene Orbitals Under normal conditions, the 6 lowest energy states are occupied with the 6 delocalized p-electrons The energy of the highest occupied molecular orbital is called the OMO. The energy of the lowest unoccupied molecular orbital is called the LUMO. layden, et. al.,. Organic hemistry, Oxford University Press, (1999). There is an energy gap between these states. 36/78 18

19 The LUMO The LUMO is the lowest energy state which is able to accommodate extra electrons. So we could add an electron to this state, and it would be transported. I.e. it behaves like the conduction band of a traditional semiconductor. layden, et. al.,. Organic hemistry, Oxford University Press, (1999). 37/78 The OMO Similarly, the OMO is the highest energy state that is completely full of electrons. So we could add a hole (remove an electron) to this state, and it would be transported. I.e. it behaves like the valence band of a traditional semiconductor. layden, et. al.,. Organic hemistry, Oxford University Press, (1999). 38/78 19

20 Organic Semiconductors So for a conjugated molecule we can picture the system as having the following density of states: LUMO+2 LUMO+1 E G LUMO E F OMO E G E E V E F OMO-1 OMO-2 39/78 Organic Semiconductors Most organic semiconductors have much more complicated density of states: Ishii et. al. Nanoscale, 4 (2012) /78 20

21 Organic Semiconductors And the delocalized molecular orbitals can be much more complicated than we have discussed so far: Kim et. al. RS. Adv., 5 (2015) /78 Organic Thin-Film Transistors 42/78 21

22 Organic Thin-Film Transistors Sometimes you will hear organic transistors being referred to as organic thin-film transistors (OTFTs). Alternatively they are referred to organic field effect transistors (OFETS). These terms mean the same thing. There a range of OTFT architectures, depending on location of electrodes, and dielectric. Bottom-Gate, Bottom-ontact (BGB) Top-Gate, Bottom-ontact (TGB) Substrate Gate Electrode Dielectric Bottom-Gate, Top-ontact (BGT) Dual-Gate, Bottom-ontact (DGB) Source / Drain Electrodes Organic Semiconductor 43/78 Organic Thin-Film Transistors We will not talk in detail about all the steps involved in forming OTFTs today. We covered much of the basics in Lectures 6 & 7. Many processes for OTFTs are still experimental. We will cover dielectrics in more detail in Lecture 10 We will cover electrodes in more detail in Lecture 11. Substrate Gate Electrode Dielectric Source / Drain Electrodes 44/78 Semiconductor Today we discuss deposition of the semiconductor. 22

23 Vacuum Deposition It happens there are already many commercial products based on organic electronic materials. These all contain organic light-emitting diodes (OLEDs). 45/78 Vacuum Deposition owever at present there are no commercial products that contain OTFTs. Yet there are plenty of proof-of-principle demonstrations using OTFTs. 46/78 23

24 Vacuum Deposition All of these devices are based on vacuumdeposition. Substrate holder Deposition rate sensor Substrate and mask To vacuum pump Substrate shutter Very similar system to we observed in Lecture 6. Only the metal source is now replaced by an organic source. E.g. pentacene: Source organic Thermal filament 47/78 Vacuum Deposition The stacking of small molecules such as pentacene are highly dependent on growth conditions. Salzmann et. al. ASNano. 6 (2012) The alignment of molecules has a strong effect on electronic properties of TFTs. 48/78 24

25 Vacuum Deposition Many factors affect the growth of small organic molecules on surfaces: Deposition Rate. Substrate temperature. hamber pressure. Surface energy of substrate. Purity of material. Oetritz et. al. Org. Electron. 14 (2013) /78 Vacuum Deposition Plastic Logic (ambridge spin-out) have a huge facility in Dresden. 50/78 25

26 Solution Deposition The great thing about organic semiconductors is that many are soluble in organic solvents. A bit like polystyrene in acetone: Often we have to use other organic solvents. 51/78 Spin-oating Because many organic semiconductors are soluble, one of the most widely-studied deposition techniques in laboratories is spin-coating /78 26

27 Spin-oating 1). Organic semiconductor (in solvent) is deposited in center of the substrate. Substrate is held in place with vacuum chuck. 2). Substrate is rotated slowly (200 rpm) to distribute material. 3). Accelerate the substrate to final speed (~5000 rpm). Spin the substrate at constant speed for 30-60s. Forms a uniform film and evaporates solvent. 53/78 Film Thickness It can be shown (we won t) that the final film thickness depends on the spinning speed via: Film thickness t 1 ω Angular velocity Actual thickness will also depend on: oncentration. Solvent evaporation rate Viscosity. Local temperature. Local humidity. 54/78 27

28 Printing Spin-coating is a very simple technique for use in the laboratory. But the goal of organic electronics is to one day be highly-scalable and incredibly cheap. For a lot of us, the end goal is printed electronics. This is a difficult engineering challenge to solve. There are a few different printing technologies: Gravure printing. Inkjet printing. Roll-to-roll printing. 55/78 Gravure Printing Gravure printing is a simple, physical deposition method. There are indentations in a roll, which define the material pattern. learly for such an approach the definition of features is going to be limited by physical features on the gravure roll. The deposition can be pretty inhomogeneous. Thiburce et. al. Adv. Electon. Mater. (2017) /78 28

29 Gravure Printing As can the performance: 57/78 Inkjet Printing Traditional inkjet printing should be familiar. 58/78 29

30 Inkjet Printing Inkjet printing for organic electronics has the same goals: desktop-printable flexible electronics. The research is in early stages. The coffee-ring effect is a notable issue: Teichler et. al. JM. 1 (2013) Teichler et. al. JM. 1 (2013) /78 Roll-to-Roll Printing But one of the ultimate goals of printed organic electronics is to be able to create products using rollto-roll printing methods. This remains a long way off commercialization. 60/78 30

31 OTFT Performance 61/78 Transport As we talked about earlier, transport in organic occurs through hybridized p-orbitals. Venkateshvaran et. al. Nature. 515 (2014) 384. So the overlap of these directional orbitals is very important, and the molecular packing is therefore very important. Unlike oxides, where disorder is less important. 62/78 31

32 Morphology So as you can expect, the micro-morphology of organic semiconductors is extremely important. Traditionally it has always been believed that crystallinity is good. There is a lot of evidence for this in vacuum-deposited films, like pentacene: Brotherton Fig /78 Morphology This was always also assumed to be the case in solution-processable materials (such as polymers) as well. Poly(3-hexylthiophene) (often just called P3T) is a good example material. The side-chains are not conjugated, and therefore do not contribute to transport. They are present to improve solubility. 64/78 32

33 Morphology So transport occurs through the conjugated thiophene faces Sometimes called π π stacking. Thiophene Because there is overlap due to π bonding Therefore how molecules pack on substrate maters: Edge-on Face-on 65/78 Morphology There is evidence to support this: This has been the long-held assumption about organic transistor design. Sirringhaus et. al. Nature. 401 (1999) /78 33

34 Morphology The morphology of organic semiconductors is normally extremely complex: Especially when multiple semiconductors are blended. Love et. al. Adv. Funct. Mater. 23 (2013) Brabec et. al. hem. Soc. Rev. 40 (2011) /78 Backbone Transport There have been some recent reports of charges being transported down the backbone of certain polymers. Potentially less dependence on morphology / order. Venkateshvaran et. al. Nature. 515 (2014) /78 34

35 LUMO OMO LUMO OMO LUMO OMO Electrons vs oles We have previously seen (Lecture 2) that the ability of charges to be injected and transported depends on the energetic alignment of the electrodes with the bands of the semiconductor: N-Type P-Type Semiconductor Electrodes Semiconductor Electrodes e - φ e - e - h + h + e - φ 69/78 P-Type OTFTs Semiconductor Electrodes So good hole-transporting (p-type) materials have good alignment of the OMO with the work function of injecting electrodes. φ 70/78 35

36 Unstable in air Stable in air LUMO OMO N-Type OTFTs So good electron-transporting (n-type) materials have good alignment of the LUMO with the work function of injecting electrodes. Semiconductor Electrodes φ 71/78 N-Type Materials It has been shown [1] that the standard redox potentials of water and oxygen with many organic semiconductors lies around 4 ev. This means that any organic semiconductors with a LUMO 4eV are likely to be unstable in the presence of water and oxygen (air). For this reason there are much fewer stable n-type organic semiconductors than p-type. 4 ev [1] de Leeuw et. al. Synth. Metals. 87 (1997) /78 36

37 harge Transport Many organic semiconductors show a gate-field dependent mobility. For this reason, and the fact that, theoretically, all transport must occur via hoping between overlapping orbitals, variable range hoping (VR) is normally used to model transport in OTFTs. E e e e e e e 73/78 OTFT Mobility OTFTs have now been studied for > 30 years, so the field is quite mature. Labram et. al. Small. 11 (2015) /78 37

38 OTFT Mobility Unfortunately, some of the most recent values > 10 cm 2 /Vs are not reliable. The gradual channel model is being applied when it is not valid to do so. L di D μ lin = W ox V D dv G μ sat = 2L W ox d I D dv G 2 Paterson et. al. Adv. Mater. (2018) /78 OTFT Mobility When plotting mobility vs gate voltage for OTFTs, normally some sort of dependence is observed. Paterson et. al. Adv. Mater. (2018) Unfortunately, some groups will take the peak value, even when it is not appropriate to do so. 76/78 38

39 OTFT Mobility So what is the real mobility of OTFTs? urrent values are only lab based, as these devices have not yet been commercialized. But the very best, reliable, values are roughly: μ h 15 cm 2 /Vs. μ e 5 cm 2 /Vs. For the few prototype commercial devices we have seen, we can probably say: μ h cm 2 /Vs. But progress continues to be made! 77/78 Next Time arbon-based systems: Other material systems. 78/78 39